A buoy-only sea surface temperature record

by Zeke Hausfather and Kevin Cowtan

A buoy-only sea surface temperature record supports NOAA’s adjustments.

Significant recent media and political attention has been focused on the new NOAA temperature record, which shows considerably more warming than their prior record during the period from 1998 to present. The main factor behind these changes is the correction in ocean temperatures to account for the transition from ship engine room intake measurement to buoy-based measurements and a calibration of differences across ships using nighttime marine air temperatures (NMAT). Here we seek to evaluate the changes to the NOAA ocean temperature record by constructing a new buoy-only sea surface temperature record. We find that a record using only buoys (and requiring no adjustments) is effectively identical in trend to the new NOAA record and significantly higher than the old one.

The changes to the prior NOAA global land/ocean temperature series are shown in Figure 1. There are some large changes in the 1930s that are interesting but have little impact on century-scale trends. The new NOAA record also increases temperatures in recent years, resulting a in a record where the period subsequent to 1998 has a trend identical to the period from 1950-1997 (and giving rise to the common claim that the paper was “busting” the recent slowdown in warming).

Figure 1

Figure 1: New and old homogenized global land/ocean records from Karl et al, 2015.

The paper that presented the revised record, Karl et al, didn’t actually do much that was new. Rather, they put together two previously published records: an update to the NOAA sea surface temperature record (called ERSST) from version 3 to version 4, and the incorporation of a new land record from the International Surface Temperature Initiative (ISTI) that makes use of around 32,000 land stations rather than the 7,000 or so GHCN-Monthly stations previously utilized. The new land record is quite similar to that produced by Berkeley Earth, though it has relatively little impact on the temperature trend vis-à-vis the old land record, particularly during the recent 1998-present period.

The slowdown-busting nature of the Karl et al paper relies almost entirely on the update from ERSST v3 to v4. During the post 1998 period, this is primarily due to a revised treatment of buoys and ship engine room intake (ERI) measurements and an improved calibration of differences across ships. During the past few decades the number of automated SST measurement buoys has expanded rapidly from effectively zero before 1980 to over 70 percent of all SST measurements today as shown in the figure below. Buoys are appealing measurement platforms, as they are not restricted to shipping routes and often have fully automated reporting via satellite uplink.

Figure 2

Figure 2: Share of SST observations by instrument type from Kennedy et al 2011. Note that this figure ends in 2006; since they buoys have continued to grow in observation share.

NOAA argues that the transition to buoys introduced a spurious cooling bias into the record. ERIs tend to warm the water a bit before measuring it (ship engine rooms being rather hot), whereas buoys do not. They identify a bias of around 0.1 C between buoys and ERIs and remove it by adjusting buoy records up to match ERI records in ERSST v4, as well as use NMAT readings to calibrate the differences across ships. These adjustments had not been done in the prior ERSST v3. As an aside, the decision to adjust buoys up to ERIs or ERIs down to buoys should nominally be trend neutral. Indeed, in their work on HadSST3 Kennedy and colleagues explicitly tested this, and found “no appreciable difference” on trends.

However, there is a rather straightforward way for us to test if the adjustments done in ERSSt v4 are proper or not: compare their adjusted record to a record made only from buoys. The buoy records are from purpose built instruments which are largely standardized, resulting in much more homogeneous records. On the other hand, the buoy record is short, and has limited coverage in the early 90’s.

The buoy-only record is prepared by calculating daily averages for each buoy. Buoys which show a large daily temperature variation are rejected: in deep water the daily temperature range is only a few tenths of a degree, but in very shallow water it can be substantial which presents problems when some data are missing. Next, the daily data are placed into 550 km equal area grid cells based on the location of the buoy for that day, and monthly averages are determined for each cell.

The resulting coverage is still limited and so produces a biased estimate of global sea surface temperature. To produce a useful comparison to ERSST, we therefore reduce the coverage of the ERSST datasets to match the buoy dataset (now using a fine 1 degree grid for all the data) and then calculate anomalies for all the datasets using a 2001-2010 baseline. The area weighted mean temperature is then calculated for each record. While this doesn’t provide a very good estimate of global SST, it does allow a strict like-with-like comparison against ERSST over the regions where they buoys have coverage. The percent of global ocean covered by buoy measurements varies from around 40% in the mid 1990s to around 70% in recent years.

combined buoy sst trends full

 

Figure 3: ERSST v3, v4, and Buoy-Only SST anomalies and trends from 1995 through the end of 2014. The trend periods shown are the full record (1995-2014) and the “hiatus” period (1998-2014). 2015 is excluded as the year is incomplete, and the period prior to 1995 is excluded due to limited buoy coverage. The anomaly graph is baselined to 1995-2005 to show the time-evolution of differences.

As shown in Figure 3, a buoy-only record is quite similar to the ERSST v4 but shows much more warming than ERSST v3 during the period from 1995 to present. This suggests that ERSST v3 suffered a cooling bias when blending buoy and ship records that is properly corrected in ERSST v4, at least for the areas where both ship and buoy records are available. Because the buoy record is relatively homogenous and requires no adjustments, it provides a good check in the validity of the combined ship-buoy series when normalized for spatial coverage.

The ship records are important because they form the foundation for a long sea surface temperature record, but they require careful calibration. The differences between HadSST3 and ERSSTv4 suggest that the finer details of the ship record are not yet settled, and as a result care is required especially when considering short term trends. However the buoy data support the NOAA claim that ERSSTv3 suffered a significant cool bias over recent years arising from inhomogeneities in the ship record and the increasing use of buoys.

Code for downloading and processing the data for this analysis is available here. Note that the underlying buoy dataset is large (approximately 44 GB). Gridded 1×1 files are also provided at for buoy, ERSSTv3, and ERSSTv4 data.

JC comments

Zeke and Kevin have taken a useful first step in deciphering what is going on with NOAA’s new ERSSTv4 data set.

The global ocean trend from 1998-2014 is cited in Karl et al. SOM as 0.097 C/decade for ERSSTv4 and 0.038 C/decade for ERSSTv3b.  In Figure 3 above, the red dot for ERSSTv4 for the regions that are sampled by buoys is about 0.12 C/decade.  Note also the slightly lower trend in Figure 3 for the buoys only (~0.102 C/decade).

There are substantial changes between v3 and v4 during this period, and the buoys certainly seem to contribute to a larger trend.

It is instructive to compare the change between HADSST2 and HADSST3, which also included a buoy adjustment in HADSST3:

HADSST2:

Slide1HADSST3:

Slide2The Huang et al. paper (from NOAA) does some comparisons of ERSSTv4 with HADSST3 for longterm trends and a few regions, but not for this particular periode. The relevant period is compared in this plot by Bob Tisdale, showing a substantially larger trend for ERSSTv4:

Slide1

You may not be aware that NOAA has another SST dataset – the OISSTv2 – for the period since the 1980’s.  This data set has higher horizontal resolution owing to the use of satellites, and references satellite and ship observations to the buoys.  Serendipitously, I spotted this article  by NOAA this morning on twitter: Exactly the same but different: why we have so many different ways of looking at sea surface temperature.  It’s a good article, and describes the advantages of OISST v2.  Criticisms of OISST are that the satellite data aren’t consistent across the period.  Well, the satellites have been pretty consistent since 1995, and the point of the buoys is to calibrate the satellite observations to adjust for any satellite inconsistencies, volcanic eruptions, etc.

Below is a plot of the  OISST v2 (provided by Bob Tisdale).

Slide3

Bob Tisdale cites a trend in OIv2 of 0.041 C during the period 1998-2014.

The bottom line seems to be that the NOAA ERSSTv4 1998-2014 trend is about twice as large as the trend for HADSST3 and OISSTv2.

So, why are these 3 analyses, primarily based on buoys during this period, so different?  A way to approach understanding this would be to do the same buoy masking on the HADSST3 and OIv2 data sets; this would help assess/understand the differences.

p.s.  Given the controversial nature of the Karl et al. paper, and the fact that I had written previously on this topic, I thought it best to include my comments in the main post, rather than in the comments thread.  Since this is a guest post, please keep your comments relevant and civil.

492 responses to “A buoy-only sea surface temperature record

  1. Pingback: A buoy-only sea surface temperature record | Enjeux énergies et environnement

  2. daveandrews723

    What I take away from all of this… they don’t have a clue about actual temperatures. All of the systems seem very suspect.

    • The real world is complex, and the point of science is to understand it. Remember, all of this is discussing trends over a relatively short period, which are highly sensitive to changes in the underlying data.

      • Zeke H, what does 1 degree area grid mean? You don’t use kriging with the buoys?

      • Hi fernando,

        A 1 degree grid refers to a grid cell resolution of 1 degree lat by 1 degree lon. In this case, you’d average the values (generally anomalies) of all stations in a grid cell to estimate the grid cell value. It’s a bit more basic than an optimal interpolation technique like kriging, but will give a pretty similar result when data coverage is not sparse (or, as in the case of this post, when we limit the analysis to areas with coverage across multiple datasets). Its also much easier to implement.

      • Curious George

        Just curious – how many grid cells are there between 89N and 90N?

      • David Springer

        Can someone please explain the 2006 step change visible when overlaying ERSSTv4 on OISSTv2? The records largely agree up until that point. The step change is the “pause buster”. ERSSTv4 has no credibility in light of this singular unexplained difference.

      • I would also appreciate any insights into the 2006 jump

      • David Springer

        The step change is visible here too in the left graph:

      • David Springer:

        Whatever gives you the idea than OISSTv2 monthly is the ultimate SST? It does not incorporate a ship-bouy bias adjustment. OISSTv2 daily does and OISSTv2 daily has a higher temperature increase than ERSST4.

        Why not compare OISSTv2 monthly vs OISSTv2 daily?

        Your step change is not so evident either.

        The effect of the bias adjustment increases when the fraction of buoys increases. No surprises there.

      • The change David Springer notes here is in line with the Karl et al paper comments that states that the total trend correction 2000 to 2014 is 0.064 degrees C per decade. A difference plot might not show a distinct jump but as the Springer comparison shows most of the change occurred in short time period. Also that Springer shows that the increase in temperature change from ERSST v3 to v4 is essentially the same for the bouy series would point to the importance of the colocation analysis of buoy and ship data. It appears that ship data of recent time has been overwhelmed by the bouy data by number and weighting and over a short time period. The findings would also be in line generally with the Hausfather and Cowtan analysis that I do not believe explained the quick change.

        I am looking at determining the proper confidence intervals for the Karl trends and attempting to find the CIs for the ship and buoy colocated data.

      • Judith should give ken a post when he finishes his analysis. It seems that he is a little more thoughtful and thorough than the average bear.

      • kenfritsch:

        So there was no step change in 2005 then. Or 2006 as Judith saw it. They cannot even agree on the time for the step change.

        Even so. If there were a bigger deployment of bouys one year, that would give av bigger bias adjustment from that year. Would that invalidate the needed bias-adjustment?

      • Between the main Karl (2015) paper and its Supplementary Materials (SM) the breakdown of the 2000-2014 trend increase of 0.064 C per decade can be attributed to 4 factors:

        (1) Continuing the ship bucket corrections beyond 1941 accounts for 0.030 C per decade whic is manifested primarily from a cooling of the ship data from 1998-2000.

        (2) The 0.12 C added to all the buoys in the data set and increasing portion of buoy data accounts for 0.014 C per decade.

        (3) Higher weighting of the buoy data and again increasing portion of buoy data accounts for 0.012 C per decade

        (4) the remaining 0.008 C per decade is accounted for by the remaining 8 corrections made.

        The Karl SM also points to the buoy data anomaly tracking that of the ERSST v4 anomaly – so I guess perhaps a quick reference to that by Hausfather and Cowtan would have been sufficient and then followed by what I consider the main issues of Karl (2015) that need analyzing, i.e. (1) rationalizing the statistics and uncertainty these adjustment noted above might impart to the final data and conclusions and (2) the correct handling of the confidence intervals for the trends in Karl (2015).

      • David Springer

        The timing of the step change appears to coincide with ARGO becoming operational. IIRC ARGO reached 2000 floats in 2006 then in 2007 reached its design goal of 3000 floats. It is now 4000 floats. Looks like ARGO data was tortured into global warming compliance between ERSSTv3b and ERSSTv4.

      • David Springer

        ejac writes: “Whatever gives you the idea than OISSTv2 monthly is the ultimate SST?”

        What gives you the idea that Engine Room Sea Surface Temperature (ERSST) isn’t the worst? LOL

      • It’s the Zeke and Deke show, minus Deke, plus Karl. Hi guys, no disrespect intended. You may remember me from the NWS Climate Forum. I have several concerns with these and other analyses. I) In Fig. 1, the warming during the 30’s was at least was warm as now (+.05), but is missing from the plot. II) The surface record continues to diverge from the satellite/balloon record. III) Where are the uncertainty estimates? IV) What about the Argo data trends, pretty flat except for El Ninos of 2010 and 2015. V) The USCRN data show the same trend for last 10 years, close to zero. I know it’s U.S. only but that brings into question the veracity of the global surface record. I trust the satellite and USCRN, the “adjustments” in the measured surface data are counter to everything I learned about respect for observations as ground truth. JerryG∞

      • David Spinger, the Karl (2015) Supplementary Materials states that Argo buoys were not used in their analysis. They say they might include those buoys in a future version ERSST.

        By the way I have not been able to duplicate the Karl trends or CIs for 2000-2014. I am using the KNMI ERSST v4 data set in the annual format and obtaining larger trends and much larger CIs. The CIs were determined using 10000 simulations of an ARMA(1,0) model for best fit aid score. The. white noise std dev was 0,,0734 and the ar1 was 0.474. Has anyone here attempted to replicate the Karl trends?

      • Keep us posted on your calcs, let me know if you want to do a guest post on this at some point.

      • David Springer:

        ARGO is not used in these SST-indexes.

        Perhaps they should be used. Surface ARGO matches ERSST4.

      • Jerry Gorline:

        Radiosondes diverges from the SST-indices. That is correct. Higher trend. They also diverge from RSS.

        https://tamino.files.wordpress.com/2015/09/rssminusrat.jpeg?w=750&h=498

        And ARGO is not flat. Surface ARGO matches ERSST4.

      • David Springer

        Still waiting for someone (ejak just dodges the question) to explain how the pause was negated in a single step change in temperature circa June 2006 added in an update from ERSSTv3b to ERSSTv4.

        Engine intakes on ships are ridiculously unsuited to the task in any case. Spatial coverage confined to shipping lanes is abysmal, calibration on and between ships non-existent, depth of intake varies greatly, wholly inadequate precision and accuracy.

        Yet another pathetic attempt to manufacture data for an imagined threat to civilization. Spare me.

      • I now have the annual data that matches that shown by Hausfather in graph of ERSST v4 in the introduction of this thread. I now duplicate the trend reported in Karl (2015) for the period 2000-2014. I should have used Geert’s KNMI conversion from monthly to annual from the start.

        The best fitting arma model was ARMA(1,1) with ar1=0.181, ma1=0.484 and with a white noise standard deviation of 0.0683. Using a Monte Carlo approach with this model and 10,000 simulations I obtained the following probabilities and trends in degrees C per decade:

        1% =-0.065; 2.5%=-0.038; 5%=-0.017; 10%=0.008; 50%=0.097; 90%=0.184; 95%=0.211; 97.5%=0.233; 99%=0.258

        From this and using a 2 sided hypothesis test that the trend for this period using the ERSST v4 mean global series is greater than 0 is not significant at the 90% level and barely so at the 80% level. Please recall that this uncertainty is only part (though a major part) of that reported by Karl as total uncertainty.

        If I can do other trend period analysis in a timely fashion I’ll report the results in a comment to this thread. I would dearly like to get hold of the data that compared the co-located ship and buoy data and estimate the uncertainty that correlation carries into the ERSST v4 final data.

      • I believe I now have the data I need to estimate the measurement and coveage uncertainty that Karl (2015) uses to obtain a total trend uncertainty. It is in Kennedy et al. (2011). NOAA has their own estimates that Karl uses, but I have not found it yet. Kennedy’s method should be general enough to apply here. The Karl paper lists the total uncertainty in trends in Supplemental Materials only.

      • I have attempted to reconcile the Karl trends and the CIs for those trends without success and will now have to contact Karl, as the lead author of Karl (2015), about the details of their calculations used in this paper.

        I obtain substantially larger CIs and somewhat smaller trends using either a linear regression or a spline smooth for determining the trend and using that trend to subtract from the ERSST V4 series to determine the residuals. From these residuals I obtain an ARMA model from which I can do Monte Carlo simulations to estimate CIs for the trends. For the spline smooth (df=7 and spar=0.75) I can use the entire 1854-2014 time period while for the linear regression trend I use the most linear part of the series and that being 1951-2014. The trends and CIs agree well using either method and mainly because the period analyzed follows a reasonably linear trajectory. I prefer, as a general approach to problems such as this one, to use the spline smooth method or alternately the Singular Spectrum Analyses decomposition and reconstruction method which do not assume the secular trends are linear.

        What I have found by simulating a long series of 161 years using an ARMA model – similar to what I found by the above described methods for the residuals – and then finding the best fitting ARMA model using the last 15 and 17 years of that series is that the shorter 15 and 17 years are too short to see most of the autocorrelation effects of the 161 years series and thus the best fitting model in the shorter series lies closer to a white noise domain, i.e. ARMA(0,0). This is what I found when modeling the last 15 and 17 years of the ERSST V4 series.

        If I restricted the ARMA model to ar1, as was the stated case for the Karl method of determining the trends CIs, the CIs would be substantially smaller if Karl used the 1998-2014 and 2000-2014 time periods for determining the ar1 than if the longer periods were used as I prescribed above. Even when I attempt to estimate CIs by what I judge to be the method used in Karl (2015) the CIs are larger than those reported by Karl.

        If indeed the ARMA models that best fit the longer and shorter ERSST V4 series were as different as one would measure using the longer and shorter time periods one would have to suspect something very different was affecting these 2 periods – like the adjustments made. I do not think that is the case, but rather the limitation of ARMA modeling short time series – as my simulations showed.

    • Dave I get a different take, that all they have are clues, which they have to tease out of the data. One point of question is how they tease out those clues. When the methods are beyond my capability to fully evaluate I have trust the people doing it. Zeke has that trust for me.

      I do get the feeling our level of understanding is far less robust than some people like to claim.

    • ZH: i see. Maybe some day I’ll stop by on my way to Oregon. Did you try making a movie of the lapse rate changes over time over the last 30 years? Compare that with the lapse rate for different models? I’m just wondering whether your work is used to trouble shoot model performance.

    • Judith,

      There may not be any jump. Look at the graph posted by ehak (@ehak1), where he/she subtracts Oiv2 monthly for ERSST4, and see if you can still discern a 2006 jump rather than a progressive (though noisy) increase from circa 2000 to circa 2012.

  3. That means ALL readings done by ship intake need to be “adjusted”, which means there is not temp increase. NOAA are crooked.

  4. David Springer

    “shows considerably more warming”

    Shocker. What are the odds? /sarc

    • Shows more warming? : ” ERIs tend to warm the water a bit before measuring it (ship engine rooms being rather hot), whereas buoys do not. They identify a bias of around 0.1 C between buoys and ERIs and remove it by adjusting buoy records up to match ERI”. So they adjust the one that is not biased, so that it matches the one that is biased.

      • Raising buoy values to ships or lowering ship values to buoys is nominally trend-neutral. As mentioned in this post Kennedy tested the order of combination for HadSST3 and found no meaningful difference.

        Here is a toy model example that roughly matches what happened in the real world showing this:

      • ” ERIs tend to warm the water a bit before measuring it (ship engine rooms being rather hot)”

        You’ve never done this have you.

      • Hi, Zeke. Here is the problem I have with the Karl, et al., paper. It is trivially obvious that it is better to adjust the less certain ship data to match the more certain buoy data. So given a choice between the two, one would naturally choose to adjust the ship data. They chose to adjust the buoy data. Their justification for this is, it doesn’t matter for the long-term (“global”) trend which choice they make.

        But the paper’s central result is not about the long-term trend; it is about the short-term trend (the hiatus). Does the choice matter for the short-term trend? We don’t know. If it doesn’t matter, why adjust the robust data to match the less robust data? We don’t know.

        The “it doesn’t matter” statement is choice-neutral. The quality of the data points towards adjusting the ship data. Combine these two together … and they adjust the buoy data? It doesn’t make any sense. And so one is left to conclude that it DOES matter, for the central result of the paper (“there is no hiatus”), or perhaps for some other aspect of the paper, and that their conclusions are not robust to this choice.

      • Ted Carmichael:

        But the paper’s central result is not about the long-term trend; it is about the short-term trend (the hiatus). Does the choice matter for the short-term trend? We don’t know.

        It appears that the choices did matter for the hiatus. From the Karl paper:

        Of the 11 improvements in ERSST version 4 (13), the continuation of the ship correction had the largest impact on trends for the 2000-2014 time period, accounting for 0.030°C of the 0.064°C trend difference with version 3b. (The buoy offset correction contributed 0.014°C dec−1 to the difference, and the additional weight given to the buoys because of their greater accuracy contributed 0.012°C dec−1.

  5. In one of my first comments here, I pointed out and agreed with Jim D that the pause was only such if using short term trends. Using a longer trend would show an overall upward trend just lower than the 1970 to 1998 trend. I never used the word pause in the description.

    Using the word pause was a very handy tool to flummox the warmists. As they went through all kinds of contortions to find missing heat, it only became more important propaganda for skeptics. The pause is ruining the cause. But it was just a ruse.

    The more important message out of all of this is that the overall trend is much lower than the IPCC models. End of story for me.

      • The 60-year observed trend, during which we have had 75% of the emissions and the best CO2 measurements, is consistent with a TCR of 2 C per doubling. The models are consistent with a TCR of 2 C per doubling. End of story.

      • Funny that:

        Guess it’s how you do the math.

    • Except there is no observed trend, because global temp is not observed. These are models.

    • The more relevant issue is the trend to the adjustments.

      I’m fine with adjustment. I’m fine with correcting the adjustments to some extent.

      The above chart shows that the whole thing has become a frickin game to the government climate scientists.

      1. Algorithms affecting historic data should only be allowed to be adjusted once every 5 or 10 years.
      2. The changes (all of them) should be reviewed by an engineering/statistician team familiar with the issues. Changes that are considered incorrect or a bad joke would be rejected.
      3. The approved changes will then be implemented.
      4. The timing should be set so the updates occur after IPCC meetings.

      The whole concept of pause buster adjustments is just ludicrous.

      The timing of these adjustments before IPCC meetings is beyond suspicious.

      And for the adjusters time should be running out. When the adjusted trend (already 56% of the real trend) becomes equal to the data trend we should RIF the departments involved and outsource it with the instructions “don’t play with the damn algorithms – just make the chart”.

      And there you go.

      • Anyone who has worked in business understands the importance of the “smell” test. When things just don’t seem right, feel wrong, then you are missing some crucial data. It may be correct, but you are still missing some crucial data, and the more the originators of the thing tell you that nothing is wrong, but do not supply the missing data, the more sure you are that there is a PROCEDURAL problem they are not discussing.

        The unidirectional and progressive changes do not “smell” right. The continual and varied explanations don’t smell right either. I’m still stuck with the procedural problem – a general technique, a base assumption that always chooses the warming side, is deep in the procedure.

        I wrote a blog on the types of certainty or perceived reality. This is a type I call Computational Certainty/Reality. The numbers add up to this. No matter what you do once you have the numbers, this is the result. Representational Certainty/Reality is what is actually going on. The difference lies in how you go from the event, i.e. the temperature the medium is, to the number you slot in the spreadsheet.

        I’ve seen so many unidirectional, progressive changes in business programs for businesses that fail. The fundamental error is in a bias of optimism. Here, the bias would be for warming.

        People do not assess uncertainty in a neutral way. They choose the one unconsciously (or consciously) that fits with their expectation (“reasonable situation”) or Emotional (“I’m afraid the world is warming”) or Ideological (“Capitalism is the devil’s system/Consumerism is Destroying the planet with CO2”) Certainty/Reality. When there is a choice, they do not choose to go up one day and down the next.

        I could be wrong. But the unidirectional, progressive AND substantial nature of the GISTemp series adjustments bothers me. If it weren’t for the adjustments, the whole CAGW story would be a non-starter. The significance of the adjustments is a huge red flag for those recognizing that interpretation, not observation, is the difficult aspect of human endeavors.

      • Those adjustments do not look random. Hard to prove it, though.

      • douglasproctor | November 23, 2015 at 6:52 pm |
        Anyone who has worked in business understands the importance of the “smell” test. When things just don’t seem right, feel wrong, then you are missing some crucial data. It may be correct, but you are still missing some crucial data, and the more the originators of the thing tell you that nothing is wrong, but do not supply the missing data, the more sure you are that there is a PROCEDURAL problem they are not discussing.

        +10

        Most global warmers have never held a useful or productive job in their lives and have no sense of smell.

      • Well, it should be even warmer, so yeah, puzzle pieces are missing.

      • edbarbar | November 23, 2015 at 10:10 pm |
        Those adjustments do not look random. Hard to prove it, though.

        Well, the adjustments have been monotonically upward in since 2000. The post 2008 changes have been captured. What are the chances that historic temperature trend of dead data increased for 15 consecutive years?

        The change is 0.25°C for the 90 year period 1910 to 2000 over a 7 year (2008 to 2015).. Lets do the math: (0.25°C * 100 years*century-1/90 years)* 100 year*century/7 years = 3.97 °C/century2 (degrees celcius per century per century) or 7.94 °C by 2100.

        I have had it with the adjusters. They have only added 0.25°C so far but have 7.44 °C of damage left to do by 2100. They should not be allowed to condemn us to a hot steaming CGAGW hell. We should fire the adjusters immediately, terminate their pension rights, and ban them permanently from government service and grants. If we take quick action against global warming adjustment, we may yet be able to save ourselves.

      • PA: Like I said, it looks unlikely, and I have a strong suspicion they are not some random walk making the record more accurate. However, how to prove it?

      • edbarbar | November 24, 2015 at 1:39 pm |
        PA: Like I said, it looks unlikely, and I have a strong suspicion they are not some random walk making the record more accurate. However, how to prove it?

        As I understand it they are still within the error bounds of the data. What they are effectively doing is pushing the present toward the upper bound and the past to the lower bound.

        Given that the record was virtually unchanged for literally decades the argument that they have to change the historic record on a daily or monthly basis now is mindlessly dishonest and irrational.

        I don’t care what they are doing as long as it stops. Congress should pass a law banning changes to the official GISS/NOAA historic record (pre-2010 data) one year from today. GISS/NOAA would have one year to play stupid temperature games and then it is game over.

        After the ban the record could only be altered by act of congress. Alteration of the record after the ban by executive branch employees would result in dismissal and loss of pension.

        And there you go, problem solved.

      • “I don’t care what they are doing as long as it stops. Congress should pass a law banning changes to the official GISS/NOAA historic record (pre-2010 data) one year from today. GISS/NOAA would have one year to play stupid temperature games and then it is game over.

        After the ban the record could only be altered by act of congress. Alteration of the record after the ban by executive branch employees would result in dismissal and loss of pension.”

        What dumb idea. You really are living in a fantasy land. Do you also have a problem with UAH adjustments?

    • If you look at actual historical ice core data and look at the warming from all the cold periods into the warm periods. The same warming easily accounts for all the warming we have had and some margin for more that we might get. That leaves nothing left for CO2 to account far. This accounts for a sensitivity of zero, plus or minus some tiny amount.

      If you look at the temperature regulation of the past ten thousand years, it is very clear that it over powers everything else in setting upper and lower bounds.

      http://popesclimatetheory.com/page85.html

      If CO2 adds energy, it will not matter, when the Polar Oceans Thaw, snowfall is increased and the upper bound of temperature is not violated.

      The major warm periods between the major ice ages did get warmer because less ice was on land on Greenland and Antarctic and Mountain Glaciers during those warm periods which allowed the oceans to get warmer and higher. Then this more warm and high oceans provided snowfall for major ice ages. We do not now have enough water in the oceans for a create a major ice age. We have enough ice on land now to prevent a major warm period. The temperature bounds of the most recent ten thousand years is the new normal

      This warm period, like the Roman and Medieval Warm periods, will bounce up and down near the same upper bound for a few hundred years while the ice rebuilds on Antarctic, Greenland and the Mountain Glaciers.

      There will be more record warm years and record open Arctic year for a few hundred years to come but it will not get much warmer. The warmer the oceans get, the more sea ice melts and produces more snowfall.

      They will continue to monkey with the data, but too many are watching, they can only monkey so much and that will not be enough.

      Only climate models will continue to go up at alarming rates. Real data has a real upper bound. The thermostats turn on cooling when the Polar Oceans Thaw, every time. The thermostats are the Polar Oceans.

      • popesclimatetheory…

        We are going to have to disagree (a little).

        Queuing analysis says that if you shorten the mean absorption distance there will be more energy in the atmosphere.

        However this doesn’t indicate how big the effect is. Using a swimming analogy – we know someone is going to jump into the pool and we know he is going make progress down the lane – assuming he doesn’t drown or get distracted. The global warmers claim the swimmer is Mark Spitz in his prime. The real swimmer is unknown..

        The evidence (22 PPM = 0.2 W/m2) indicates that there is someone in the pool but he swims about as well as I do (I get get from one end of the pool to the other about half the time without drowning).

  6. David Springer

    In 2008 the ARGO record was a cooling trend.

    http://earthobservatory.nasa.gov/Features/OceanCooling/

    • It always gives me a warm fuzzy feeling when my government uses words to disparage those who pay their salaries as NASA did using the word “denier”. They just couldn’t help themselves.

      I noticed they admitted to still not being able to close the sea level rise budget when trying to reconcile with the altimetry data. Good luck with that one.

  7. So…just what did happen in 1910?

  8. Thank you for putting up this post Dr. Curry. I like to hear both sides and when you post it I know it’s genuine not spin. Also it enhances your reputation as honest broker.

  9. I’m afraid there are some unresolved questions concerning the OISSTv2 trends. I downloaded the OISSTv2 daily anomaly grids from ftp://ftp.cdc.noaa.gov/Datasets/noaa.oisst.v2.highres/
    and integrated them down to monthly temperatures using CDO. I get a trend on 1998-2014 of 0.13 C/decade, which is about 3x Bob’s value and in good agreement with ERSSTv4. However I would like to redo the integration with my own code as a cross check.

    I understand from a contact in the field that there are two different versions of OISSTv2 – an older monthly product, and a newer daily product. The older version does not include adjustments for ship biases. I think KNMI has the older version. Nick Stokes also quotes a weekly product – I don’t know which one that is related to. However I haven’t checked any of the other versions myself, and I may have garbled the explanation.

    • Kevin says: “I’m afraid there are some unresolved questions concerning the OISSTv2 trends. I downloaded the OISSTv2 daily anomaly grids from ftp://ftp.cdc.noaa.gov/Datasets/noaa.oisst.v2.highres/
      and integrated them down to monthly temperatures using CDO. I get a trend on 1998-2014 of 0.13 C/decade, which is about 3x Bob’s value and in good agreement with ERSSTv4. However I would like to redo the integration with my own code as a cross check.”

      I suggest you do that cross check, because my source for the Reynolds OI.v2 data is available online to anyone. My source is the KNMI Climate Explorer. And my earlier source was the NOAA NOMADS website. Maybe you also need to cross check the results of this post if you’re having trouble processing data.

      Cheers

      • I have cross-checked the CDO (Climate Data Operators) result against my own code and can confirm that the trend in the daily data is 0.13C/decade.

        I’ve also downloaded the OISST monthly data from here:
        ftp://ftp.emc.ncep.noaa.gov/cmb/sst/oimonth_v2/GRIB/
        For these data I get a trend of 0.054C/decade. That does not match the trend from KNMI. However the KNMI data do seem to be the older data, since they are on the monthly (1 degree) grid rather than the finer daily grid.

        So there’s still a mystery here.

      • Kevin: Further to your comment and my reply. You are correct that the monthly Reynolds OI.v2 data are not bias corrected for ship inlet and buoys. Reynolds et al. (2002)…
        ftp://ftp.emc.ncep.noaa.gov/cmb/sst/papers/oiv2pap/oiv2.pdf
        …state that:
        “Figure 4 suggests that present in situ bias error limits can exceed 0.05 deg C for globally averaged monthly SSTs. We have not corrected the OI.v2 in situ data by the factors in Table 2 because of the uncertainties of the biases in the table. However, any correction of satellite data is further complicated by in situ biases and their uncertainties.”

        Curiously, NOAA uses the monthly Reynolds OI.v2 SST data for quality control of the ERSST.v4 data. See Huang et al.:
        http://rda.ucar.edu/datasets/ds277.0/docs/ERSST.V4.P1.JCLI-D-14-00006.1.pdf

        How do we know it’s the monthly data and not the daily data? Because Huang et al do not reference Reynolds et al (2007), which presents the daily OI.v2 data:
        http://www1.ncdc.noaa.gov/pub/data/oisst/papers/daily-sst.pdf

        And yes you are correct that the daily OI.v2 have bias adjustments for ship inlets and buoys. But that still doesn’t overcome the uncertainties shown in Reynolds et al 2002, which was the reason they didn’t include them in the monthly data.

        Additionally, they also use different satellite observations in the daily data than the monthly data, so it’s a totally different dataset.

        Last, if NOAA was so confident in the bias adjustments in the daily version, why do they still produce a monthly version without the bias adjustments? Answer: they have two totally different purposes. The monthly Reynolds OI.v2 data are used for satellite-era sea surface temperature observations (still remain as an option for GISS LOTI data) and the daily OI.v2 data are to be used for reanalyses and weather models.

      • Regarding OISST:

        “At CPC there is a daily OISST dataset that is used in our models. However, this is a more recent product. Before the daily product existed, there was (and still is) a weekly OISST dataset. This weekly product is different entirely and is not simply the seven-day average of the daily OISST for a variety of reasons (see references below). Since both of these products are satellite-based — they do not measure sea surface temperatures directly — there are adjustments done before we get the final SSTs (these adjustments change from month to month). Different methods and statistical techniques in the daily and weekly product lead to different numbers in the two similarly named datasets. For instance, ship and buoy SSTs are measured differently and at slightly different water depths, thus there is a statistical difference of ship-buoy SSTs of about 0.12ᵒC. The weekly OISST was developed when ships were the main source of in situ observations, and there is no ship-buoy SST adjustment in it. On the other hand, the latter daily OISST was developed when more buoy data became available, and there is a ship-buoy SST adjustment in it, as in the latest version of ERSST.”

        https://www.climate.gov/news-features/blogs/enso/exactly-same-completely-different-why-we-have-so-many-different-ways

        So weekly (and presumably monthly) OISST would not have buoy corrections.

    • OK, I’ve finally sorted out the OISST mess.

      The older monthly files are available as netcdf, grib, or Fortran binary files:
      http://www.esrl.noaa.gov/psd/data/gridded/data.noaa.oisst.v2.html
      http://ftp.emc.ncep.noaa.gov/cmb/sst/oisst_v2/GRIB/
      ftp://ftp.emc.ncep.noaa.gov/cmb/sst/oimonth_v2/

      However these files are … counterintuitive. The SSTs are infilled over land, so integrating them produces results which are slightly off.

      So I downloaded the netcdf file from KNMI. Geert has very sensibly masked the files to remove the infilled land cells, so that interpolation in KNMI gives a sensible result. Unfortunately that means that in order to reproduce the KNMI results we need to reproduce the land mask exactly.

      Fortunately there’s a trick to do this: I used the KNMI file itself as a mask to mask my NOAA file, and redid the integration of the masked NOAA data. Once I do that, my results agree exactly with the KNMI results. (Unsurprising, since I believe KNMI makes a lot of use of CDO internally).

      I then went back to check the OISSTv2 daily data. These are not infilled, and the land cells are correctly set to missing, so the trend of 0.13C/decade stands.

      I’ll send Geert my scripts and see if he can include OISSTv2 daily on KNMI so more people can experiment with it. (But unfortunately sorting that out taken all the spare time I have today.)

  10. Hi Zeke and Kevin. First, you wrote: “Buoys which show a large daily temperature variation are rejected: in deep water the daily temperature range is only a few tenths of a degree, but in very shallow water it can be substantial which presents problems when some data are missing.”

    Those large variations exist in the record and are therefore part of it. By eliminating coastal waters you’re opening yourself to criticism. In other words, someone will claim you’re cherry picking grids to meet your objective. So, can you also provide trend comparisons including the highly volatile data you rejected in order to overcome that possible criticism? Will you provide maps to show which parts of the oceans are included in your evaluations, and to show if those regions remain constant over the term of your analyses?….in other words, there are different types of buoys—the NOAA TAO project and NOAA drifter program (which does not include ARGO) for example—which have different start dates and locations (TAO buoys are limited to the equatorial Pacific while the drifters are global). Another reason for the maps: do you include the polar oceans?…in other words, are you biasing your results with differences in how sea ice is addressed?

    Second, you wrote: “The buoy records are from purpose built instruments which are largely standardized, resulting in much more homogeneous records…”

    “…largely standardized…”? “…more homogeneous records…”? While your wiggle-worded claims sound logical, those are big assumptions. Please document, because your analyses depend on them.

    Third, why did you include trend comparisons starting in 1995 in the right-hand cell of your Figure 3?

    That start year has nothing to do with Karl et al. or the hiatus. Karl et al use 1950, 1998 and 2000 for the start years of the trend comparisons in their Figure 1, not 1995.

    So we can disregard the left-hand trend comparison in your Figure 3 because it’s misdirection. The right-hand comparison starting in 1998 shows the pause buster ERSST.v4 data have a noticeably higher trend than your buoy-only data. In other words, you’re showing that NOAA overcooked the ERSST.v4 data during the hiatus in those (not illustrated) portions of the global oceans to which you elected to limit your analyses.

    Fourth, you mention the UKMO HadSST2 data are used to bias adjust the ERSST.v4 data, but your analysis does not address why the HadNMAT2 data (which do not have ship-inlet and buoy biases) have a noticeably lower warming rate than the ERSST.v4 data during the slowdown period:

    HadNMAT2 data end in 2010, which is why that comparison ends there.

    Fifth, further to the comparison between ERSST.v4 and HadNMAT2 data, there is no difference in the warming rates between those two datasets from 1998 through 2007:

    That of course indicates that the difference between the ERSST.v4 and HadNMAT2 warming rates from 1998 to 2010 occurred due to adjustments after 2007. As far as I recall, there was a major uptick in the number of ICOADS SST observations in 2005.

    Sixth, as Judith mentions, you also do not address why HadSST3, which has also been bias-adjusted for the ship inlet and buoy differences, show a warming rate from 1998 to 2010 that’s comparable to the HadNMAT2 data and therefore have a much lower warming rate than the pause buster ERSST.v4 data:

    Seventh, are ARGO buoys included in your buoy analyses? They are not included in sea surface temperature datasets.

    Last, not to be nitpicky, but you list ERSST.v3 as the former NOAA sea surface temperature dataset. That’s incorrect. The former in-situ-only NOAA SST dataset is ERSST.v3b. If you’re not aware of it, ERSST.v3 was a satellite-enhanced dataset that was introduced in early 2008 but replaced with ERSST.v3b later in 2008 because the satellite-enhanced version didn’t show as much warming during the hiatus.

    That’s it for now.

    Cheers.

    • Bob: I’ve added coverage maps for every month in ascii form in this file;
      http://www.ysbl.york.ac.uk/~cowtan/ftp/coverage.log
      Top is the Arctic, left edge is the dateline. If you look at the end of the file you can see the world map clearly. We mask ERSST down to the same coverage as the buoy record, so it is a like-with-like comparison, although it only tells us about cells where the buoys have coverage.

      I’ve done some preliminary analysis on the homogeneity of the buoy records by comparing buoys within 25km. For most buoys the differences are 0.1-0.2K, which suggests that the buoys are both more precise and more homogeneous than say the land stations. But I have found a few cases with rather larger differences. The handful I have checked correspond to buoys which have large diurnal cycles, which we are eliminating. I am planning a more systematic study – there’s a little discussion in the README.

      To remove the diurnal range check, edit line 75 of icoads.py and remove the final condition (or all of them if you prefer). I’ve just done a rerun with all the buoys and the results are unchanged.

      For trends to 2014-12 from various start dates, see this figure (inspired by Deep Climate. The confidence intervals are 1sigma, ARMA(1,1)):

      As to the difference with HadSST3, I have no basis for choosing between them except during WW2, where I have unpublished results which reject the spike in ERSSTv4. For the remainder of the record I cannot reject either as a plausible record of global SST.

      • I appreciate the use of error ranges for your graph (if at 1 sigma) in this comment. But what was the confidence interval selected for statistical testing of your primary conclusions?

      • Top is the Arctic, left edge is the dateline. If you look at the end of the file you can see the world map clearly.

        sphaira bovis

      • Kevin Cowtan writes: “I’ve done some preliminary analysis on the homogeneity of the buoy records by comparing buoys within 25km. For most buoys the differences are 0.1-0.2K…”

        Isn’t that greater than the buoy bias adjustment?

      • Mr. Bob has got his eye on these youngsters.

      • davideisenstadt

        removing buoy data that exhibits characteristics you dont like after collection is ex post selection, no?

      • Bob should always have his eye on these “youngsters”. Cowten’s always had a good cherry harvest.

    • Bob,

      Regarding the trend periods shown, 1995-2014 is used because that is the full period of the dataset (where buoys have reasonable spatial coverage). I included it because not showing the trend of the full period graphed would be cherry-picking.

      We will be looking at HadSST3 in detail masked to the same areas we have buoy and ERSST data for to look at the differences in more detail. This initial analysis focused on the NOAA adjustments, so we didn’t examine Hadley’s SST series initially.

    • Dear Bob,
      You raise some interesting points, which I’d like to expand on a little. I’ve used your numbering.

      First, coastal “SST” from drifters can exhibit large variations because there can be large variations in coastal areas. Also, sometimes, buoys wash up on beaches and start measuring air temperature rather than SST. It’s also common to see drifting buoys reporting erratic measurements shortly before they go offline, wherever they happen to be. Occasionally, they get picked up by ships and, for a short period, record air temperatures on deck. This paper goes into some of the problems that ship and drifter data suffer from:

      http://onlinelibrary.wiley.com/doi/10.1002/jgrc.20257/full

      Second, drifter design was standardised in the early 1990s. Since then, the only major change I know of has been in the size of the buoys: modern mini drifters are smaller than their non-mini predecessors. Different manufacturers make buoys to the specifications laid down in the standard design. Metadata for buoys is not especially easy to get hold of (for ships there’s ICOADS and WMO publication 47), but work is ongoing to organise the metadata and to see if there are measurable differences between drifters from different manufacturers. Work has also been done to fit a small number of drifters with higher-quality thermometers alongside the standard thermistor. See e.g.

      http://journals.ametsoc.org/doi/abs/10.1175/2010JTECHO741.1

      The results suggest that individual buoys can exhibit a variety of problems. On average, though, they seem to be unbiased relative to the true SST. Individually, they are higher or lower, with calibrations that vary by a few tenths of a degree.

      There can occasionally be large calibration errors (of a degree or more). Nowadays, there is constant monitoring of the drifter network by a number of different centres. Large calibration errors are usually identified quickly. Sometimes these can be fixed remotely, sometimes they can’t and the buoy goes onto a list (see, for example, http://www.meteo.shom.fr/qctools/ ). Monitoring of the early data was less thorough.

      As a result of the above considerations, everyone who uses drifting buoy data applies some level of quality screening to it. What is generally accepted is that the average drifter makes a much better SST measurement than the average ship (though there are exceptions, of course, in both directions).

      Third, I’d note that drifter coverage is not so great prior to 1995 (I think Kevin said the same), so the relative effect of calibration errors would be more pronounced as well as the difficulty of making a solid comparison with fewer data points. I think, more generally, it’s useful to know how consistent the trends are across a variety of periods. As your graphs show, looking at a variety of periods can reveal different aspects of the data.

      Fourth, (I think you mistyped HadSST2 when you meant HadNMAT2, or did I misunderstand?). Question: are the coverages of HadNMAT2 and ERSSTv4 in your plot the same? Coverage of NMAT is confined to areas where ships go, and ship coverage has declined somewhat over this period, whereas ERSSTv4 is more or less global.

      The closeness with which NMAT and ERSSTv4 should track each other is something to consider also. The ERSST ship adjustment is smoothed so that variations of shorter than a few years (approximately) are not resolved. My understanding of this is that it’s necessary to reduce the effect of random measurement errors on the estimated bias. By smoothing over several years, the effect of random measurement errors average out, so what’s left is largely due to systematic errors (which is good because that’s what they are trying to assess). On the other hand, it means that the method can’t resolve changes in bias that happen faster than that.

      Fifth, the uptick in the number of ICOADS SST observations in 2005 coincides with a large increase in the number of drifting buoy data. Depending on the version of ICOADS used, there’s also often a change in the number and composition of observations at the switch from delayed mode to real time. I think for ICOADS 2.5, that’s the end of 2007.

      Sixth, don’t forget that there are 100 different estimates of HadSST3 – which together span estimated uncertainty in the bias adjustment – and additional measurement, sampling and coverage uncertainties which can also affect the trends over shorter periods such as the ones being discussed here. In brief, the trend over this period as estimated by HadSST3 is uncertain. The same goes for ERSSTv4: there is an uncertainty analysis (Liu et al. 2015 published at the same time as Huang et al. 2015). One should be wary about drawing conclusions from a comparison based only on the medians.

      http://journals.ametsoc.org/doi/abs/10.1175/JCLI-D-14-00007.1

      Best regards to one and all,

      John Kennedy

      • I would like to commend Liu, et al., for expressly stating their significance testing method at various critical points in their paper. For one example, when mapping different results for SST warming trends they helpfully state:

        In all plots, results are calculated from monthly data, shown in color, and only illustrated when they exceed a 95% significance based on a two-tailed Student’s t test.

        That is a standard of clarity we should all applaud. And require.

        I have raised the question of how (and when) the decision was made in Karl, et al., to apply a more lenient test of 90% but have never received a satisfactory answer beyond general agreement that post hoc decisions to loosen the test parameters can be problematic.

        Kent

      • Sorry for the delay. I was off visiting with family for most of the week.

        Thanks, Kevin Cowtan, Zeke Hausfather and John Kennedy, for your thoughtful replies.

        I’m looking forward to the similar comparison using the HADSST3 data.

        Cheers.

  11. As with atmospheric measurements, satellite provides so much better coverage and resolution. But so many corrections, it’s difficult to have confidence in any measurements.

    The small scale features in the OISST do inspire greater confidence of what’s going on spatially and of the extent of spatial variation.

    • If we weren’t talking about fractions of a degree over decades they would all be in basic agreement.

  12. David Springer

    What’s the trend from ARGO alone?

    • Here are the trends over the ARGO period (Jan 2005-December 2014) for each series:

      ARGO – 0.12 C per decade (per Willis’ figure below)
      ERSST v3 – 0.07 C per decade
      ERSST v4 – 0.12 C per decade
      Buoy-Only – 0.13 C per decade

      So it looks like ARGO agrees pretty well with other Buoys and the adjusted ERSST v4.

      • Who picks these colors?

      • It might be worth recalculating the ARGO trend. I’m getting +0.04C/dec. Not to say I’m right… it was a rather quick exercise on my part. But my graph looks almost exactly like the top frame posted by jim2 and my mean is 19.91 vs. 19.89.

      • When I calculated the trend on the monthly anomalies (instead of the monthly values), I get a trend of 0.114C/dec. Didn’t think the difference in the trends would be so high.

      • If I take the trend of the monthly means repeated 10 times I get -0.07C/dec. Looks like this resolves the difference. In this case, perhaps the trend should not be calculated on the monthly anomalies as this adds a spurious value. Maybe it would be ok to do so if the anomalies were calculated by subtracting a trendless sinusoid?

      • … and scratch that last bit… nonsense on my part… it’s removing a spurious negative trend.

      • @climateadj: Just curious… did anyone ever volunteer to port your “65-year HadCRUT4 with and without TSI” code to R?

        I’ve uploaded clim60vp2.R (mwgrant’s R implementation of the original program) to http://clim.stanford.edu > MATLAB > Clim60
        Let me know if you run into any problems.

    • David Springer

      ARGO confirming the ERSSTv.4 trend inspires confidence in the latter AFAIC.

      The global map of ARGO trends is far more interesting than the global average however. The almost complete lack of SST trend in the tropics reinforces what I’ve been saying for a decade: there are hemispheric heat engines running and they simply speed up in response to CO2 forcing moving excess energy from the tropics to the poles.

      There’s a maximum attainable SST of 30C due to strong convection becoming dominant at that point. This effectively transports energy (carried insensibly a.k.a. latent heat of vaporization) away from the surface like an express elevator to radiate away near the top of the troposphere where it become sensible again and there is little resistance left to upward infrared propagation to space and great resistance downward to the surface.

      The strong convection point puts a cap on how much of a top-of-atmosphere energy imbalance can be sustained and we mustn’t forget that most of the solar energy entering the system enters through the tropics due to angle of incidence.

      What happens instead of the tropics getting warmer is the tropical climate zone is widened extending farther north and south. The temperate zone also extends north and south while the polar zones shrink. Antarctica was once deciduous forest. The normal state of the earth in the long view is no polar ice caps.

      Land of course follows what the ocean temperature is doing for the most part due to “continentality” where the divergence increases as does distance from shore.

      So there are two competing concerns; one beneficial and one adverse. Warmer climates are more biologically productive and CO2 enrichment adds to that effect. So the planet gets greener as it warms. The adverse effect is sea level rise but that’s mostly a human foible. No other species builds permanent infrastructure at the land-sea interface in anticipation of it it never changing.

      At a rate of 0.3 centimeters per year rise in sea level due to thermal expansion and ice cap melting this doesn’t seem to be an immediate problem and due to limited amount of economically recoverable fossil fuels the anthropogenic CO2 forcing can’t continue long enough to become a huge concern.

      Thus we come to the root problem in my view: finite recoverable fossil fuels. We have no choice about entering a so-called low carbon economy – one way or another. Either technology comes to the rescue again or Erlichman’s population bomb finally detonates.

      I’m a big believer in humanity’s capability to bring about rapid technological advances provided it isn’t hamstrung by nattering nabobs of negativity like climate change fanatics. Abundant, inexpensive fossil fuel is the goose that lays the golden eggs as far as advances in technology. Those advances come from the industrialized world with vibrant economies flush with funding for long-term R&D into alternative energy sources which will be needed within a century or two at most. Squandering R&D bandwidth on half-measures to reduce CO2 emission are foolish in the short-term and disastrous in the long term.

      • The rate of evaporation starts going exponential at 295K. At 303K it is pretty hard to heat the ocean much, it just boils away.

  13. The satellite data series will never be much help in determining longer term climate trends in our lifetime and there are obvious issues with the earlier recorded series that continue to make life difficult for present day climate researchers, both in terms of measurement bias and the lack of resolution.

    The post is an excellent one thanks Judith and the authors are to be commended for their balanced analysis and careful conclusions. Too much research is currently being pushed through in a rush to have papers available for the Paris Conference, but this paper is an exception.

    • One solution would be a one year hiatus on new papers by the IPCC before a conference.

      It is going to be 10-20 years (at least) before the IPCC gets this right. The IPCC has been working on it for almost 30 years and has done little or nothing to tighten the bounds on climate sensitivity.

      The only reason to “rush” papers is to sneak them in without a lot of critical analysis. There is no rush. The IPCC should take its time and do it right.

      The US congress should ban US action on climate until the sensitivity is bounded to +/- 10%. That would put the emphasis on science (bounding sensitivity), instead of propaganda.

      • What are you talking about? The IPCC only reviews the existing scientific literature. They don’t do any original research or “rush” anything..

      • They don’t do any original research or “rush” anything.

        Some times they rush to accept things like this:

        and then embarrassingly never mention again.

      • Joseph,

        I don’t know about TE, but to conflate human emissions of CO2 in such a way as to imply that climate change (a perfectly natural phenomenon, being simply the average of ever changing weather) is somehow preventable, or even evil in some way, shows you are delusional.

        You appear to be in denial of the simple facts that the Earth has cooled since its creation, CO2 is necessary for plant life, and more is better (to a point, of course), and nobody has ever demonstrated the alleged heating effect of CO2.

        There is no greenhouse effect. Why would anybody deny it doesn’t exit, apart from deluded Warmists?

        Cheers.

      • This comment is rather confused.

        1) The IPCC doesn’t publish papers. They publish 1 report every 6 years or so, and occasional special reports on specific topics. As far as I know its been years since they published anything.

        2) The IPCC actually has a cutoff for including papers in the reports. Its ~6 months before they are finalized or so, IIRC.

        3) Peer-reviewed papers are rarely rushed. They generally have a period of at least 6 months between submission and publication. This was true for the Karl et al paper as well (that all this fuss was about).

      • Mr. Hausfather:

        I would take these criticisms seriously if not for two things (the short list):
        1. the rush to papers before each meeting. They need a longer cooling off period

        2. Further: there is World Wrestling Federation literature used by the IPCC. While it is noted in the document it doesn’t appear in the free standing citation list, The “scientist”s names don’t even show in the free standing citation list, one example is:

        Christie, P. and M. Sommerkorn, 2012: RACER: Rapid Assessment of Circum-Arctic
        Ecosystem Resilience, 2nd edn., WWF Global Arctic Programme, Ottawa, Canada,
        72 pp.

        The World Wrestling Federation (or any other WWF acronym) is not known for objectivity or science.

      • @PA: The World Wrestling Federation (or any other WWF acronym) is not known for objectivity or science.

        Unlike PA.

      • 3) Peer-reviewed papers are rarely rushed.

        Yeah, Mike’s Nature Trick was so sound, and such good objective science that the IPCC didn’t even need to mention Hockey Stick ever again.

        And when AR4 declared A temperature rise of about 0.2 °C per decade is projected for the next two decades for all SRES scenarios., they knew there wouldn’t be any problem if in AR5, they ran away from any rate of warming discussion at all and returned to unfalsifiable and nebulous ‘sensitivity’ arguments instead.

        IPCC is, as indicated, summary science, but there is garbage in garbage out. And since people expect something new from IPCC reports, the IPCC is obliged to include the latest garbage.

      • TE you understate the problem.

        We have the climategate mails so we know exactly what they did do.

        1. The IPCC editors edited the peer review so it was pal review.
        2. The IPCC editors edited the journal editors so they had favorable editors.
        3. The IPCC editors edited the science. They decided what science they wanted, made it up, and rushed it through the edited journals.

        The statement that the IPCC doesn’t publish papers is on its face true. But the editors and their minions do publish papers, and they do sneak them in under the wire, they even sneak in grey literature and pal reviewed (see above) literature.

        The climategate emails indicate that ZH’s points are incorrect. The IPCC rules were discussed in Pirates of the Caribbean, “It’s not really a rule … its really more of a guideline”.

      • No confusion here. True the IPCC does not rush anything, however the IPCC sets a deadline and people rush to get their papers published or in print or in review before the IPCC deadline.

        It has been documented that sometimes the rushing of these papers is facilitate by reviewers, publishers and even the IPCC if the conclusion of the paper has an agreeable flavour.

        To pretend this is not a real issue is disingenuous.

      • “Unlike PA”

        But PA doesn’t pretend to be objective and full of the sanctity of peer reviewed science like the IPCC pretend their assessment reports are.

      • And oh, yeah, most people think that climate change is a matter of Radiative Forcing(RF). But the AR5 spent a lot of pages entertaining speculation that RF is out and what matters is Effective Radiative Forcing(ERF). We’ll see if that persists, or if it’s yet another artifact that is quietly omitted going forward because it was prematurely included.

        It’s a reminder that the IPCC is inherently political, and not just because it comes from a political body ( the UN ) or from the efforts of an activated, self described socialist who fled ( without much shame ) to China after defrauding the UN ( Maurice Strong ). But the IPCC is political also because internal contributors want their own research included and will jockey for power to do so.

      • “But the editors and their minions do publish papers, ”

        Quite an elaborate conspiracy theory you have there, PA.

      • PA it was good to see you making this point on the recent Lomborg post. Not!

      • human1ty1st | November 23, 2015 at 1:26 pm |
        PA it was good to see you making this point on the recent Lomborg post. Not!

        I don’t pay much attention to the cost/impact discussions because it is a sick joke.

        If you plot the rate of increase in PPM/Y against cumulative emissions you get an odd looking graph.

        There ism’t much hope for pushing the rate of increase above 2.5 PPM/Y.

        The forcing in the surface layer (the bottom 480 meters) is 0.2 W = 22 PPM, from only actual study (Feb, 2015) of downwelling IR. The surface averages 288K and a 1K increase takes 5.5 W/m2.
        If you do the 0.2 W/m2 * ln(2) /ln (392/370) magic you find the TCR forcing is 2.4 W/m2, or 0.44 K.

        Less than 500 PPM and 0.44 K isn’t going to be problematic even if you project a 0.66K ECS. Since the real world (UAH/RSS/Radiosondes) don’t show evidence of much effect above the surface layer that is pretty much it.

        Someone smarter than I with access to the models could provide the TCR = 2 (the IPCC mean) GCM modeled surface forcing for the 370 PPM to 392 PPM transition for comparison.

      • Jeff Norman: But PA doesn’t pretend to be objective

        You on the other pretend to speak for him.

        I’d find that more believable coming from PA himself.

    • TE, you doubt that humans through CO2 emission have no effect on climate? Are you a greenhouse effect denier?

  14. Am I understanding this correctly? Hausfather and Cowtan say there is a rather straightforward way to test if Karl et al adjustments are proper or not – compare Karl et al adjusted record to a record made only from buoys.

    Ok, that makes sense. So, how to make a record made only from bouys:

    1. Buoys which show a large daily temperature variation are rejected:
    2. daily data are placed into 550 km equal area grid cells based on the location of the buoy for that day
    3. monthly averages are determined for each cell.

    Whoa nellie. Step 1 is to reject bouy data? Charming.

    So, now they’ve got this incomplete monthly averaged grid cell guff – what’s the upshot?

    “The resulting coverage is still limited and so produces a biased estimate of global sea surface temperature.”

    Whoa, hang on. “Biased”? Biased how? They don’t say, but what they mean is “biased in that it does not yet align to Karl et al which is our aim”.

    Typical pre-determined outcome. So, how to get there?

    4. reduce the coverage of the ERSST datasets to match the buoy dataset (now using a fine 1 degree grid for all the data)
    5. calculate anomalies for all the datasets using a 2001-2010 baseline.
    6. area weighted mean temperature is then calculated for each record.

    Whoa, more data rejection? A 9 year baseline for calculating ‘anomalies’? Area weighting? Hey presto, aligned to Karl et al.

    Those guys must be exhausted.

    • It seems to me they may have simply created a low pass filter that picks up only the long term warming which is largely agreed upon.

      • The long term (warming) trend is what is of interest is it not? The removal of outliers for that purpose is fine, but if one is looking at the extent of natural (and unnatural) variability then the outliers must be included in the analyses.

  15. Buoys provide many types of “sea surface temperatures”. This from people who apparently know a thing or two –

    “Understanding Sea Surface Temperature

    SST is a challenging parameter to define precisely as the upper ocean (~10 m) has a complex and variable vertical temperature structure that is related to ocean turbulence and air-sea fluxes of heat, moisture and momentum. A theoretical framework is therefore required to understand the information content and relationships between measurements of SST made by different satellite and in situ instruments, especially if these are to be merged together. The definitions of SST developed by the GHRSST SST Science Team (agreed at the 2nd and 3rd workshops) achieve the closest possible coincidence between what is defined and what can be measured, taking into account current scientific knowledge and understanding of the near surface thermal structure of the ocean.
    – See more at: https://www.ghrsst.org/science-and-applications/sst-definitions/#sthash.p4CXF4bH.dpuf

    Now which buoy measurements did they use? Who knows?
    Maybe, with any luck, they used –

    “The foundation temperature (SSTfnd)

    The foundation SST, SSTfnd, is the temperature free of diurnal temperature variability, i.e., SSTfnd is defined as the temperature at the first time of the day when the heat gain from the solar radiation absorption exceeds the heat loss at the sea surface. For conditions, when the SST increases or decreases monotonically over several days, the Tfnd occurs on a given day when the time rate of change of temperature is at a minimum (increasing SST), or a maximum (decreasing SST). If such a point in the daily time series cannot be identified, the SSTfnd should be set to a clearly stated time. SSTfnd is named to indicate that it is the foundation temperature upon which the growth and decay of the diurnal heating develops each day. Only in situ contact thermometry is able to measure SSTfnd and analysis procedures must be used to estimate the SSTfnd from radiometric retrievals of SSTskin and SSTsubskin taken at other times of the day.”

    Or maybe they used one or more of the various other temperatures, all of which apparently need quite a lot of massaging to make any sense.

    Hausfather and Cowtan state –

    “We find that a record using only buoys (and requiring no adjustments) is effectively identical in trend to the new NOAA record and significantly higher than the old one.”, which all sounds very sciencey, as long as we are prepared to believe that exactly the same type of data has been used for the “new” and the “old”, and that one can be demonstrated to be inferior to the other. We also have to believe that “effectively identical” is good enough.

    The temperature taken at 20 to 30 cm below the surface is inconstant, depending on surface winds, buoy movement and various other factors.

    Of course, after adjustments by all involved, as has been noted by NOAA –

    “The newest version of ERSST, version 3b, is optimally tuned to exclude under-sampled regions for global averages. In contrast to version 3, ERSST v3b does not include satellite data, which were found to cause a cold bias significant enough to change the rankings of months.”

    Optimally tuned, and ignoring inconvenient satellite data,

    So, a buoy only SST record is completely meaningless, without much further explanation of terms and methodology. Either that, or accept that the authors are grasping at waterlogged straws to avoid drowning in an undertow of their own devising!

    Another complete waste of time and effort. At least it keeps them away from interfering with real scientists. A small price to pay, I guess.

    Cheers.

    • Sorry, the ‘significantly higher’ line slipped by me in proofreading. I have not done any significance tests. I would rather invert the statement and say that the buoy trends give us no reason to doubt the ERSSTv4 trends.

      Doing a significance test is not as simple as looking at the error bars on the trends (whether OLS or ARMA). To illustrate: while the difference in the trends between buoys and v3 (using ARMA) is much less than the uncertainty in the trends, the trend in the difference series is much greater than the uncertainty in the trend in the difference series. But I’d need to think about the sources of correlation more carefully before drawing a conclusion. If you want to take a go, the data are in the file ‘all.temp’ in the zip file (data, v3masked, v4masked, buoy).

      Sorry, got to go to bed now.

    • Kevin Cowtan,

      Just as a matter of interest, you wrote –

      “Sorry, the ‘significantly higher’ line slipped by me in proofreading.” This doesn’t inspire confidence, that you apparently proof read only once. Did the error slip by your co-author as well?
      Yes, I’m picky, it’s supposed to be science.

      But thanks for the response. What sort of sea surface temperature are you talking about? What sort of sensors, and where were they placed?

      A few things seem to slip by NOAA from time to time, so I’d be less than inclined to take everything NOAA says as gospel. As you point out, their data series purporting to represent SST depend on your definition of SST. If they choose to define an apple as an orange coloured citrus fruit, you might leap to an incorrect conclusion.

      What is your work supposed to achieve? To show that the presence of CO2 in the atmosphere created warming? This would seem to contradict the obvious fact that the Earth has cooled since its creation, regardless of atmospheric composition. As to thermometers showing increased temperatures in various places, maybe an examination of the sources of heat that cause this might be worthwhile. Or maybe not.

      So far, a waste of time and effort, without a bit more effort into demonstrating you have examined the assumptions, premises, methodology and the equipment which were involved in the numbers you so blithely accept as meaningful.

      That’s all part of the scientific method.

      Cheers.

      • Mike,

        The initial post (in Figure 3) shows uncertainties in trends using a simple OLS approach. For this the trend differences are significant. For more complicated autocorrelation-correcting approaches like AR(1) or other ARMA models it likely depends on the model chosen and the time period, and as Kevin mentions we haven’t done any explicit testing.

        Regarding your discussion of CO2, it’s not really germane to this discussion. We are talking about evaluating the effectiveness of adjustments done to ocean temperature records.

      • For this the trend differences are significant.

        ? Clarify statement with respect to adjective ‘significant’. Sigh.

  16. 1910 to 1940. Quite a steeple. Now, if we anthros are responsible even for a modest 25% (it might be 75%, but even the French tax man can’t easily score that percentage)…I repeat: What were we up to around 1910? What had we stopped doing around 1880 and around 1940?

    I know, I know. Don’t mention.

    • What had we stopped doing around 1880 and around 1940?

      We stopped spinning the Earth so fast. Spinning the Earth fast throws up 1000 °C magma through the mid-ocean ridges which heats the oceanic mixed layer (OML).

      • There was a dance called The Stomp, popular only in Australia for a couple of years in the 1960s. It threatened the flooring of surf clubs and had to be banned from some premises. Could that have slowed the spinning of the globe and brought on the Global Cooling scare?

        Trying to look beyond GHGs and aerosols here, Vaughan, for that anthro footprint (or stomp-print).

      • Vaughan Pratt,

        Apart from the silliness of us slowing the Earth to any degree, you’re probably headed in the right direction.

        According to NOAA –

        “Submarine Volcanism
        The EOI Geology Program aims to understand how submarine volcanic activity affects hydrothermal venting and life in the oceans. The ultimate heat source for all hydrothermal venting is molten rock within the Earth. Individual volcanic events bring this heat close to the surface where it can be released into the ocean, along with chemicals that sustain unique biological environments, both on and below the seafloor. Therefore, we are especially interested in studying the sites of recent submarine eruptions.”

        There are unknown numbers of thermal vents under the oceans, and also unknown numbers of active and inactive volcanoes there. I am aware that you think the interior of the Earth is “rock solid”. Maybe you should let the people at NOAA just how wrong they are, and straighten them out. Let me know how you get on.

        Cheers.

      • VP:

        Spinning the Earth fast throws up 1000 °C magma through the mid-ocean ridges which heats the oceanic mixed layer (OML).

        Centripetal Volcanism is a fascinating concept worthy of its own CE post.

        I look forward to it, particularly any correlation of CV to glacial periods examining the global redistribution of forces due to the ebb/flow of ice masses and the corresponding increases/decreases in mid-ocean ridge heat output.

      • opluso: Centripetal Volcanism is a fascinating concept

        Thank you. (But it’s centrifugal. Centripetal would be the force that prevents the magma escaping. Opposite, but only equal in equilibrium. Luckily your acronym still works.)

        worthy of its own CE post.

        Or even an AGU poster. San Francisco, December 17, morning (but the poster will remain up until 6 pm).

        particularly any correlation of CV to glacial periods

        That’s like comparing UHF TV to AM radio. CV effects are on a fast half-century period, glacial periods are more than three orders of magnitude slower at around 100,000 years. Any such correlation would be like picking up KQEH Channel 54 on your AM receiver. There is no way volcanism of any origin could possibly usurp Milankovitch theory.

        Besides which no one was around prior to the 17th century to measure LOD to within even a second. Moreover no one’s found a way to extract LOD from the geologic record to an accuracy of better than a few minutes (thus far done only via Fourier analysis of tidalites).

        A far better correlation is between LOD and AMO. In this theory (and it’s only a theory) the AMO is the result of CV, which it fits extraordinarily well. In the Wyatt-Curry stadium wave theory the AMO acts as a clock but there is nothing remotely like precisely measured LOD that explains the AMO itself, either its frequency, phase, or amplitude.

      • VP:

        I assume that shifting glacial masses would impact length of day, though perhaps not rapidly enough for what you are talking about. And while Milankovitch cycles would control (particularly landlocked ice) I was assuming your hypothesis included warming the oceans enough to raise SST and shift currents sufficient to increase melting rates for coastal ice.

        Yet even if it isn’t enough to melt coastal glaciers, you could still get this published in a top tier journal if you estimate the warming in terms of hiroshima bomb equivalents.

      • Vaughan Pratt,

        It’s a good thing that gravity stops the Earth’s interior from going anywhere. Where pressure of unknown origin forces molten rock through the crust, gravity makes sure it doesn’t fly off into space.

        I’m not sure about centrifugal volcanism. I assume you are joking, but If not you might explain how centrifugal force causes volcanoes in Antarctica. The oblate parts of the spheroid are at the poles, I believe, where centrifugal force is presumably least.

        I’m always willing to learn. When and where are you publishing this novel theory?

        Cheers.

      • I found the video of Prof. Pratt explaining his OML heating hypothesis:

      • Apparently I found it twice….

      • I cheerfully admit it’s a crazy theory, opluso. My worry is that the truth may turn out to be even crazier.

      • @opluso: I assume that shifting glacial masses would impact length of day, though perhaps not rapidly enough for what you are talking about.

        Yes and yes. A while back (far enough that Pekka Pirilla was participating in the discussion) I calculated that even moving enough land ice from the polar regions to create a belt of water 100 m high and 1000 km wide around the tropics would be insufficient to account for any observed LOD fluctuations in recorded history.

        And while Milankovitch cycles would control (particularly landlocked ice) I was assuming your hypothesis included warming the oceans enough to raise SST and shift currents sufficient to increase melting rates for coastal ice.

        Neither Milankovitch nor melting enters into any part in my CV (centrifugal volcanism) hypothesis. Furthermore the SST is cooled by more-than-average slowing such as experienced between 1879 and 1911. The hypothesis has no implications whatsoever for a bias towards warming. (Average slowing means the day lengthening by 1.6 ms per century, which is the estimated average over the past several millennia, and its implications for any long-term cooling trend is miniscule to nonexistent relative to other natural climate fluctuations.)

      • @MF: I am aware that you think the interior of the Earth is “rock solid”. Maybe you should let the people at NOAA just how wrong they are, and straighten them out.

        Look, I just explained to you a few days ago that magma and mantle are not the same thing, even quoting from the Wikipedia article on magma to that exact effect. Magma, the molten rock that is called lava when erupted, resides in the crust, not in the mantle, which is the solid rock that makes up the bulk of the Earth.

        Either you have a serious attention deficit disorder, or geophysics is not your strong suit, or you are just trying to be difficult. I’ll let others decide which. I’m guessing all three.

      • Just curious… did anyone ever volunteer to port your “65-year HadCRUT4 with and without TSI” code to R? It might be more interesting to see if one (I that is) could replicate your results using your description of the process. Might make for a fun project on a snowy January day.

      • Yes, mwgrant ported the original version to R. I tested it and it works fine. I’ll see if he’s ok with my putting it on http://clim.stanford.edu/ which I’ve recently organized slightly better.

        I’ve been taking care to make my MATLAB scripts also run under octave, which unlike MATLAB is free and no porting is needed. I’d do an Excel version but that will now have to wait until after AGU (i.e. after Dec. 18).

      • Vaughan – when I looked at Octave, it appeared to be a mish mash of disparate pieces. I run Linux. Did you have to install a bunch of pieces or just one installation?

      • Vaughan Pratt,

        When you explain that solid rock ascends through other solid rock so that adiabatic decompression enables it to melt, it would appear that you are using a different definition of solid to the one I use.

        Let us agree that when solid rock ascends through other less dense solid rock, this could, in fact, be described as convection. The direction of movement is away from the source of heat, and towards the colder surface, which you might not consider as solid as the mantle, but it seems to exhibit less decrompression melting, as a rule.

        Sorry. Still no global warming. Still no centrifugal volcanism.

        Unfortunately for your hypothesis, gravitational force stops the interior of the Earth being ejected through the surface, centrifugal force or no.You might have noticed that dirt on the surface at the Equator does not fly into space. Neither does the rock beneath, all the way to the core. This has been a fact few a few billion years.

        Maybe you could stick to Warmist denialism, or playing with trivial computer models producing pointless analyses of the past.

        By the way, when your presumably molten magma gets flung up through the crust (which through the miracle of CO2 stubbornly refuses to be flung), what fills in the space left by the now vanished magma? More magma, perhaps? Solid rock? From where?

        If you can’t answer these simple questions, some might think you don’t know what you’re talking about.

        Cheers.

      • @jim2: Vaughan – when I looked at Octave, it appeared to be a mish mash of disparate pieces. I run Linux. Did you have to install a bunch of pieces or just one installation?

        Not much. I installed it under Cygwin, with the following in /usr/bin:

        -rwx—r-x 1 pratt pratt 15360 Jun 1 02:07 octave-4.0.0.exe
        -rwx—r-x 1 pratt pratt 28160 Jun 1 02:07 octave-config-4.0.0.exe
        -rwx—r-x 1 pratt pratt 9216 Jun 1 02:07 octave-cli-4.0.0.exe
        lrwxrwxrwx 1 pratt pratt 16 Jul 8 21:22 octave -> octave-4.0.0.exe
        lrwxrwxrwx 1 pratt pratt 20 Jul 8 21:22 octave-cli -> octave-cli-4.0.0.exe
        lrwxrwxrwx 1 pratt pratt 23 Jul 8 21:22 octave-config -> octave-config-4.0.0.exe

        I would imagine the same pieces would work under Linux. Let me know either way how it works out.

      • @MF: Unfortunately for your hypothesis, gravitational force stops the interior of the Earth being ejected through the surface

        Volcano deniers. Now I’ve heard everything. :)

  17. ERSST firsst add 0.12 C degrees ter readings
    collected by buoys ter make ’em compatible ter
    ship bucket readings. ERSST next give buoy data
    extra weight in computations, then adjust post
    1941 ship bucket data, including a large cooling
    adjustment applied ter readings over 1998-2000.

    http://www.rossmckitrick.com/uploads/4/8/0/8/4808045/mckitrick_comments_on_karl2015.pdf

    Serfs think ‘ ERSST is WERRST.’

    • Toffs know that one can’t just show up at a Paris reception empty-handed. At least guests have been willing to throw together something for the event. It’s the thought that counts.

  18. This discussion is completely overblown. The adjustment amounts to about a twentieth of a degree C over 15 years, which killed the pause nevertheless showing how weak it was in the first place. Fifteen-year trends are very noisy over time and vary by almost 100% of their mean. Just the 15 years ending in 2007 had nearly 0.3 C per decade as a trend which was twice the long-term mean, and nobody talks about that. In contrast, thirty-year trends have held very steadily within 10% of their mean for decades, and that is a much better climate metric than the 15-year roller coaster.

    • ‘Must give us pause. There’s the respect
      That makes calamity of so long warming.’

      H/t H.

    • Whether or not there’s a pause is not material.

      CO2 should tend toward warming.

      But, that facts that:
      the longer term warming is less than low end projections,
      rates of radiaitve forcing have declined since peaking in 1979,
      emissions were flat in 2014,
      rising temperatures should lead to less intense storminess,
      etc. etc. are reasons to reject the exaggerations and ulterior motives of ‘climate change’.

    • The pause that is killing the cause can’t be buried with some vague little adjustments, yimmy. It’s part of history. We all lived through it and the desperate attempts by the red-faced settled science crowd to explain it away. Anyway, it has already done it’s thang. No binding agreement from the Paree party boys and girls, other. No bags of dollars for the third world despots. And The Donald rises, inexorably. Life is good! Sorry, yimmy.

    • This is a good point. Short-term trends are quite noisy, and the current El Nino will quickly make the “hiatus” or “pause” discussion rather moot. Looking at longer-term trends is more meaningful and less subject to sub-decadal influences of solar, volcanoes, or ENSO.

      • Subsequent to the El Nino hot year is an 18-month of big cold, showing that the El Nino discharges more than the current year’s energy. So the skeptics will cheer when the bottom drops out, the warmists will say it is just a temporary post-El Nino adjustment, and we’ll have to wait another 18 months to see which way the trend is going.

        We’re at the point where the “coolers”, of which I am one, are being tested. If the 2018 temps aren’t significantly down from 2015, then the imminent cooling, pre-Dalton isn’t likely to happen. If they aren’t significantly above the 2015 temps, then the IPCC rates are REALLY too high. And it will matter, because any agreement done iin Paris will take until 2018 to get any traction.

        Here’s something weird to consider. The committemtn for100 billion/year, with the US at an initial 3B. Sounds like a lot, right? Alberta, Canada, just passed a carbon tax that is said to be raising 2Billlion/year. We are minnows in tiny fishbowl, and are able to come up with 2B with a snap of the fingers. The 3B is nothing. The 100billion for all the developed world is near nothing. The US is/was printing 85B/MONTH.

        The pushback on the money is made to sound like the money is a big deal. It isn’t. Nobody wants to do it. Period. Give it to the developing world to be squandered on something that doesn’t matter? Don’t think so.

        The shouting is covering something else. It is very useful politically and makes some ideological and strategic sense (replacing fossil fuels for some energy source of greater density, like nuclear). But the INDCs don’t do the trick, the “threatened” portions of the world are not alarmed enough to do anything by themselves, and the developed world is not concerned enough to print more paper to cover their alleged costs (the US Fed can’t print 88B one month instead of 85, or redirect 3 of the 85 new paper for one month?).

        Either we are in a another time of the “madness of crowds”, or something else is going on under the smokescreen.

      • the current El Nino will quickly make the “hiatus” or “pause” discussion rather moot.

        Huh???

        Fine, when La Nina comes through the “non-warmers” will go on and on about how warmers lied through their teeth about the pause being over.

        Which they did.

        I’m more interested in the 2020 rate of CO2 increase and the 2020 temperature and will let others amuse themselves making claims and counter claims about short term temperature changes.

      • The current trend suggests that global warming is very unlikely to be anything other than beneficial, how is a 6 month blip going to change that? What happens if atmospheric temps dip next fall and plataeu below the past decadal average?

        If anything, this highlights the need for this el nino to end in another step change for GHG emission not to be essentially a moot point for climate climate change.

    • Any line with zero slope can easily obtain a positive or negative slope if you change the data at each end in opposite directions. I agree that it is better to look at longer time periods but you don’t want to fit a straight line to data unless it really is “straight”.

      • you don’t want to fit a straight line to data unless it really is “straight”.

        Agreed. The AMO and TSI have done a great job bending HadCRUT4 every which way, making it impossible until now to see a straight line.

        But if you remove the 65-year-period AMO with a 65-year filter (gee, why didn’t anyone think of that before?) to give the blue curve in this graph, and before doing that subtract the expected solar forcing of TSI from HadCRUT4 to give the red graph, you actually do get an astonishingly straight dependence of global mean surface temperature on CO2 forcing, aka log(CO2).

        So impossibly straight in fact that there can be no other explanation than that the folks at the Met Office Hadley and at East Anglia Climatic Research Unit have conspired over the past few decades to adjust that data so that it would turn out straight after removing all the shorter-term natural fluctuations. ;)

        Note that this includes all feedbacks and impacts of other GHGs besides CO2, because they’re in linear proportion to CO2 and therefore do not distort this linear dependence. CO2 itself is merely the 90 lb weakling applying judo to the big guy, water vapor, with 20 lb methane etc. dutifully trotting along behind CO2 to shout encouragement.

      • Vaughan Pratt,

        You wrote –

        “So impossibly straight in fact that there can be no other explanation than that the folks at the Met Office Hadley and at East Anglia Climatic Research Unit have conspired over the past few decades to adjust that data so that it would turn out straight after removing all the shorter-term natural fluctuations.”

        You appear to be accusing them of fraud, or conspiracy or something. There are many other explanations – excessive enthusiasm, gullibility, stupidity, collective delusional behaviour, ignorance . . . the list goes on.

        I believe the Met Office has lost at least one contract to provide the BBC with weather forecasts, in spite of their supposed “expertise”. You may be aware of “climategate”, which showed possible attempts to subvert peer review, data manipulation tricks and a few other less than desirable practices. Senior researchers lost data, were unable to understand simple spreadsheets, and complained that if they were forced to release publicly funded data to the public, people might criticise the researchers.

        Gee. Criticism of climatologists? Is it even possible?

        Cheers.

      • Nice Vaughan.

        We took out co2 and the residual looked like amo

      • VP:

        Note that this includes all feedbacks and impacts of other GHGs besides CO2, because they’re in linear proportion to CO2 and therefore do not distort this linear dependence.

        This seems to address the question I had for you on a different thread regarding captions and explanations:
        https://judithcurry.com/2015/11/16/400-years-of-warming/#comment-745215

        Nevertheless, your thoughts on (all forcings) + (all feedbacks) = (observed temperature for 100+ years) and TCR ≈ ECS would be appreciated.

        Kent

      • VP & SM – interesting, independent complementary calcs, very nice. Gets my attention.

      • VP does this analysis go beyond 1990? What’s the source of the graph? I’d like to read more.

      • richardswarthout

        VP

        Your graph is interesting. I do not think, as Mike Flynn suggests, that you accused the met office of malfeasance. My take was that it was sarcasm; a jab at those who criticize the met office.

        Also, Moshers comment infers that you found nothing significant. My take is quite different. You calculated a 65 year rolling average, which should be significant in the field of climate science, a field with professed interest in the long term.

        Richard

      • Vaughan, a climate sensitivity of 1.73C per doubling of CO2 looks believable, about 50% more than what would be expected with CO2 without any feedback. The UAH/RSS satellite temperature measurements don’t show the hot spot in the tropical upper troposphere is not appearing as predicted by the Hansen, et al model with large positive feedback due to water vapor.

      • My take was that it was sarcasm;

        Mike Flynn quoted me out of context. He omitted the two characters that followed, namely “;)” which fully justified your take, as I’d intended.

        No one does “troll” better than Mike Flynn. He’s a specialist, just as the Mafia specializes in extortion, though I can’t imagine trolling pays as well. :)

        You calculated a 65 year rolling average, which should be significant in the field of climate science, a field with professed interest in the long term.

        Indeed. I believe this is a chaotically excited natural resonance governed by nothing more than two properties of the mantle: its moment of inertia and its Young’s modulus of elasticity.

      • vp, “Indeed. I believe this is a chaotically excited natural resonance governed by nothing more than two properties of the mantle: its moment of inertia and its Young’s modulus of elasticity.”

        Nice, but I still like my three basins controlling the resonance. The changes in the panama flow and Antarctic circumpolar currents would vary the transfer and the frequencies. I believe the VC variables are a bit more fixed in that case. The three basins are north atlantic, north pacific and southern hemisphere oceans with volume ratios of 1:2:4 roughly.

        Depending on atmospheric circulation and the rate that the basins can equalization you could have the north pacific sea level a foot higher than the north atlantic which could deform the surface a touch stimulating more NH volcanic activity, Then there would be variations in snow fields and glacial extent with an N-S variation not dependent on average temperature but temperature distribution..

        Hey, a few hundred years from now we can find out right? :)

      • @eem: Vaughan, a climate sensitivity of 1.73C per doubling of CO2 looks believable, about 50% more than what would be expected with CO2 without any feedback.

        Quite right, Erik. Bear in mind however that what we’re currently experiencing is neither equilibrium climate sensitivity (ECS) nor transient climate response (TCR) but the empirical or observed climate sensitivity (OCS, I suppose).

        ECS is the eventual temperature rise following a doubling of CO2, which may take centuries to reach that temperature, necessarily long after the doubled CO2 has stopped changing. As the temperature approaches its equilibrium value its slope decreases until in the limit it’s flat, the equilibrium condition.

        TCR is the rise in 20-year climate during 70 years of a CO2 CAGR of 1%. It is measured at the start and end of the 70 years, and the rise is the difference. A CAGR of 1% over 70 years corresponds to a doubling in CO2 over that period. Unlike ECS, there is no requirement that either CO2 or temperature cease increasing at any point. Currently TCR is only meaningful for models, but RCP8.5 (“business as usual’) CO2 for 2010-2080 should be quite similar on average since it will hit the 1% CAGR mark shortly before mid-century and then rise above it.

        Empirical climate sensitivity is like TCR but with CO2 doing its own thing instead of obeying some rule like a fixed CAGR. In the 1960s CO2 CAGR was around a quarter of a percent, by 1980 it had climbed to 0.36%, by 2000 0.5%, and today it’s around 0.65%.

        The advantage of plotting climate against forcing instead of time is that it tends to factor out any changes in CAGR because CAGR is defined in terms of time while forcing is not. This is what makes it possible to observe a straight line after removing the known oscillations and TSI. The longer the observation, the more confident one can be about whether the trend line has any real meaning.

        I say “tends to” because in practice if forcing were to slow down it would give the ocean more time to absorb heat and rise in temperature, making the situation slightly more like the equilibrium case. While this might sound hard to model, it turns out to be much simpler than you’d expect if you judged it by the enormous complexity of global ocean-land circulation models (GCMs) like the CMIP5 suite.

        So far forcing has not slowed down, and so we have had no opportunity to date to observe what would happen if it did. We can only theorize, but that’s not necessarily a tall order, the complexity of the CMIP5 models notwithstanding.

        An empirical climate sensitivity of 1.7 or 1.8 °C is entirely consistent with an equilibrium climate sensitivity of 3 °C. For the purposes of 21st century forecasts, ECS is entirely irrelevant. Moreover if CO2 stops following RCP8.5 then even observed climate sensitivity will become less relevant. Theory (good theory anyway) then becomes the more reliable basis for projection.

        If CO2 continues to follow RCP8.5 it will reach 936 ppm in 2100. If in doing so climate continues along that straight line with slope 1.73 °C per doubling of CO2, climate in 2100 will be 1.73*log2(936/400) = 2 degrees hotter than today (so 3 degrees hotter than in 1910).

        Naively using the 3 °C/doubling figure for equilibrium climate sensitivity might make you think that the rise would be 3*log2(936/400 = 3.7 °C hotter than today in 2100. However that’s not what equilibrium climate sensitivity means. It means, very roughly, that if CO2 stops rising after 2100 then a few centuries later (depending on how long the climate takes to reach equilibrium) the climate will have risen 3.7 °C over today.

        Actually even that is not quite right. More precisely the definition of ECS assumes that the starting temperature is measured while in equilibrium, whereas today we’re well out of equilibrium.

      • @cd: I still like my three basins controlling the resonance. The changes in the panama flow and Antarctic circumpolar currents would vary the transfer and the frequencies.

        While I can’t see it, cd, isn’t that what belief is all about? The ability to see only your own belief and not those of others.

        If I could see two hypotheses equally clearly I’d be of two minds on the question and would be casting around for a crucial experiment that could improve the odds of one of them.

        A year ago I was of several minds on the matter. What eventually became the knock-down argument for me is the correlation between the AMO and not the LOD itself, which is not that great, but rather its expected thermal impact if the mechanism is centrifugally driven lava, for which the correlation is great.

      • @eem: Vaughan, a climate sensitivity of 1.73C per doubling of CO2 looks believable, about 50% more than what would be expected with CO2 without any feedback.

        Thanks, Erik, that’s indeed the case.

        One should bear in mind however that what we’re currently experiencing is neither equilibrium climate sensitivity (ECS) nor transient climate response (TCR) but what some have called the empirical or observed climate sensitivity (OCS might a good acronym).

        The differences are substantial.

        ECS is the eventual temperature rise following a doubling of CO2, which may take centuries to reach that temperature, necessarily long after the doubled CO2 has stopped changing. As the temperature approaches its equilibrium value its slope decreases until in the limit it’s flat, the equilibrium condition.

        TCR is the rise in 20-year climate during 70 years of a CO2 CAGR of 1%. It is measured at the start and end of the 70 years, and the rise is the difference. A CAGR of 1% over 70 years corresponds to a doubling in CO2 over that period. Unlike ECS, there is no requirement that either CO2 or temperature cease increasing at any point, but if the CO2 did stop rising at the end of 70 years the temperature would continue to rise, much as a capacitor’s voltage will continue for time RC to rise when a ramp from 0 to 1 volts is applied across a series-connected RC network. Hence ECS is guaranteed to be no less than TCR and is almost certainly considerably more given the expected settling time of hundreds of years. Currently TCR is only meaningful for models, but RCP8.5 (“business as usual’) CO2 for 2010-2080 should be quite similar on average since it will hit the 1% CAGR mark shortly before mid-century and then rise above it.

        Empirical climate sensitivity is like TCR but with CO2 doing its own thing instead of obeying some rule like a fixed CAGR. In the 1960s CO2 CAGR was around a quarter of a percent, by 1980 it had climbed to 0.36%, by 2000 0.5%, and today it’s around 0.65%.

        The advantage of plotting climate against forcing instead of time is that it tends to factor out any changes in CAGR because CAGR is defined in terms of time while forcing is not. This is what makes it possible to observe a straight line after removing the known oscillations and TSI. The longer the observation, the more confident one can be about whether the trend line has any real meaning.

        I say “tends to” because in practice if forcing were to slow down it would give the ocean more time to absorb heat and rise in temperature, making the situation slightly more like the equilibrium case. While this might sound hard to model, it turns out to be much simpler than you’d expect if you judged it by the enormous complexity of global ocean-land circulation models (GCMs) like the CMIP5 suite.

        So far forcing has not slowed down, and so we have had no opportunity to date to observe what would happen if it did. We can only theorize, but that’s not necessarily a tall order, the complexity of the CMIP5 models notwithstanding.

        An empirical climate sensitivity of 1.7 or 1.8 °C is entirely consistent with an equilibrium climate sensitivity of 3 °C. For the purposes of 21st century forecasts, ECS is entirely irrelevant. Moreover if CO2 stops following RCP8.5 then even observed climate sensitivity will become less relevant. Theory (good theory anyway) then becomes the more reliable basis for projection.

        If CO2 continues to follow RCP8.5 it will reach 936 ppm in 2100. If in doing so climate continues along that straight line with slope 1.73 °C per doubling of CO2, climate in 2100 will be 1.73*log2(936/400) = 2 degrees hotter than today (so 3 degrees hotter than in 1910).

        Naively using the 3 °C/doubling figure for equilibrium climate sensitivity might make you think that the rise would be 3*log2(936/400 = 3.7 °C hotter than today in 2100. However that’s not what equilibrium climate sensitivity means. It means, very roughly, that if CO2 stops rising after 2100 then a few centuries later (depending on how long the climate takes to reach equilibrium) the climate will have risen 3.7 °C over today.

        Actually even that is not quite right. More precisely the definition of ECS assumes that the starting temperature is measured while in equilibrium, whereas today we’re well out of equilibrium.

  19. I welcome any effort to improve the observed temperature data sets. It may well take some time and effort to analyze this most recent attempt.

    Recent efforts have shown that obtaining an accurate observed series is work in progress and that the results like those from satellites remain open to interpretation.

    We have been taught and I have long been convinced that the natural variability is such in the observed temperature that a 15 year pause or longer is difficult to conclude. That a recent rendition of a mean global temperatue trend over recent time indicates that the efforts to explain it by noted climate scientists was in vain is not surprising and neither should be the observation that this new series will not seriously affect the mismatch of the observed series with the CMIP5 modeled temperature series over the period 1880 to 2005. Here I am not only considering the trends but differences in white and red noise and ratios of trends of the northern and southern heispheres

  20. C’mon, c’mon. Never mind all the fiddly bits.

    We have to find out who-did-what in 1910 in case someone does it again. Since the climate’s so anthro, we need to focus on human activities in the pivot year. What were people up to back then? Telephones? Kerosene? We mustn’t evade such an obviously critical point.

    1910, people!

    • PDO ramp up; solar; some anthro; OHT

    • 1910 was at the center of the lowest solar activity in the 20th century, so anyone starting with that year is cherrypicking.

      • A sunspot guy!

        I see great opportunities for adjustments here…although you’d have to do something about that high-soaring 1950s-60s cycle. Trying to flatten out solar cycle 19 could really strain your handle.

      • Yeah, that solar cycle (#14) was similar to the one (#24) we are in right now. Were temperatures the same as now?

      • stevenreincarnated

        “Were temperatures the same as now?”

        I would expect them to be if ECS was near instantaneous. Under what circumstances would you expect that outcome?

    • @mosomoso: We have to find out who-did-what in 1910 in case someone does it again.

      I’m not saying you did it, but something more convincing than your birth certificate that you aren’t the reincarnation of the culprit would help. :)

      “It” in this case is the angular velocity of the Earth, which was slower in 1910 than it has ever been at any other time in its 4.5 billion year history.

      You may need a spin doctor to get out of this one.

      • Vaughan, all the best spin doctors are heading to France, and I’m not consulting a crummy locum. No way.

        And how will we keep them down on the farm after they’ve seen Pa-ree?

      • Vaughan Pratt,

        You wrote –

        “It” in this case is the angular velocity of the Earth, which was slower in 1910 than it has ever been at any other time in its 4.5 billion year history.”

        I’m curious. Do you you mean the speed of rotation of the Earth about its axis, or the angular velocity of the Earth as it orbits the Sun?

        Can you give a reference showing either of these was at a minimum in 1910? I’m not disputing your statement, but it seems odd. Maybe I’m just unaware of the reason. Thanks anyway.

        Cheers.

      • @MF: Do you you mean the speed of rotation of the Earth about its axis, or the angular velocity of the Earth as it orbits the Sun?

        The former. Earth’s spin is what defines its length of day, typically given as the excess in milliseconds over 86,400 seconds.

        Can you give a reference showing either of these was at a minimum in 1910?

        https://hpiers.obspm.fr/eop-pc/earthor/ut1lod/lod-1623.html

        The longest day ever in Earth’s recorded history (averaged over a year) was in 1912 and lasted 86400.00389 seconds. An even longer day could happen in this century though it’s not guaranteed.

      • Vaughan Pratt,

        Many thanks.

        Also from your source is the following –

        “Actually the rotation axis is neither fixed with respect to the crust nor with respect to a celestial system, and the Earth’s rotation rate undergoes slight variations. The changes of the Earth rotation vector are caused by the gravitational torque exerted by the Moon, Sun and planets, displacements of matter in different parts of the planet and other excitation mechanisms. The observed oscillations can be interpreted in terms of mantle elasticity, earth flattening, structure and properties of the core-mantle boundary, rheology of the core, underground water, oceanic variability, and atmospheric variability on time scales of weather or climate. The understanding of the coupling between the various layers of our planet is also a key aspect of the research on the Earth’s rotation.”

        One of the problems with ascribing behaviour of any of the factors causing changes of rotation rate as resulting from causes of the rotation rate, is of course that the argument is circular.

        As an example, theorising that changes in rotation cause atmospheric variability, when atmospheric variability may cause rotation rate changes, is the same as saying that atmospheric variability may be responsible for atmospheric variability.

        I assume you will accept the authority of the link you provided.

        Quite complex, the Earth system.

        Cheers.

      • @MF: I assume you will accept the authority of the link you provided.

        Certainly. But how would it follow that I should therefore accept your total adulteration of it?

        If you live in one of those countries where the penalty for adulteration is stoning, you might want to keep your head down.

      • @MF: One of the problems with ascribing behaviour of any of the factors causing changes of rotation rate as resulting from causes of the rotation rate, is of course that the argument is circular.

        Please stop blaming your circular arguments on me.

      • Vaughan Pratt,

        I merely quoted from your source.

        What part of what I wrote do you disagree with? Or are you trying to blame me for pointing out you didn’t even read your own reference properly?

        A fact or two might be handy, but Warmist deniers seem partial to hand waving and evasion, rathet than discussing reality. You could try a spot of aspersion casting. If you need a bit of advice, I’d be glad to help.

        Cheers.

      • @MF: What part of what I wrote do you disagree with?

        Sorry, sometimes you ask fair questions, sometimes it sounds like you’re just trying to be difficult, sometimes it’s hard to tell. Judging by your tone just now it sounds like I should assume the former this time around.

        Everything in what you quoted is completely correct, is consistent with what I wrote, and is not at all circular. I could fill in the details here, but the answer is long enough and (I think) interesting enough to be worth a separate post on CE. I’ll send Judy something, though I imagine Thanksgiving is keeping her occupied just now.

  21. Tonight we left the hockey game @ 6;50 PM and my car temperature reading was 28 F. The hockey arena is at city center. The half moon in a clear sky provided light to my vehicle. A three mile drive into the suburban region and the car temperature reading was 24 F.

    It appears to me that the late Fall, early winter night time temperature readings, particularly to determine the Urban Heat Island effect, is substantial and probably provides a better assessment of the Urban Heat Island effect than daytime readings.

    I would be interested in the night time temperatures being collated and reported than the daytime temperature. Is there such an animal?

  22. Thank you, Professor Curry, for this guest post in support of NOAA’s temperature adjustments.

    Seth Borenstein expresses an opposing view of NOAA’s temperature adjustments below:

    https://stevengoddard.wordpress.com/2015/11/22/government-scientists-continue-to-ramp-up-their-criminal-activity-ahead-of-paris/

  23. Adjusting the record with number of days per year that the Tioga Pass was closed, the yearly start dates for Washington DC cherry blossom blooms and the number of malaria deaths per year in sub-Saharan Africa in children under 5 years of age might increase the accuracy the NASA temperature record even more… whatever shows more warming, right?

    • The whole point of this post is to evaluate the effect of NOAA adjustments using a single-instrument dataset (buoys) that do not require any adjustment.

      • Well, whether or not the point of the post is as you claim, there is also the fact that such studies will be held up by warmunistas, the clapping seals masquerading as journalists, and left leaning politicians as further proof that cagw is real and requires urgent and drastic action to decarbonize the world.

        As for your earlier comment about the IPCC sticking to a 6 month cut off date, Donna Laframboise has evidence to the contrary which she documented in the Delinquent Teenager.

      • The problem is that what should be a historic record is being used as a political tool.

        1. You can use the temperature record as a scientific toy that you play with.

        2. You can use the temperature record as an official record for official pronouncements.

        NOAA/GISS don’t get to do both. Attempting to do both should be banned by law.

        If the historic temperature record is going to show the stability of a video shot in an off road rally car there is no point in allowing it to be used for historic comparisons and that use should be banned by law.

      • Zeke

        Bearing in mind that we have no proper idea of a reliable ‘global’ sea temperature back to 1860 or 1880, at what point does the weight of reliable data become sufficient to use as a sensible base line? The 1950’s. The 1960’s?

        In other words how far can we look back to see if any definite trends are emerging? Thanks

        Tonyb

      • Judge ye by results
        and cui bono.*

        *H/t Cicero.

      • TONY!!!!

        can you at least stay on topic.

        Scientists were accused of fraud.

        congress has demanded their mails.

        Zeke has shown that using the best data he gets the same answer
        as these guys accused of fraud

        Judith has said the buoy data is the best.

        STOP changing the subject when men’s character has been assasinated.

      • MOSH!!!!!

        How many times have I said I do not believe in Fraud, Hoax or any permutation thereof?

        . Surely it is perfectly valid to point out that we are building a straw man by examining Global ocean temperatures that didn’t exist in any meaningful form until…Until When? By selecting a valid start date we can then start to see if any claims of incorrect adjustments are accurate or not.

        Please note that my counter blast is more powerful than yours as I have used FIVE exclamation marks and not a measly four. Anyway what are you still doing up?

        tonyb

      • I agree, TonyB, that the SST series haven’t been produced long enough for determining trends over the longer term. Mosher seems to think that this paper rebuts the critics of the Karl et al paper, simply by the production of another series of low resolution data with high error bands? ALL data to date, is unsuitable for policy makers IMO.

      • how far can we look back to see if any definite trends are emerging?

        About 13 billion years.

        A definite trend is a definite trend, Tony. If you saw a definite trend in CET, would you disqualify it merely because you were looking back more than three centuries?

      • VP your 13 billion period for trend detection in the 21st century seems to long. You only need about 100 years of data to show a trend IMO. The problem is to filter out the white, red, green and purple noise!

      • Vaughan

        Do you really believe that the trends in the Holocene ocean temperatures have any relationship to what happened in other eras when the world, its continents, its oceans and its climate were ordered entirely differently?

        My point is that we need to determine from when we have reliable data and it only since then that we can start to determine what has happened over that short period. Which may or may not have any relevance to the much longer time needed to take into account cycles within cycles.

        tonyb

      • Peter

        “I agree, TonyB, that the SST series haven’t been produced long enough for determining trends over the longer term. Mosher seems to think that this paper rebuts the critics of the Karl et al paper, simply by the production of another series of low resolution data with high error bands? ALL data to date, is unsuitable for policy makers IMO.”

        1. Do you think Karl committed Fraud. yes or no?

        2. YOU are not a policy maker, they can use anything they damn well please.

      • ==> “2. YOU are not a policy maker, they can use anything they damn well please.”

        Peter said that in his opinion, the data collection period wasn’t sufficient to inform policy. He didn’t say that policy-makers couldn’t or wouldn’t use those data to inform policy. (In other words, read harder).

      • Steven, I don’t believe that Karl et al was being deliberately fraudulent, biased yes but not fraudulent. J (whose name shall not be mentioned!) had expressed correctly what I was trying to say: its my opinion only and I have always consistently commented that people are trying to do too much with too little data. Policy makers form part of this population.

      • Try to get a holt of yerself, Steven. The Congressional oversight committee is not accusing the NOAA people of “fraud”, yet. They are investigating whistleblower allegations of undue political influence, and failure to follow established NOAA procedures and methods. Stop the hysterical foolishness.

      • It isn’t a case of fraud per se. The data was altered to serve political ends in a way supported by the biases of the scientist involved.

        That isn’t fraud. The adjustment itself might actually be defensible. It could have been done on his own initiative. But the problem has been around for about 20 years and the timing is odd.

        The question that warmers keep dodging is was the change in response to outside influence (administration or otherwise)?

        The claim isn’t that the “books were cooked.”

        The situation situation as related by the whistleblowers is more like:
        Karl is braising the data on the grill while partaking of beer and chips. Obama while doing wine and cheese taps him on the should and says, “I want my data burned crispy”.

        We have a legitimate right to know what exchanges happened between Karl and outside (possibly administration) influences.

      • Very good, PA. Somebody is paying attention to the crux of the matter.

        We would also like to know how many perfectly plausible adjustment scenarios they rejected, because they didn’t result in erasing the pause that is killing the cause. And what took them so many freaking years to do this? Why didn’t researchers working on other series do it long ago, if it is so necessary and evident? Smells fishy.

      • The SST smells fishy. because there are fish in it.

      • It’s clear to me that something is not right with the “hiatus buster” paper by Karl et al. The Congressional investigation is just sauce for the goose. Anyway, I was wondering as a Veteran of the U.S. Navy, buoys can demonstrate a positive bias at night in calm seas as the structures are heated during the day and radiate into the immediate area at night. This effect is similar to a parking garage rising in temperature at night after a hot day, acting like an oven, rising higher than the daytime max. I was wondering what your thoughts were on this possibility. I know that water has a high heat capacity but a local environment may be constructed around the buoy, enough for a positive bias reading, especially in littoral areas. Maybe there were in-situ experiments performed?

        Respectfully,
        JerryG∞

  24. Thanks to the authors for writing about and discussing this.

  25. All this started with an AGW hypotheses about man made CO2 causing a tropospheric hot spot, melting of both the Arctic and Antarctic, and atmospheric temperature increase. None of that appears to be happening, so we are talking about sea surface temperatures. Are we now discussion a new hypotheses? If so what shall we call it?

  26. by Zeke Hausfather and Kevin Cowtan

    “A buoy-only sea surface temperature record supports NOAA’s adjustments.”
    yet
    “ship and buoy SSTs are measured differently and at slightly different water depths, thus there is a statistical difference of ship-buoy SSTs of about 0.12ᵒC.”

    So much to ask.

    Which buoys are we talking about?
    What level are we measuring for SST?
    How can we compare buoys with ships NOAA adjustments when they occur at different levels?
    Are their important differences between the different types of buoys used?
    Who owns/runs the different buoy sets?
    How many buoys are there?
    Is ARGO included?

    Then important stuff,
    If there are only a few thousand active buoys, with limited half lives, floating in the ocean hence moving in position,how many are original from 1995? how many have been replaced?
    [Mosher is very insistent that a replacement actually causes a new record to be formed].
    How many new ones are being added?
    what do you do with data when a buoy “dies” which happens reasonably often going by ARGO?

    Finally,
    Are the results and grids then infilled [because there are so few accurate working buoys]?
    How many infilled grids are there?
    90% of the 70% areas we have buoys in, real and modeled?
    How much of the data of the buoys is actually synthetic [modeled] data?
    Why do you chuck out the inconvenient stuff like buoys that do show real variation?
    [You also did this with your “inconvenient truth ” Russian island arctic temp measurements].
    I would happily state that if put 10 of your buoys in a half kilometer radius anywhere in the sea that the variations would be at least 0.2 degrees C in most locations, not 25 kilometers.
    Why do you Guarantee that such accuracy can extend 25/50 or 550 kilometers ?

    Quotes you really said , fellas.

    “the daily data are placed into 550 km equal area grid cells based on the location of the buoy for that day,”
    “preliminary analysis on the homogeneity of the buoy records by comparing buoys within 25km. For most buoys the differences are 0.1-0.2K”

  27. “A buoy-only sea surface temperature record supports NOAA’s adjustments.”

    As a statement means zip.
    “makeup sales were way up, which scientists believe is a bad sign for the economy.”
    Same logic, or lack thereof, confusing coincidence with causation or proof.

  28. It seems everyone including Judith has missed the point.

    1. Karl did adjustments.
    2. Skeptics screamed fraud, and demand emails.
    3. Skeptics said the bouys the bouys the bouys are best

    Now H&C demonstrate that a bouy only record gives the same
    answer as Karl.

    Raising the issue of other NOAA datasets.. is a diversion from the real issue.

    That issue is the crass claims of fraud and the fricking witch hunt.

    • What else is newt?

    • Steven Mosher,

      You wrote –

      “Now H&C demonstrate that a bouy only record gives the same
      answer as Karl.”

      You seem to be claiming that something gives the same answer as something else based on something different, provided it is properly adjusted.

      H says it has nothing to do with CO2, apparently. As a matter of fact, one completely pointless assertion has been claimed to be the same as another. Both are supposed to be useful for something, but nobody knows what it is.

      I don’t know about fraud. Maybe silliness, incompetence and ignorance, driven by a desire for recognition are sufficient explanation.

      How are your efforts to convince the Chinese that smoking is a health hazard going? I haven’t seen anything in the news, lately.

      Cheers.

    • Mike Flynn wrote:
      You seem to be claiming that something gives the same answer as something else based on something different, provided it is properly adjusted.

      At the very top ZH and KC write
      A buoy-only sea surface temperature record supports NOAA’s adjustments.

      … We find that a record using only buoys (and requiring no adjustments) is effectively identical in trend to the new NOAA record and significantly higher than the old one.

      and Mosh wrote

      Now H&C demonstrate that a bouy[sic] only record gives the same
      answer as Karl.

      Who is saying what seems pretty clear. Take a deep breath.

    • Well, the point is this…

      The ONI is going to be taking a nose dive shortly and will be entering La Nina territory some time around August/September.

      Some time between August 2016 and December 2017 either a new “pause breaking” adjustment will come along, or the monthly “adjustments” will continue their inevitable climb toward infinity.

      At this point it doesn’t matter what rationalization is produced. Once the trend becomes the product of the post 2008 adjustments the claim that what is being presented is an accurate “surface temperature record” becomes a joke (if it isn’t already).

      Then we can start with a cultured and refined claim of fraud and start a non-fricking perfectly justified witch hunt.

      • @PA: Then we can start with a cultured and refined claim of fraud and start a non-fricking perfectly justified witch hunt.

        Trying for a Guinness record in being wrong at the top of our voice, are we?

      • The ONI is going to be taking a nose dive shortly and will be entering La Nina territory some time around August/September. …

        This is what could happen, but there is no way to know. ENSO defies predicability at that distance in the future.

        It could also settle into a long period of neutral, in which the temp might slide back to the GMST depths of 2014 – the warmest ENSO neutral in the instrument record.

        Or is could be like monstrous la Nina events of the past when the gMST went up because of the ramp up of the PDO. Wouldn’t that be a hoot? Skeptics could get their chilly La Nina dream, only to wake up in a warm ACO2 blanket.

      • Two points.

        1. VP I have an objective standard. Take the data with 1999 as a base year. Substrate the next year and compute the net trend. If it is positive it is Heads if it is negative it is tails.

        Compute the number of heads and tails and when the result is 2 sigma we start firing people.

        People with common sense know that historic data is dead and this whole thing is a joke anyway. The concept that historic data is moving around is absurd.

        2. JCH:

        You know what they say about karma. You knew La Nina was coming. La Nina is coming. Predicted trend indicates she isn’t in a good mood.

        The “deniers” will be dancing around and screaming “the pause lives” and of course be right. The trend out to 2030 is widely predicted to be flat. The pause denial is a stupid waste of time and money done for political reasons.

      • It’s predicting better odds for ENSO neutral in JAS than for La Nina.

        And it does not matter. La Nina is not going to fix your bad science.

      • JCH, What science? PA, has admitted he isn’t a climate scientist. I am think Dunning Krueger applies in his case.

      • “I am think Dunning Krueger applies in his case”

        Gee, I’m just reading a chart. Are you claiming there won’t be an La Nina? Are you claiming it won’t happen in the last half of next year?

        Roll the evidence. Present your data.

        “.And it does not matter. La Nina is not going to fix your bad science.”

        I’m not the one dancing through hoops to eliminate the pause.

        I expect it to be warming. I’ve said that before. It takes 260 WY/m2 to warm the top half of the ocean 1°C. At 0.2-0.3 W/m2 it takes a long time. The warming from the solar change in the last century hasn’t been fully incorporated.

        The pause has surprised me because it means GHG forcing is relatively insignificant. It isn’t driving the bus – it is a little kid in the back seat.

        Global warmers take great pains to combine the effects of ALW, AGW, and natural variation and attribute it all to global warming. That’s goofy.

        The way to figure out whether these adjustments to the historic record are legitimate is to assign an engineering/statistical team to “cool the planet”. They would go through the data and the current methodology and exploit every adjustment that cools the data and invalidate every adjustment that is technically problematic (statistical, theoretical, or other issues) They would get a bonus of 100% for each 0.1°C they cool the planet. The only restriction is they have to be as honest as the current team (if they cheat – they have to point to where the current team cheated), and they have to have a justification for their adjustments..

        Now do you think their temperature trend will be higher, lower, or the same?

        If the science is solid there should be no difference because the positive adjustments are all solid science and there are no significant cooling adjustments left.

        There is a claim out there that the current adjusters because of cognitive bias are picking more warming adjustments than cooling adjustments from the hundreds if not thousands of potential tweaks available.

      • The pause has surprised me because it means GHG forcing is relatively insignificant. It isn’t driving the bus – it is a little kid in the back seat.

        This is what I mean. You really don’t have the requisite background knowledge to draw that conclusion. You confidently pontificate on a number areas related to climate science and other subjects as though you do.

      • Joseph | November 23, 2015 at 12:19 pm |
        ..
        You confidently pontificate on a number areas related to climate science and other subjects as though you do.

        This is indefensible. This is wrong. Historic data is not a moving target.

        CGAGW is 55% of AGW and rising. The problem isn’t the 55% the problem is that it is constantly rising. This should be stopped, by law if necessary.

        If adjustments to the past are frozen, the adjustments to current data would become very significant and difficult to explain. They have essentially have added 0.25°C+ to the 21st century temperature record since 2008 by spreading it back through the previous 105 years.

      • This is indefensible. This is wrong. Historic data is not a moving target.

        I don’t know enough about the subject to give an opinion. So you expect me to believe that those who are experts in this area are wrong and you are,right? Why?

      • There could be a La Nina. There could be prolonged ENSO neutral, followed by either an EL Nino or a La Nina. There could be a very brief neutral followed by either an El Nino or a La Nina.

        That forecast, the one you are showing, is showing neutral is the most probable.

        The temp will go down. That won’t save you. Having more “the warmest” La Nina events in the record will not help you. La Nina does not reverse EL Nino. But heat the tar our of some section of the globe. There is no respite. That is why the pause can barely fog a mirror.

      • Joseph | November 23, 2015 at 1:17 pm |
        This is indefensible. This is wrong. Historic data is not a moving target.

        I don’t know enough about the subject to give an opinion. So you expect me to believe that those who are experts in this area are wrong and you are,right? Why?

        You do know weather stations measure the temperature don’t you? You can buy a thermometer at the store, put it outside on top of a pole in a slatted white box, and the thermometer will automatically go into CRS emulation mode.

        Now in theory moving the station, measuring temperature at some time other than midnight, replacing the instrument, etc. induce biases. Adjustments in theory remove this bias.

        But midnight isn’t randomly shifting between 6 PM and 6 AM, and no one is going back in time to move the station or swap instruments. So the historic temperatures (pre-2000) shouldn’t be changing on a monthly basis for a 0.25°C+ change since 2008.

        When you mention adjustments to most of the other engineers on the forum the response is probably “Well, Duh!”. But I suspect the engineers in the “historic temperature is a moving target” camp is a much more selective clientèle..

        However – since we are on the boaty and bobby things thread, it occurs to me that a plot of the raw CRS, MMTS, ASOS, etc. surface data trends grouped by instrument type would be interesting.

        BEST should already be doing this division since a change of sensor should be treated as a new station.

      • Are you going to answer my question, PA?

      • Lots of things would be interesting. None of them will save you guys. ACO2 should have made the earth warmer to date, so the monitoring systems have likely missed some of it. That does not mean it is not here; it means it is here.

      • David Springer

        Joseph no one here expects you to believe them and it isn’t because they aren’t qualified. You don’t believe Curry, Spencer, Lindzen, Ball, or any number of others who are qualified so there’s no reason at all you’d believe anyone less qualified. Your belief system is that of a fanatic and they only believe what they want to believe. No amount of simple contrary evidence will change your beliefs unless you first wanted to change your beliefs.

        Thanks for playing. With friends like you the CAGW crowd doesn’t need enemies.

    • Steven Mosher:

      Now H&C demonstrate that a bouy only record gives the same answer as Karl.

      Look at the data again. Read more, talk less. Yadayadayada ;-P

      They don’t give the same answer. They give similar trends if you cherry pick the interval, but that’s about it. They are very dissimilar between circa 2000-2008.

      In fact, I’d say there’s “no agreement among the three series for that period”.

      I agree with Steven about fraud (but my feelings are a bit too harsh towards the faudulators to relate them in polite company). Though I don’t doubt that there is a substantial confirmation bias in terms of which periods (and regions, cough) get examined whilst people grapple with why the measurements don’t agree better with the models.

      I think Judith is closer when she says there are still unanswered questions here.

      I find it a bit curious to see Zeke insisting that an offset shift is “trend neutral”. (I happen to agree that an offset shift needs to be applied.)

      Of course it’s not trend neutral! It’s simple mathematics that it will increase the trend (for this particular offset shift). And that’s why the trend is “slowdown busting”.

      It happens, by basic math, that as you increase the period you’re looking at, the influence of that offset shift on trend gets watered down, but that’s a very different thing that claiming an offset shift trend neutral.

      • Hi Carrick,

        I’m not arguing that an offset shift is trend neutral vs. no offset shift. Rather, I’m pointing out that the sign of the offset shift (e.g. buoys up to ships or ships down to buoys) is nominally trend neutral, pointing out that Kennedy and colleagues explicitly tested this with HadSST3.

        I agree that there are still some unanswered questions (after all, HadSST3 also gives rather different values over this period). But the buoy record lends support to the idea that the ERSST v4 adjustments are not unwarranted politically-inspired meddling as some have claimed :-p

      • Whether the buoy adjustments are warranted is an issue that looks to be unsettled. Another issue is: Were the adjustments prompted by unwarranted politically-inspired meddling and were they done disregarding established NOAA procedures and methods? You wouldn’t know anything about the latter issue.

      • Carrick

        “They don’t give the same answer. They give similar trends if you cherry pick the interval, but that’s about it. They are very dissimilar between circa 2000-2008.

        In fact, I’d say there’s “no agreement among the three series for that period”.

        How long will you feed the likes of goddard and the tribe at WUWT?

        You know exactly what they are charging. You know that Zeke’s test
        shows their criticisms and conspiracy theories are bunk.
        If Karl said 10.23 and zeke said 10.22…. would you still pull a brandon
        and say there was no agreement?

        The the agreement will never be perfect but its substantial and should indicate to any rational person that the books are not being cooked.

        Watch how critics all slither back and change the topic to other datasets.

        Watch how tony B pops up to slime the rest of the record.

        At what point is the statement “Karl commited fraud, or malpractice, ” falsifiable?

        Not to mention the dopes who are using KNMI as a definitive source

      • Mosh

        Having already said twice on this thread that I don’t believe in fraud and that goddard needs to go to peer review if he wants credibility in his vendetta against giss i think that was a very foolish thing for you to say I would pop up to ‘slime the record’

        Where have I said that? The data is going to be fine from a certain point in time. Prior to that our records are far too irregular and the methodology to poor to have any scientific reliability. At what point that changeover occurs is surely a useful thing to ascertain?

        Tonyb

      • Zeke thanks for the reply. That helped me understand what you were saying.

        Don Monfort—it’s clear to me that the two data sets (meaning ship vs buoy) have very different sampling regions. So a correction of some sort needs to be made. I’d say we’d end up very lucky if it were as simple as a pure offset. The use of an offset in lieu of a more complex model certainly needs to be justified, not just assumed.

        Steven Mosher—as far as I’m concerned Goddard is a charlatan. If an error in the Pause Buster, it has been made in focusing on corrections that push towards agreement with preconceptions. This does happen in science (I’ve witnessed 9-sigma results vanish after independent review for example). In that respect the buoy comparison is an important effort for validation.

        I would say the agreement is far from perfect for circa 2000-2008. It’s because the two series agree at the start (1995-2000) and at the end (2010-2014) that the trends match up so well. As you know, linear trends weight the points near the edge the heaviest.

        I probably would have used 2000-2014 as a second period to avoid concerns about (unconsciously) cherry picking the intervals that give the best correspondence.

      • David Springer

        Hausfather you still have not explained why we see a step change of 0.06C in the summer of 2006 when ERSSTv4 is overlaid on OISSTv2.

    • You are ignoring the central fact, Steven. Whistleblowers from the NOAA, described as scientists, have made allegations. Some anonymous band of skeptics didn’t create this situation. The Congressional committee is investigating the allegations of the whistleblowers. You got a problem with that?

      • Interesting. Do you have a link to this whistleblower story?

      • Whistleblowers are protected. where are the statements.

        Also, WHY doesnt smith want to make transcripts available?

      • The only whistle blower allegation I have seen is a rush to publish which could be something like announcing ERSSTv4, but v4 only runs from 1995 to present and the rest of the 1880 to 1995 is sort of not finished quite yet. Zeke and Kevin have done a fine job ‘splaining what is available, but is Karl et al a complete ERSSTv4 or just a part with great headlines?

      • Whistleblowers are protected.

        Ask Edward Snowden how well that works in practice.

      • Read more, comment less.

        Chairman Smith spells out what information he is operating on in the letters he has sent to NOAA. How many times do you people have to be told. It ain’t primarily about the rushing. That is a red herring the stonewallers have waved under your noses.

        Hey, ask Judith. She has talked with NOAA people. What has she said about the investigation? Read more, comment less.

        And the pause is killing the cause.

      • Smith doesn’t want to make transcripts available so that the other side doesn’t know what he has, until he has gathered as much evidence as he get his hands on and interviewed all the people he wants to interview. Watch some cop shows on TV, and comment less.

      • Smith doesn’t want to make transcripts available so that the other side doesn’t know what he has

        Well, Don, if he made the transcripts available then it would make his case for getting the emails stronger and weaken the political resistance. I don’t know what letting the other side know has to do with it. What are they going to do with that information?

      • Don, “It ain’t primarily about the rushing.” I don’t know about that, rushing to publish and rushing to publish prolifically is a major problem that leads to lots of other problems.

      • So no story until he releases something. Until then we can all tell each other made up stories about what this all means.

      • First thing yoey, what transcripts are you talking about? Second, his case for getting the emails is as strong as it is going to get. How about the Congress has broad oversight responsibilities and powers per the U.S. Constitution and various supporting Supreme Court decisions. That includes the power to subpoena. The political appointees who run federal agencies and their minions, including scientists, don’t get to decide what they will turn over. Read the Supreme Court decisions and comment less. Google it. I don’t have time to educate you jokers.

      • Please, capt. Read the letters to NOAA by Chairman Smith.

      • Don perhaps you have a link to the letter you are concerned with. The Nov. 18 letter appears to be the first mention of whistle blowers and that just indicates the whistle blowers mentioned rush to publish despite objections, but not what the objections were. The first letter is July 14 and it mentions reasons for rushing, CPP and Paris indicating that NOAA may be playing politics, but Obama is kind of the boss when it comes to scheduling. What I don’t see is any specific problems mentioned by whistle blowers other than implied by their objection to rushing the project. What else specifically did the whistle blowers object to?

      • The letter clearly lists the categories of objections by the whistleblowing NOAA scientists, Tony. The specifics will come out in due time. The committee is not going to tell the NOAA stonewallers all the details. That is not how an investigation is conducted. The committee is asking the questions.

        http://cdn.arstechnica.net/wp-content/uploads/2015/11/lamar_smith_noaa_comm_letter_nov18.pdf

        It’s self-explanatory. All BS aside, the NOAA is obligated to answer the committee. Period.

      • PS: Focusing on the word “rushing” is missing the point. The objections of the NOAA scientist whistleblowers are not about going fast, but about not completing steps in NOAA’s established procedures and methods. It’s all in the letter and the committee doesn’t have to explain squat to the NOAA.

      • Here is some context, Tony. Smith’s reply to that letter from Eddie Bernice Johnson, the scheister defending NOAA from Congressional oversight:

        https://science.house.gov/news/press-releases/smith-questions-ranking-member-s-commitment-oversight

      • The timing of the complaints, from the letter to NOAA, does not make sense. The paper was submitted in December 2014, so by April-May 2015 it was already in the final revision stage and it was accepted for publication on 21 May, and so the June complaints were even after that. Why were all the internal complaints only well after it would have been all-but-accepted, and why would they be complaining about an accepted paper in the first place?

      • There is a lot of good background information in there. This is particularly pertinent for those who yammer about fishing expeditions, yes you yimmy!:

        “Additionally, your characterization that oversight should only be performed if evidence of misconduct, fraud, or abuse of discretion is found is mistaken. This Committee has a duty to “determine whether laws and programs addressing subjects within the jurisdiction of [this] committee are being implemented and carried out in accordance with the intent of Congress and whether they should be continued, curtailed, or eliminated.”[26] The Rules of the House direct standing committees to conduct oversight on a continuing basis, regardless of whether wrongdoing is suspected. To that end, Congress has a constitutional responsibility to perform oversight, and that oversight may be far-reaching. In Watkins v. United States, the Supreme Court ruled that the “power of the Congress to conduct investigations is inherent in the legislative process. That power is broad. It encompasses inquiries concerning the administration of existing laws as well as proposed or possibly needed statutes.”[27]”

      • Eddie Bernice Johnson was involved in a scholarship scandal a while back.

        Scholarship checks for her out of district (ineligible) relatives were sent to their homes instead of the college.

        It is interesting reading. She does not seem to be a paragon on virtue.

      • Somebody at NOAA was trying to quash, delay or slow-walk an already accepted paper that they knew would kill “the pause”. They are the ones that need investigating.

      • Don, Well, “rushing” due to a political need should be pretty obvious and isn’t particularly uncommon. All the whistle blower complaints are pretty common for any rushed project. I have written a few CYA letters myself on government jobs.

        What is boils down to me is who discovered the need to produce a fairly small change and hype it just in time. If it was Obama, unfortunately that is his privilege. If it was Karl, then that is bordering on activist “science.” There are plenty of “acceptable” ways to fudge things a touch if properly motivated. Some require more CYA letters than others. Science isn’t completely immune no matter how noble scientists think they should be.

        btw, I don’t think rushing is a particularly useful squirrel for the believers and since so many government screw ups are caused by rushing, it hits the nail on the head. Just about every real problem in Climate Change boils down to rushing and motivating “immediate action” using worst cases and precautionary principles. That eliminates the need to worry with those pesky error bars. Administrative opinions of nuclear power inspired the same “science” in first one then the other direction.

    • Steven Mosher, there is entirely too much time spent on this thread by some people flipping off the work done by Karl et al and here by Cowtan and Hausfather instead of doing a detailed analysis and/or at least looking at detailed analyses before making comments. I welcome work on new temperature data bases and probably as much for being able at some point in time doing an analysis of that work to determine how strong the evidence is for its correctness and whether all the proper statistical and analytical tools have been applied. I would do this as an interested and skeptical party.

      I therefore have a problem with your continual reference to skeptics in generalized terms and almost always in negative terms. Surely you must agree that there are “skeptics” on all sides of the AGW issue – otherwise there would not be any science in the matter.

      I am hoping that the day is fast approaching when all the algorithms applied in adjusting the temperature series for the various better known and used data sets are tested with a benchmarking test and a benchmark that truly tests the limits of the adjustment methods. Now that is something that a true skeptic can get his/her teeth into.

      • Lots of folks are working on benchmarks at the moment. For land records both Williams et al 2012 and Venema et al 2012 did a pretty good job at evaluating the effectiveness of homogenization using blinded tests. The International Surface Temperature Initiative has a working group to create a new series of benchmarks that various groups around the world working with temperature data can submit their methods to be evaluated. You can find updates on their progress here: http://www.surfacetemperatures.org/benchmarking-and-assessment-working-group

      • Both pause and pause bust are tenuous ( given the brief duration ) and not mutually exclusive ( No MUTEX ).

        SSTs and Land Ocean Indices are at all time highs. Well, were AGW the only game in town every year should be an all time high, not just El Nino years with the Blob prevailing for the first half.

        On the other hand, MSU trends are still negative since 2001 and the Land Ocean index is about a quarter of the 0.2 per decade that the IPCC guaranteed.

        None of the trends are statistically significant.

      • Zeke, I am aware of the status of the recent benchmarking initiative and was hoping for more rapid progress.

      • ken

        ” I welcome work on new temperature data bases and probably as much for being able at some point in time doing an analysis of that work to determine how strong the evidence is for its correctness and whether all the proper statistical and analytical tools have been applied. I would do this as an interested and skeptical party.”

        You possess the skills to do your OWN homogenization approach and let others have a wack at it.

        or..

        Start with NOAA’s approach. Show how it can be improved

    • Mosh, you and Karl have to put on your big boy pants. Pielke Sr. has pointed out that Karl suppressed research in the past. See https://pielkeclimatesci.wordpress.com/2011/07/11/new-article-comments-on-%E2%80%98%E2%80%98observed-trends-in-indices-of-daily-temperature-extremes-in-south-america-1960%E2%80%932000%E2%80%99%E2%80%99-by-r-j-stone/ Additionally, James Hansen advocates the jailing of people who provide energy. If the warmists are going to suppress and imprison people, they have to accept that they will be investigated. Whether or not wrongdoing is suspected, Congress, and others, have the right to investigate. When warmists scientists get out of the business of advocating huge economic changes, they have the right to be mostly left alone. That is not the case now.

      Additionally, Pielke Sr. noted one time that Karl peer-reviewed himself. There is no reason to give his work any presumption of validity. Pielke stated: “In fact, as stated above, the CCSP Report “Temperature Trends in the Lower Atmosphere: Steps for Understanding and Reconciling Differences“, with its documented bias, was chaired by the same person as the Review Editor of the IPCC WG1 Chapter 3 Report (Tom Karl)! Regardless of his professional expertise, he is still overseeing an assessment which is evaluating his own research. There cannot be a clearer conflict of interest.”

      JD

    • Mosher, you forgot about the whistleblowers.

  29. Zeke Hausfather,

    You wrote –

    “Regarding your discussion of CO2, it’s not really germane to this discussion. We are talking about evaluating the effectiveness of adjustments done to ocean temperature records.”, in response to me asking what your work was supposed to achieve.

    I appreciate that you may feel an uncontrollable desire to furiously analyse historical figures due to some form of OCD. I agree that you don’t need a particular reason, and it may have nothing at all to do with the peculiar idea that CO2 in the atmosphere makes the surface warmer, winter or summer, day or night.

    I restate my original contention that if you haven’t any idea why you are trying to make two dissimilar sets of figures agree, then you are wasting time and money. If it’s your own, I don’t give a fig. If it’s someone else’s money, that could have been spent usefully, I might well demur.

    Your choice, I guess. It still seems about as pointless as Steve Mosher’s bizarre preoccupation with historical temperatures. They were what they were. They do not carry any predictive ability whatsoever. Carry on regardless!

    Cheers.

  30. What a surprise – the “new” version shows more warming than the “old” version. An interesting project would be to look at all the new versions that have been introduce to GHCN, HADCRIT, NOAA etc and work out what proportion of them made the warming look greater and what proportion made it look smaller.

    • Karl shows a graph with no adjustments over the last century, and the change in the 1940’s was much larger and in the opposite direction towards a cooling trend. No one complained about that one, and many have forgotten.

  31. Interesting post.

    From visual inspection it seems the deviation mostly occurred in 2005-2010.
    Would be interesting whether it is possible to track down further what exactly changed in those years. The boy coverage seems to have increased a further 10% or so. If a ten percent increase in coverage has such an impact I believe further scrutiny might be warranted. And one would want to wait what the next 10% might bring.

  32. Karl introduced a spurious adjustment to temperature trend given by 0.12(?) x (slope of proportion of buoy to ship data). Helpfully figure 2 above shows this is a large positive adjustment. This is schoolgirl mathematics – no amount of ‘big data’ analysis can change that.

    A question is how a properly implemented adjustment (although how?) for the changing composition of the data might compare to this botched adjustment. But this post takes (by the authors’ own admission) partial and rapidly changing buoy coverage over 20 odd years. The scope for all sorts of bias in this should be obvious. What is the point? If anything, it leaves me thinking a constant offset looks even more ludicrous.

  33. Icosahedral-symmetry coverings, packings and max vols of the sphere are better than the lat-long approach, available for points into the tens of thousands.

  34. I look at Figure 3 after reading comments on trends and in particularly differences in trends…and then shake my head and walk away. Pretty darned presumptive stuff going on.

    • Error bars in RHS of Figure 3 are not defined and are totally useless.

    • Use visual and call it a day. Good enough for government work.

    • ZH&KC–Thanks for the good post and the code. BTW ran on OSX 10.11.1 by stripping comments from the file and C&P to running csh shell in bash. [Comments (#) not playing nice in script.]

    • Hi mwgrant,

      We should have had more detail in the original post on what the error bars in Figure 3 represented. They are simple 95% CIs for an OLS regression on monthly anomalies. If we had used an approach that accounted for autocorrelation (AR1 or various ARMA), they would have been wider.

      • Hi Zeke,

        Thanks for crossing that ‘t’. That was my presumption but I have grown very cautious with blog numbers. [SLR of course assumes independence of the observations…Thinking about it, I assume you and KC were downplaying quantitative characterization of the trend(s)? Afterall it is the close tracking of the two line plots that tells the story and chasing statistical significance is an overly messy complication at this time and place. Apologies for the diversion. ]

  35. Assuming he pays attention to this, Smith can go either of two ways. He can retract his subpoena, due to the Karl paper being easily supportable by independent evaluations, or he can look for emails between the Obama administration and these authors too.

    • Jim D,

      Or me might do whatever he wants, and ignore both of the choices you gave him. Who is Smith, anyway?

      Cheers.

      • Smith is the chair of the science committee who didn’t like Karl’s results and so wants to see his emails to check if Obama sent him anything.

    • See Carrick above, yimmy. Chairman Smith has information and very likely emails and other documentation of the Karl shenanigans from whistleblower NOAA scientists. It’s going to ugly, yimmy. The enforcement powers for Congressional subpoenas are iffy and can take forever, but the emails will be pried out of those clowns through a citizen FOIA.

    • Jim D, The information Smith has has to do with two issues:

      1.)The research, considered a bombshell in the climate change debate, set off alarms among skeptics. Smith, a prominent congressional skeptic, claimed that scientists manipulated data to advance President Obama’s agenda and timed the study’s release to coincide the the administration’s new limits on emissions from coal plants.

      2.) Whistleblowers have told the committee, according to Smith’s letter, that Thomas Karl — the director of NOAA’s National Centers for Environmental Information, which led the study — “rushed” to publish the climate study “before all appropriate reviews of the underlying science and new methodologies” used in the climate data sets were conducted.

      https://notalotofpeopleknowthat.wordpress.com/2015/11/19/whistleblowers-in-noaa/

      I had linked it previously here:
      https://judithcurry.com/2015/11/18/iatrogenic-climate-policy/#comment-744778

      This was also linked by Goddard linked by omanuel above:
      omanuel | November 22, 2015 at 10:05 pm | Reply

      https://judithcurry.com/2015/11/22/a-buoy-only-sea-surface-temperature-record/#comment-745489

      Goddard is not my favorite (understatement) so I’m always suspicious of his musings.

      Now whether or not this is a fishing exhibition or is a tempest in a teapot remains to be seen. Smith apparently believes he has something. Personally I don’t much like this clock and dagger stuff. I didn’t like it when Grijalva was spouting his nonsense (that did appear to be a tempest without a teapot):

      https://judithcurry.com/2015/08/18/industry-funding-witch-hunts/#more-19659

      “So I guess Grijalva witch hunts against climate skeptics are ok, but there shouldn’t be witch hunts against consensus scientists?”

      So I wouldn’t like it anymore if Smith comes up short. But that’s politics. I doubt the press will be as kind to Smith, should he fall short, as they were about Grijalva (silent). As Joshua might say, ‘sameole sameole’.

      So I guess there is witch hunts against concensus scientists after all.
      Oh my!

      • Government Scientists Continue To Ramp Up Their Criminal Activity Ahead Of Paris — By SETH BORENSTEIN Nov. 22, 2015 10:51 AM EST

        The same criminals who claim (when convenient) that surface temperatures are more accurate, also feel the need to massively adjust them (when convenient) – because of their gross inaccuracies.

      • Ordvic

        I think the graph came from Steve Goddard who was referencing Seth.

        http://realclimatescience.com/2015/11/government-scientists-continue-to-ramp-up-their-criminal-activity-ahead-of-paris/

        if NASA or NOAA have been fiddling the temperature figures Steve has the scoop of the century. He needs to demonstrate his claims in peer reviewed journals rather than in his own blog. Alternatively putting them directly to the agencies involved and getting their responses might help to focus his mind

        tonyb

      • Hardly a witch hunt. When whistleblowers bring to the attention of a Congressional oversight committee allegations that there has been undue political influence and a failure to follow a federal agency’s methods and procedures, the committee is supposed to investigate. It’s their job. Period.

        Here is another one:

        http://www.foxnews.com/politics/2015/11/23/emails-show-dod-analysts-told-to-cut-it-out-on-isis-warnings-ig-probe-expands.html

        If you don’t like Fox, here is the story softly portrayed by the Obama media:

        http://www.nytimes.com/2015/11/22/us/politics/military-reviews-us-response-to-isis-rise.html?_r=0

        Obama’s political toadies are running agencies not for the people, but to please their narcissist boss. IRS, Justice, State , NOAA, HUD, Homeland, all of them. And when they get found out they stonewall. I know people in Centcom. Those who don’t toe the party line don’t get promoted, or they are just shuffled out of the service. The military command structure has been systematically purged of anyone who won’t enthusiastically carry out the Obama agenda.

        Of course, the Obama goons would manipulate the NOAA science. You can’t trust any of the crap coming from an administration that will make up that “video caused Benghazi” lie. They insist that we believe that Islam is a religion of peace, but a freaking video causes the faithful Muslims to go into a spontaneous murderous rage.

        Shhh! Don’t say it’s radical Islam. Anyway, they are not a mortal threat to the U.S. Which means they can’t kill all of us. So don’t worry, be happy. Delusional and dangerous ideologues are in charge of our country.

      • Another good(ard) chart:

      • Record Crushing Fraud From NOAA And NASA Ahead Of Paris

        Posted on November 19, 2015 by tonyheller

        Gavin and Tom delivered their fraud right on schedule ahead of Paris, just as I predicted they would. They claim that October had the highest temperature anomaly ever recorded for any month.

      • 1. It wasn’t a bombshell. The adjustment amounted to a small fraction of a degree. In the current El Nino the temperature has already gone up five times as much in just months which has a major effect on the trend. Not only that but Karl’s Fig. 2B shows that if you don’t adjust for observing system changes at all you end up with even more warming in the past century, so net adjustments are still negative.
        2. It was first submitted in December 2014, and took 6 months to publish. That is not rushed, and may even be considered slow. If “rushed” is all they are saying now, it is weak tea.

      • JimD:

        “It wasn’t a bombshell. The adjustment amounted to a small fraction of a degree.”

        Well quite. That was all that was “needed”.

        The point missed by those who are trying to support the Karl et al paper, or by Stephen Mosher attacking those who are critical of it, is very much the one you’ve just made. But that was not how it was promoted.

        The timing of it, the way it was released to the press, has all the hallmarks of politicisation of science and not of objective analysis of data. I don’t think there would have been the push back if the results had been released in a timely fashion and announced as a “very minor correction” which shows a slight statistically significant warming trend in relatively sparsely covered sea surface temperatures over the duration of the “pause” in increase in GMT.

        Given the experience of climategate, it’s not unreasonable to be suspicious of the way this study was done, it’s conclusions, the way it was promoted by the lead author, and the timing.

      • agnostic, you can probably see where the concern is. The skeptics/whistleblowers, whatever you want to call them, have now had months to come up with a scientific rebuttal and have access to all the same data, and have drawn a blank. Given that we have a scientific study with proper results published, and no sign of a rebuttal, there is no reason to go around requesting the emails of the authors. As the AMS letter says, this has a chilling effect. In the future, a scientist that has results that they might think that the chair of a committee doesn’t like, may hold on to them rather than publish, and we are back to the muzzling of government scientists again. It won’t stop the science in the US because of the academic freedom that universities have to publish, so there are ways around it, but having to work around such congressional threats is just another obstacle to government science.

      • You are engaging in flimflammery, yimmy. The Congressional oversight committee is doing oversight. The NOAA is a creation of the Congress. The Congress, whose members are elected by the people every two years, appropriates funds to support NOAA’s operations. The oversight committee has the responsibility to make sure the people’s money is well spent. Some scientists working at NOAA have alleged that in the case of Karl et al. the work was influenced by politics and that established scientific procedures and methods were not followed. The oversight committee is supposed to investigate that, yimmy. You little lefties love whistleblowers, when the pressure is being put on Republicans. It’s your turn in the hotseat. Stop whining.

      • The seat is not hot; it’s not even lukewarm. It’s kool. Bring it, losers.

      • It’s just kool? It’s not like it’s really cold like a toilet seat in the morning or anything? It’s sooo not like we’re losers lol

  36. Re ARGO, see this recent post
    https://judithcurry.com/2015/07/09/recent-hiatus-caused-by-decadal-shift-in-indo-pacific-heating-2/

    Note, they cited a 0.01C/decade OISST trend since 2003

    • Except for the fact that OISST is not the same as ARGO. Nieves has the same trend for ARGO 0-10 meters as the trend for ERSST4. So if you want to use ARGO to show the wrongness of SST that must be OISST. Or more precice: OISST monthly. Not OISST daily which is in line with ERSST4. Ref earlier in this thread.

    • Shucks, article protected. I trust Josh though.
      Hi Dr. Curry. Good luck next week. I wish you focused thoughts and articulated words of truth. We need a Champion, a warrior princess like Diana, the Huntress. Sorry for the drama, that’s me. You have demonstrated your courage in the face of adversity. Again, good luck and best wishes for more grants.

      Regards,
      JerryG∞

  37. Thanks for the article. Appreciated.

  38. Claims for new observed temperature data sets and increased trends should be compared with existing observed data sets looking for statistically significance differences. The effects of autocorrelatio should be included in the comparison and can be estimated using simulations of an appropriate ARMA model. That ARMA model is usually (2,0) or (1,1). These calculations are straight forward and should be included in detail with any claims – and particularly in the den where the uncertainty monster lives.

  39. Odd, you can justify Karl et al. adjustments to surface temperature by showing the method they selected to begin with was inferior. Adjustments aren’t required, they are just required by their choice of methodology.

    You can do the same thing with ersstv4, but that doesn’t indicated that Karls method is superior to hadcru or oisst, just that it is okay. However with all this focus on buoys, the grand leap circa 1945 is huge in Karl et al. indicating that they will need to do some more adjustments on their adjustments which must be great for job security.

    • A buoy-only SST record is decent from 1995 onward (though even then its spatial coverage isn’t perfect). To create a SST record prior to 1995 or so you need to use ship-based instruments. Karl et al are attempting to construct a SST series from 1880 to present, something that necessarily requires weaving together multiple data sources as they change over time.

      The test we performed is useful, however, as much of the recent controversy has focused on the effect of adjustments during the modern era, when we have good data from multiple instrument types to work with.

    • captdallas2 0.8 +/- 0.3 says, November 23, 2015 at 10:46 am:

      Odd, you can justify Karl et al. adjustments to surface temperature by showing the method they selected to begin with was inferior. Adjustments aren’t required, they are just required by their choice of methodology.

      That’s a very good point! In fact, it renders this whole endeavour rather pointless.

      Karl et al. claim to get a higher trend post 1998 because there are more and more buoys in the data, and these are cooler than the ship intake data, and so they somehow bias the total down. Hence the claimed need for upward adjustments.

      But like you said, according to Cowtan and Hausfather here, that apparently wasn’t necessary in the first place, because the buoy trend is – allegedly – much higher than the ship intake trend (and the satellite trend, and the night marine air temp trend, mind you) already to begin with.

      Rather hilarious!

      • yeah, pretty funny, but this is the circle jerk called post normal science. Every new adjustment just teases the uncertainty margins in the politically correct direction. If it were normal science we would be seeing averages of all products, uncertainty margins and less hype pre-conference. You might even see a scientist in the media say something like “I dunno” or “Good news, it isn’t as bad as we thought!” You just can lose that urgency momentum even with 40 million refugees wandering around thanks to bio-fuel demand increases in basic food stuffs.

        .

      • From Karl’s supplementary material (page 2 of 6):

        The factor that contributed the largest change in SST trends over this period was continuing to make corrections to ship data after 1941.These corrections are based on information derived from night marine air temperature. This correction cools the ship data a bit more in 1998-2000 than it does in the later years, which thereby adds to the warming trend. To evaluate the robustness of this correction, trends of the corrected and uncorrected ship data were compared to co-located buoy data without the offset added. As the buoy data did not include the offset the buoy data are independent of the ship data. The trend of uncorrected ship minus buoy data was -0.066°C dec-1 over the period 2000-2014, while the trend in corrected ship minus buoy data was -0.002°C dec-1.

        As I understand Karl’s description, uncorrected buoy data (that is, prior to adding 0.12 C?) was used to test of the validity of their decision to extend the mid-century ship adjustments into more recent years.

        If extending pre-1941 ship data adjustments into recent decades by itself produced nearly perfect agreement (-0.002 C/decade) with the uncorrected buoy data trend line — why did they need to add an additional adjustment to the buoy data afterwards?

        I can only assume I misunderstood Karl’s description and would welcome a clarification on this point.

  40. I am not a scientist but I have been following the climate debate for over a year and my simple mind has reached some conclusions.

    Since the end of the Holocene optimum, the Eart is in a definite cooling trend, with spikes of occasional warming, each successive warming less than the previous warnings. Arguments concerning such short time frames as this and tenths of degrees are not useful in determining where we are heading.

    The CO2 theory does not explain past warmings, periods of glaciations, or non ice ages.

    Like everyone else, I learned about the Greenhouse theory in school. But when I read about the Robert Wood experiment in 1909, I’m stuck for an explanation.

    It seems to me that the temperature and climate on Earth is dependent on the amount of energy received by the sun and that weather is caused by unequal heating of the Eath’s surface.

    The only theory I have seen that can explain the Earth’s climate history is Svenmarks cloud theory. More clouds results in cooler temps, less clouds means warmer temps.

    Perhaps that is too simple to consider.

    • 100% CORRRECT.

      The news is the cooling trend since the Holocene Optimum is still intact.

    • +10. Clouds are right up there as a known unknown factor causing the negative feedbacks that has kept climate since the LIA within remarkably stable bounds. Its my guess that sudden climate shifts are caused by various combinations of forcings occurring at the same time but that individual forcings have only marginal impacts on their own.

  41. The only difference of any significance between ERSST.v4 and NOAA OISST.v2 is a distinct 0.06K uptick in the former relative to the latter in the summer of 2006. Nothing before, nothing after. You only find this sudden divergence in the ERSST.v4 dataset. Not in ERSST.v3b. Not in HadSST2. Not in HadISST1.

    Is Karl et al. discussing this single step anywhere? Even mentioning it? Why exactly in the summer of 2006? Why the entire lift in that one step alone? Why no gradual adjustments? Why nothing before? Why nothing after?

  42. All of these adjustments which have been going on are in word BS.

  43. ” During the past few decades the number of automated SST measurement buoys has expanded rapidly from effectively zero before 1980 to over 70 percent of all SST measurements today as shown in the figure below. ”

    Alas, the “figure below” is not found in Kennedy et al. (2011) and the claim seems credible only in light of the fact that ship-board observations are made only four times daily at the synoptic GMT times, whereas buoy observations are typically hourly. While the prospect of aliasing diurnal variations into much lower frequencies thus is virtually nil with buoy data, they nevertheless are largely unavailable for fixed locations.

    The ploy of coarsely gridding ARGO and shipboard data in order to “localize” daily averages no doubt introduces additional bias and variability not considered by the authors. Many grid boxes simply do not contain enough data to obtain reliable climatological summaries, let alone credible time-series. Thus the determination of global or hemispheric SST averages via other than satellite observations remains unconvincing.

  44. Roy Spencer did a comparison a few years back between the “TRMM satellite Microwave Imager (TMI)” SSTa and the equivalent “AMSR-E” on the one hand and the in situ records of “HadSST2” and “ERSST.v3b” on the other (~40N-~40S), and found no disagreements worth noting:

    As should be well-known, the HadSST2 and the ERSST.v3b agree pretty well with OISST.v2. There is absolutely no sudden step up in mid 2006 to be seen in any of these datasets. Rather, they all very much seem to corroborate (validate) each other. Here’s OISST.v2:

    Spencer also checked his TMI results against the global moored buoy network SSTa measurements (40N-40S):

    Not even a hint of any step change occurring during 2006 …

    • Nice reminder. The temptation to “fix” what is not broken nevertheless remains irresistible to most warmistas.

    • David Springer

      Cowtan and Hausfather fail to respond again.

      The 2006 step change in ERSSTv4 is a smoking gun. One single adjustment unique to ERSSTv4 beginning in 2006, that no one is talking about, is the whole enchilada.

      • Cowtan and Hausfather have done left the building. Mr. okulear needs to make sure the investigators for the people on the House Science oversight committee are aware of suspect uptick.

      • Hate to burst a good bubble (or put out a smoking gun?), but the difference between ERSST v4 and v3 doesn’t show a step-change in 2006 in the difference series:

      • Is there a little blue pill for smoking gun dysfunction?

      • That is interesting, Mr. Zeke. Please explain for the unwashed masses how thoroughly that contradicts:

        “As should be well-known, the HadSST2 and the ERSST.v3b agree pretty well with OISST.v2. There is absolutely no sudden step up in mid 2006 to be seen in any of these datasets. Rather, they all very much seem to corroborate (validate) each other.”

      • Also, are you saying there is no step change in ERSSTv4 in 2006, Mr. Zeke?

      • David Springer

        The step change in 2006 can be easily eyeballed by overlaying OISST v2 and ERSSTv4

        https://judithcurry.com/2015/11/22/a-buoy-only-sea-surface-temperature-record/#comment-745645

        Offhand I’d say ARGO data was changed because that’s the approximate start date of all 2000 floats going online and the new data was stitched in at that point in yet another trick to hide the decline. These boys just never learn.

      • Zeke here of course knows perfectly well that I didn’t discuss an apparent 2006 step change between ERSST.v4 and ERSST.v3 at all; rather I discussed an apparent step change between ERSST.v4 and Reynolds (NOAA) OISST.v2.

        Here’s the difference between ERSST.v3b and OISST.v2 (1995-2015), with trendline:

        Here’s the difference between ERSST.v4 and OISST.v2 (1995-2015), with trendline:

        What happens if we adjust ERSST.v4 down by 0.0615K en bloc from June 2006? The following:

      • David Springer

        Okulaer shoots! He scores!!!

        The crowd goes wild!!!!!!!!

  45. This is great. Half the warmer scientists are burning budget trying to explain the hiatus, while the other half (Zeke & Co) are busy explaining that it does not actually exist, so presumeably cannot be explained. Uncertainty personified. The proper conclusion would seem to be that we do not know what is going on, but that would be inconvenient, to say the least.

  46. These adjustments should come as no surprise to anyone since they we’re planned and announced long before the “hiatus” was even an issue, in the 2007 paper by Smith, Reynolds etc. Improvements to NOAA’s Historical Merged Land–Ocean Surface Temperature Analysis (1880–2006)

    “Because ships tend to be biased warm relative to buoys and because of the increase in the number of buoys and the decrease in the number of ships, the merged in situ data without bias adjustment can have a
    cool bias relative to data with no ship–buoy bias….At present, methods for removing the ship–buoy bias are being developed and tested.”

    (emphasis added)

    • “Because ships tend to be biased warm relative to buoys and because of the increase in the number of buoys and the decrease in the number of ships, the merged in situ data without bias adjustment can have a cool bias relative to data with no ship–buoy bias….”

      Only if the buoy ANOMALY TREND is cooling relative to the ship ANOMALY TREND. It shouldn’t matter if the buoy temps are cooler in absolute terms than the ship temps. What we’re looking at here are the anomalies, particularly the ANOMALY TREND.

      • This is equivalent – only inversely so – to the well-known argument about whether or not removing ‘cold’ weather stations and increasing the amount of ‘hot’ weather stations instead has had any effect on the total long term trend. The Hausfathers and the Moshers of this debate have been adamant the whole time – and rightfully so – that it doesn’t matter, as long as the ANOMALY TRENDS are the same. We’re talking anomalies, not absolute values here …

      • But the effect upon gridded data amalgams of gradually introducing very different sensor readings is precisely upon the anomaly trend. This is analogous to the gradual introduction (or development) of UHI-afflicted stations in estimating the “regional” average, instead of maintaining a uniform set of non-urban stations throughout the period of record. That’s why fixed buoy data correlate much more strongly to satellite observations.

    • That’s a hoot, Slanger And the adjustment was so simple in concept and so necessary that it took them eight years to get around to doing it, just before the last chance to save the earth from burning up Big Partee in Paree.

      • You see the significance, Don. “Rushed” it wasn’t. The contrary, in fact.

      • Little yimmy is tossing out the ole stinky “rushed” red herring again. If the adjustments are freaking obvious and necessary, why did it take them 8 freaking years to do it? The answer is in the emails and in the honest testimony of the NOAA scientists.

        The refusal to co-operate indicates, when all other stonewalling tactics fail, we are very likely to see hard drives getting struck by lightning, eaten by dogs and numerous invocations of the 5th amendment.

        If this climate science stuff is so important to our kids’ survival, why do they want to hide stuff from us? It messes up their credibility. Is it because if the truth comes out, it will be much worse for them? Help us out, yimmy? What does huffpo say?

      • They are not hiding anything. The data is there. People can test what they said in the paper themselves. Some already have, as we see here. It’s the paper, not the people, that this is centered on, but Smith lost focus somewhere along the line and is just thrashing out.

      • It is a big lie that they are not hiding anything, little yimmy. They are hiding the emails and the other info the Congressional committee has asked for. The committee wants to know if the whistleblowers’ charges are true. If the NOAA alarmist sighentists were influenced by political motives to get rid of the pause and if they followed the established NOAA scientific procedures and methods. They can’t keep that stuff hidden, yimmy. It will come out.

      • If the screws are turned, they will fold. In any case ain’t no virgins involved.

      • If they want to check if the whistleblowers are lying, they can check the whistleblowers emails first. Then they can separate facts from idle speculation.

      • You are still clueless, or pretending to be clueless. You don’t get to tell the Congressional investigators what to look at first, yimster.

  47. richardswarthout

    Appears that Mosher got his tail on fire reading blogs and then came here to yell at the denizens. Perhaps the stress of living in California is too great to bear. He should take my previous advice and return to the good life in Michigan.

    Richard

    • Richard

      On occasion Mosh needs to read more and insult less and take a nice calming cup of cocoa early evening to reduce his stress levels. A couple of chocolate digestive biscuits at the same time would then set him up for his forays into the deniersphere where he can face the awful prospect of people asking questions.

      What’s with this Michigan comment?

      Tonyb

      • questions from honest people are fine.

        sadly a bunch of people are fabricating crap about Karl’s science.

        That maybe funny to you.

      • Mosher, “sadly a bunch of people are fabricating crap about Karl’s science.”

        You referring to the NOAA Whistleblowers Mosh?

      • Mosh

        There is no sceptic on this blog who defends scientists more than i do against accusations of fraud. You know that. I also do it elsewhere.

        I was over at WUWT commenting on the space heater/CET alarm way before you were and at Bishop Hill. I get flack for that. Suggesting you need to relax with a cup of cocoa is hardly in the same league as being accused of ‘sliming the records’ is it?

        As Richard suggests, perhaps number crunching isn’t always satisfying?

        I bear you no ill will. I am sorry this does not seem to be reciprocated.

        tonyb

    • richardswarthout

      Tony

      Just to fill in a blank – My suggestion to Mosher, to teach at my granddaughters high school had a twofold purpose. First, for him, guiding high school girls through the great books could only be, IMO, much more gratifying than crunching numbers at Berkley. Secondly, for me, I wanted my granddaughter to gain the experience of being taught the great books from a master.

      Richard

    • David Springer,

      If he believes he can debate several sceptics at once, wouldn’t that make him a mass debater?

      Sorry. An attack of CWP (Compulsive Word Play).

      Cheers.

  48. “The new NOAA record also increases temperatures in recent years, resulting in a record where the period subsequent to 1998 has a trend identical to the period from 1950-1997 (and giving rise to the common claim that the paper was “busting” the recent slowdown in warming).”

    Zeke, eyeballing the graphic with the new and old adjustments makes me think that perhaps both sides of this issue are overreacting. We agreed, or at least I thought we agreed, that 15 year or even longer trends of this magnitude in the recent observed temperature record are difficult to impossible to show statistical significance pauses and that is due to the natural variation in observed temperatures. That graphic shows a bending downward trend from a straight line from the 1970s to current time. (I think we probably also can agree that it is unlikely that the temperature trends should be linear). The advent of large increases in the atmospheric CO2 levels would call for looking at that period and not a 1950 starting time.

    I am more motivated than ever here to get the Karl data and look for statistically significant differences with other temperature data sets and with the CMIP5 historical model runs.

    Busting is a cute term for what you do when you find your teenager in the “cookie jar” but probably is going to provoke some reactions not compatible with a reasoning scientific analysis.

    • Ken,

      My current inclination is that the existence or non-existence of a hiatus is not important–basically because there is always the next year’s uncertainty, the next five years’ uncertainty, etc. Lacking an accepted validated conceptual physical model statistics will come up short. (To be clear, this should not be taken as a reason for no action on the policy front. That is an entirely different matter.) Looking at the hiatus is important from the perspective that it may provide insights and ultimately detail on the [yet unspecified] mechanisms at play.

      • mwgrant, my point on the hiatus is to focus attention on the noise levels and natural variations making it difficult to determine trends or lack thereof of 15 and even more years. In fact we should not ignore the fact that GHG warming and its effects on global temperature trends have covered only about 40 odd tears of the observed record and natural variations and noise make even that period of comparison of models with observations a formidable task. I have been using the 1970s to near present time to compare model and observed trends. It is there were increased degrees of freedom are your friend in finding statistically significant differences in these trends that we see difference between models and observed global temperature series.

        It is not only trends where significant differences exist between models (CMIP5) and observed series but in white noise and autocorrelations and trend ratios between the hemispheres.

        Obviously predictions for the future will require validating climate models with observed data. In fact, given the problems with temperature reconstructions going back in time and validating proper temperature proxies, we will probably have to depend on validated climate models to tell us about the millennial past temperatures.

      • ken,

        In fact we should not ignore the fact that GHG warming and its effects on global temperature trends have covered only about 40 odd tears of the observed record and natural variations and noise make even that period of comparison of models with observations a formidable task.

        A couple of thoughts here. First, looking at the larger time period this makes a lot of sense–particular after your clear statement. Also noting the model/observation differences in white noise and correlations as well as the hemispheric trend ratios is basically indicating that the models do not have the same statistics as the observations and that is a potentially very big red flag (IMO). It is also interesting because it suggests [to me anyway] an approach to initially [semi-]quantitatively evaluating/screening models without having details of mechanisms in hand–depth and breadth look at statistical consistencies and inconsistencies between a model and observation.

        In fact, given the problems with temperature reconstructions going back in time and validating proper temperature proxies, we will probably have to depend on validated climate models to tell us about the millennial past temperatures.

        You are a bundle of optimism…probably real fun at all the climate parties. But it is what it is. Keep on truckin’.

      • My current inclination is that the existence or non-existence of a hiatus is not important–basically because…

        Because it’s almost completely nearly dead!

      • Well, there was a pause from 2001 through 2013 ( actual negative trend ).

        At the same time, the trend and correlation with RF is also consistent.

  49. I want to thank Zeke Hausfather and Kevin Cowtan for their essay, and for their responses to comments.

    • But I have a question: If there has been no pause, what happens to all of the explanations of how the pause came about? Science Magazine went so far as to publish a paper showing how the pause was predictable from the science known before it occurred.

      • Matthew

        The Met office published three papers and hosted a conference on the subject of the pause.

        At a climate conference held in exeter two years ago the panel of IPCC reviewers including Thomas stocker were talking freely about the pause.

        Tonyb

      • They should do the right and honorable thing. They should deny any conferences ever took place and that any papers or discussion were simply a figment of the imagination of anyone who thought they saw what they saw or heard what they heard. Wiping of hands. Poof, gone.

      • tonyb: At a climate conference held in exeter two years ago the panel of IPCC reviewers including Thomas stocker were talking freely about the pause.

        I am hoping for a response from Housefather and Cowtan.

    • Of course they talked about it. That was before I almost completely killed it and then Karl erased it.

  50. How much of the new analysis is based on the same sort of infilling and krig-ing that the BEST record is based on? What is the expected error in the krig-ing involved? As I understand it, the BEST supplemental data claims a minimal error for any krigged value of 0.49ºC — the difference between two points krigged even if they are in exactly the same location.

    Any of that going on here?

  51. Ah, reading more closely we see that the results depend on the fine details, where they reject or include data, how the averages are determined, how the averages are combined, ad infinitum. I would like to see the whole paper and see if differing methods of calculation were attempted and if the results from everything tried, every method tested, every result. I would like to see the results of every reasonable approach.

    The question is: Are we seeing the results of only the combination of methods that “Came out right”?

    • The NOAA whistleblowers saw all that, Kip. And they say the adjustments are tainted by undue ppolitical influence and a failure to follow established NOAA scientific procedures and methods. That is what is being investigated by the people’s elected representatives. The Obama political flunkies who run NOAA and their minions, Karl et al., respond that they don’t have to show us no stinking emails.

      • 1. We are being told that this is an honest result of objective science.

        2. If this is the result of an administration request or pressure this is not the honest result of objective science.

        3. If this is the honest result of objective science it would be claimed to support Obama’s policies.

        4. If this is the result of an administration request or pressure this is a stupid political stunt and beneath contempt.

        Now it is pretty clear that if there was pressure it would undermine any political advantage from Karl’s adjustment and make it look like an unethical political stunt.. Karl has no right to avoid disclosing his administration interactions. If they were harmless, Smith has egg on his face. But the delay and resistance makes it look like Smith will be eating his eggs with some of the Administration’s bacon.

      • Even NOAA has members of the Tea Party working for them. Most likely these are disgruntled marginally competent partisan hacks who had their careers switched over to the sidetrack down in the basement when Obama won reelection, and now they’re looking for their 15 minutes of fame. Has that stench. No scientist would do this. They’re reaching for the zenith of their scientific careers.

      • Dr. Judith Curry:

        “I’ve heard enough behind the scenes (including discussions with NOAA employees) that I am siding with Rep. Smith on this one.

        The politicization of climate science has gotten extreme. I don’t know where to start in trying to ameliorate this situation, but Congressional oversight and investigation into what is going on in government labs does not seem inappropriate under these circumstances.”

        I am guessing that she was not talking to janitors at NOAA, but to fellow climate scientists.

      • Then there is the alternate crackpot conspiracy theory espoused by the putz, above.

      • Based upon peer-reviewed science, at least 3% of them would talk to her.

      • OMG! More conspiratorial anti-Obama Tea Party whistleblowers:

        http://www.thedailybeast.com/articles/2015/11/23/analysts-accuse-centcom-of-covering-up-cooked-isis-intelligence.html

        These clowns under investigation decided to delete the incriminating emails. A big shot Maj. General Obama political sycophant, in this case. Another Lois Lerner. He will obviously be promoted to Chief of Staff. The most transparent administration in history. You can see right through them.

      • JCH | November 23, 2015 at 7:59 pm |

        No scientist would do this. They’re reaching for the zenith of their scientific careers.

        Go read the climategate files. That statement is simply wrong.

        However, I’m not sure what the problem is here. Given an adjustment that is at least somewhat defensible they shouldn’t be hiding their emails.

        I tend to believe they are covering up some other hanky-panky or are just unbelievably arrogant and stupid.

      • JCH,

        “Most likely these are disgruntled marginally competent partisan hacks who had their careers switched over to the sidetrack down in the basement when Obama won reelection, and now they’re looking for their 15 minutes of fame.”

        Of the many whistle-blowers in the US gov during the past generation or two, can you name a few who meet your description? “Hacks” “looking for fame”.

        In fact govt whistle-bowers often provide valuable information to the public at great cost to themselves. As a group they are brave men and women who have contributed much of what we know about our governments’ actions.

        We don’t know anything about the alleged NOAA whistle-blowers, but history provides little support for your theory — and suggests an open mind on the subject would be more appropriate.

        https://en.wikipedia.org/wiki/List_of_whistleblowers

      • Whistleblower scientists blow their whistles in journals.

      • JCH,

        “Whistleblower scientists blow their whistles in journals.”

        Chairman Smith cites a whistleblower’s information regarding possible violations of NOAA’s internal policies. In America such things are investigated by the Executive or Legislative branches of government, not outsourced to non-governmental organizations.

        Also, are such violations even within the usual field of interest of “journals”? Some cites of prior examples would be interesting support for your assertion.

        http://cdn.arstechnica.net/wp-content/uploads/2015/11/lamar_smith_noaa_comm_letter_nov18.pdf

      • Whistleblower scientists blow their whistles in journals.

        They’re tooting their own horns, not blowing whistles.

      • I hope our little alarmist drone friend is around when the Karl mob are perp-walked in front of the committee and meekly read off the 5th, from scraps of papers shoved in front of them by their scheister progressive lawyers. Oh, the humanity!

    • (Trying to reply to Kip H again)

      Isn’t all the data used (and not used) available to all?

      • All the data might be available, but we want to know why they did what they did with the data. Why did they decide to take the fork in the road that led to the hyped up pause killing BS, rather than going in some other direction, or standing still? Is it because they were following the procedures and methods of NOAA, or was it done to come up with the prescribed politically correct answer?

      • “All the data might be available, but we want to know why they did what they did with the data. Why did they decide to take the fork in the road that led to the hyped up pause killing BS, rather than going in some other direction, or standing still? ”

        Ask the people who did the original work.. long .. like 2007

      • Why didn’t somebody make the adjustments back in 2007? Or 08, 09, 10, 11,12,13,14?

      • Don M wonders:

        “Why didn’t somebody make the adjustments back in 2007? Or 08, 09, 10, 11,12,13,14?2

        Because they wanted to do the adjustment as good as possible. Worked on in for more than 5 years.

        That means “rushing” for Lamar Lysenko and his compadres.

      • The stinky “rushing” red herring again. You kids are incorrigible. The dissenting scientist whistleblowers at NOAA say the Karl crowd did incomplete work. Did not follow the scientific procedures and methods established in NOAA. They had years and years, but they still cut corners. They are not getting paid by the taxpayers to waste years and years cutting corners and turning out product that is not the result of sound scientific procedures and methods. And now they stonewalling. Is that clear now, joker?

      • “Rushed” suddenly became a red herring for the Don.

        Well, that was all the “whistleblower” could come up with. Months after the paper was submitted for publication. A whistleblower herring then.

      • Little ehack tosses the well-worn and very stinky “rushed” red herring, again. What a putz. The NOAA scientist whistleblowers have accused Karl et al of being unduly influenced by politics and not following established NOAA scientific procedures and methods. Stop the lying that’s it about rushing. You clowns appear not to understand that you are continually diminishing the credibility of the AGW case by lying and stonewalling. If you really believed that Karl et al had done nothing suspect, you would be encouraging them to co-operate with the investigation and get it over with.

      • richardswarthout

        Ehak

        “Months after the paper was submitted for publication. A whistleblower herring then.”

        Darn. Somebody screwed up; every government employee must be made aware that the whistleblower statute of limitations is measured months.

        Richard

  52. Looking at our ATSR-based obs4MIPS dataset, the global mean SST trend (not including sea ice areas) over 1998 to 2012 [1] is 0.085°C, which is 0.06°C/decade [2].

    Karl et al.’s “new” value of SST change over this interval therefore fits pretty well with our independent [3] satellite data. These data featured in IPCC AR5. (They also agree well with the Hadley Centre in situ ensemble in the same figure.)

    So, how new Karl et al.’s result is depends on what data you previously paid attention to.

    And I would still describe 0.06°C/decade as a slowdown compared to the 1980s and 1990s.

    • Will Judith highlight this in the same way as she highlights OISSTv2 monthly (and ignores OISSTv2 daily)?

      Don’t think so. Climate Politization etc.

      • I have emailed Dick Reynolds (father of OI) for his take on the different data sets and the changes.

      • Excellent Judith.

        In the meantime Geert has shown himself as a really fast mover and put Oiv2 daily on Climate Explorer. Here we go:

        Trend 1998 to now for Ersst4: 0.12/dec. For OISSTv2 daily: 0.15.

      • Apparently daily took the little blue pill.

      • Kinda interesting.

        The Blob is gone and it looks like this El Nino is getting a little long in the tooth:

      • November 23rd this year:

        November 24th last year:

        SOI is way up. The El Nino is forecast to peak November-December… old news.

      • The El Nino is forecast to peak November-December

        Right. And that would make it an early party-pooper:

      • richardswarthout

        Ehak

        Your question and your answer to your question is uninteresting. And your injection into the discussion of daily v monthly reveals much about you. Dr Curry, by enlisting the advice of Dick Reynolds, is being reasonable, while you cannot wait for reason.

        Richard

      • Right. And that would make it an early party-pooper:

        Although, it could be a head fake.

      • TE – you’re lost. GMST during ENSO neutral was sky high. The blob was just an excuse to not face the music – the cyclist didn’t want to believe a cycle could fire off 20 years ahead of their prediction – 20 years. All their chips are on the other cycle, and I am going to win that one too.

      • I was talking about the ENSO – you were talking about T.

        Still not clear that anyone really understands the relationship between the two.

      • Haha, anyone noticed how the ‘new’ HR OI.v2 analysis (claimed to be from daily AVHRR readings) at KNMI Climate Explorer starts one month before the original analysis AND already has the November and December 2015 values in!!!

        This is modelled data. Complete fabrication. Utter BS. For crying out loud, look at that graph! It’s insane!

        Who buys this stuff? Everything’s apparently totally normal from 1981 to about 2000/01. And then all of a sudden it just takes off, aiming for the sky. The 2009/10 Niño peak now 0.15 degrees higher than the 1997/98 peak, when in the original version it’s almost 0.05 below? C’mon, people.

        These people have one aim and one aim only: To bust the “Pause”.

        Look at the satellite microwave imagers, TMI and AMSR-E:

        They don’t agree AT ALL with what’s claimed here to be AVHRR, but which is clearly just modelled data.

        This is just too stupid to be real.

      • What is going on here!?

        Here’s the “Daily AVHRR (?) OI.v2” vs. the “Monthly OI.v2”:

        First of all, notice how the monthly (yellow) is ‘always’ a month ahead of the daily (blue).

        But more interesting (I would say stunning) is the fact that the two more or less agree from 1982 to 2001/02 AND from 2007 till present, but for some unexplicable reason, the daily rises 0.13-0.135 degrees relative to the monthly between 2002 and 2007, across a period of less than 5 years!

      • Consider the possibility that there is an error in Climate explorer okulaer. Like setting a wrong start month.

        And if you had paid attention you should have noticed this link:

        https://www.climate.gov/news-features/blogs/enso/exactly-same-completely-different-why-we-have-so-many-different-ways

        The monthly version does not have the ship-buoy bias adjustment. Daily has.

        The match in the first part and the last part in monthly vs daily is no big surprise. In the first part there are predominantly ship-intake measurements. In the late part mostly buoy measurements. That also means that in the coming years with mostly buoy measurements this bias adjustment will have small or negligible effect for the trend.

        Fact is: The adjusted series corresponds better to buoy only series than non-adjusted. Even compared to ARGO. Which I find surprising given the fact that the ARGO product were not designed to be a surface temperature replacement.

        Btw: Your “cold station” bias is taken care of in Karl 2015. They brought in more “cold stations”. Result: higher temperature increase. Reason: The “cold stations” have higher temperature increase.

  53. Consider three or more straight lines of constant slope.

    y1 = mt + b1
    y2 = mt + b2
    y3 = mt + b3
    y4 = mt + b4

    A composite record made by combining mixture of measurements y1, y2, y3 and y4 that changes with time can be adjusted to produce almost any slope by changing the offsets/biases b1, b2, b3 and b4 between the four lines. Zeke and Kevin are asserting that they can tell if offset b4 is correct by comparing y4 alone (buoys) versus a composite y1-y2-y3-y4 (ERSST v4 or v3). This appears to be nonsense. The best composite record will be created by optimizing b1, b2, b3, and b4 simultaneously, not by some piecemeal process whose final result depends on the order in which records are introduced into the composite. Lacking that, I like to be able to study all pairwise comparisons between each method of measuring SST. Hopefully, the BEST project is testing strategies for how to do this. And maybe some real scientist (i.e. one that conduct experiments rather than massaging inadequate data at their desk) will spend a year on a boat conducting experiments in different locations with various methods for measuring SSTs. Then maybe we have some solid information about engine intact, canvas buckets, and wooden buckets.

    • “And maybe some real scientist (i.e. one that conduct experiments rather than massaging inadequate data at their desk) will spend a year on a boat conducting experiments in different locations with various methods for measuring SSTs. Then maybe we have some solid information about engine intact, canvas buckets, and wooden buckets.”

      You might want to read the literature on how bucket adjustments were decided.

      • Steve: I read some literature awhile ago and didn’t review it before commenting – perhaps too harshly. As best I could remember, real field work had not been done – corrections are mostly based on models. Here is the background section from a 2002 paper

        “The most important studies of the problem of bias in historic SSTs were performed at the Met Office. Among the published results are Folland et al. (1984), Bottom- ley et al. (1990), and Folland and Parker (1995). These studies show that before 1942, the global-average SST has a cold bias of between 0.1􏰆 and 0.4􏰆C, with respect to the average SST after 1942. The Folland et al. (1984, hereinafter FPK84) bias correction is the simplest of the three. With each new paper, the adjustments were re- fined. In the last two papers, the bias corrections include models of the evaporative cooling of canvas and wooden buckets. The modeled bias was affected by variables such as the marine air temperature and both ship and wind speed. To properly use the models, it was neces- sary to estimate how the relative number of canvas and wooden buckets changed with time, as well as how typ- ical ship speeds and deck heights changed with time. These assumptions lead to a comprehensive model for estimating SST bias.
        In addition, Bottomley et al. (1990) and Parker et al. (1995) made adjustments to nighttime marine air tem- peratures (NMAT). They suggest several similar bias- correction schemes for NMAT prior to 1930 and during World War II. Beginning in the nineteenth century, ships gradually increased in size with time, with a corre- sponding increase in the height of the NMAT. Bottomley et al. (1990) used boundary layer theory to estimate bias induced by these height changes. During World War II nonstandard NMAT measurement practices were used, such as reading the thermometer inside to avoid showing a light on deck (Folland et al. 1984; S. Levitus 2000, personal communication). These practices caused pos- itive bias in NMAT in the early 1940s. In Bottomley et al. (1990), four NMAT bias-correction schemes are de- scribed. All four are all similar and give the same gen- eral corrections, but details and some local corrections differ. The Bottomley et al. (1990) scheme D corrects for changes in deck heights, World War II practices, and also include some local corrections in the nineteenth century. We applied these adjustments to the Compre- hensive Ocean–Atmosphere Data Set (COADS) NMAT to reduce the influence of NMAT bias on our SST bias estimates. In scheme D, some local adjustments for the nineteenth century use the corrected SST anomaly in regions where NMAT is less reliable. For those local adjustments we used the Folland and Parker (1995) SST corrections to adjust NMAT. Thus, for part of the nine- teenth century our bias corrections will be influenced by the Folland and Parker (1995) corrections. As we show later, our results are consistent between the nine teenth and twentieth century, suggesting that the influ- ence is small.
        The final Met Office adjustments to SST included geographic and seasonal variations, which tended to be larger in extratropical latitudes. The Folland and Parker (1995, hereinafter FP95) bias model was well researched and was shown to give reasonable results based on com- parisons of average SST and adjusted NMAT. However, it incorporates many assumptions to compute the bias correction.”

        http://www1.ncdc.noaa.gov/pub/data/cmb/temperature-monitoring/Smith_Reynolds_2002.pdf

        The key reference FP95 (Folland and Parker (1995) is behind a paywall. It’s abstract says:

        “We describe a physically based empirical technique for correcting historical sea surface temperature measurements for time-varying biases. The corrections are based on models of heat and moisture transfers from uninsulated (canvas) and partially insulatcd (wooden) sea temperature buckets exposed on deck. One of the canvas bucket models is tested using measurements on board ship and published wind-tunnel measurements. The method gives geographically and seasonally varying bias corrections through the period 1856 to 1941. The corrections are fairly insensitive to uncertainties such as the size of the bucket or the details of its exposure on deck. A discussion of the history of sea surface temperature observations providcs a background to the procedure. The rcsulting globally and seasonally averaged sea surface temperature corrections increase from 0.11 degC in 1856 to 0.42 degC by 1940. The corrections are compatible with recent measurements made at sea of the errors of canvas buckets. Global and hemispheric time series of corrected sea surface temperature and night marine air temperature data show good agreement: more detailed verifications of the corrections will be reported elsewhere.”

        “ONE of the canvas bucket models was tested”. I don’t see any evidence that canvas, wooden and insulated buckets were compared SIDE-BY-SIDE with engine intake and day and night marine air temperature measurements at various heights above the surface in the vicinity of buoys and compared with simultaneous satellite measurements for the same location. How robust are the measurements to variations in how the sample is collected (including depth and quantity) and measured? Perhaps I misunderstandd what experiments have been done, but what I have read is mostly about models, theory, adjustments, bias; not definitive side-by-side measurement under a variety of conditions.

        Looking back at really old work from a half century ago:

        “More recently James and Fox (1972) have made a comparative study based on a large volume of data comprising matched pairs of bucket and intake temperatures obtained from various ships in the world oceans during 1968-70. Their results showed that, on the average, the intake temperatures were 0.3±1.3°C higher than the bucket temperatures. Similarly, Walden (1966) found that the intake temperatures were also 0.3°C higher than those taken with the bucket.”

        http://appconv.metoffice.gov.uk/hadsst3/references/Tabata_1977.pdf

        The focus of modern work is to homogenize the large amount of data that exists; not understand how reliable that data is and how much uncertainty is introduced by homogenization.

        I see the same problem with the USCRN, which consists of equipment that is significantly different from existing stations. No side-by-side data is being collected using newer equipment and the older Stephenson screens. Instead, the USCRN temperature field will probably be compared with neighboring stations collecting data under the usual poorly controlled conditions.

    • Imagine we are interested in temperature trends. We have a temperature series that has been neasured in °F. We then start to get a few thermometers in °C. This will cause the temperature to appear lower if we do not correct to the same temperature. As time progresses more and more of the thermometers are in °C. The trend will appear to be much less if we do not make a correction.

      It does not matter if we convert °F to °C or vice versa, but we must do one or the other if we are to see the trend

    • franktoo:

      Oceanographers have spent much more than a year in comparing various methods for measuring SSTs, but that hands-on experience is seldom reported in the academic literature, being more the stuff of technical notes and reports about techniques used on scientific cruises.. Alas, there’s seldom any experienced oceanographer involved in assessing the reliability of SST data from ships of opportunity, which is widely recognized in the profession as being almost worthless for scientific purposes. That recognition hasn’t prevented “climate science” from attributing an undeserved reliability to such data or from the exercise of sheer hubris in applying ad hoc statistical massages.

  54. richardswarthout

    The NOAA administrator, by not complying with the subpoena, is not helping Karl; this will not go away. Perhaps she is trying to hide her own complicity.

    Richard

    • Richard,

      +10. Evasion of this kind of information request has been the typical response of activists seeking public policy action to fight climate change, and one reason for their failure to date.

      Civil servants citing protection of the integrity of the scientific progress as a reason for non-compliance with a Congressional subpoena seems odd at best. Science is not a Constitutionally protected activity.

    • The “most transparent administration in U.S. history.”

      This is almost funny. The problem with regressives, like the current president, is they don’t have a clue what ethics and transparency really mean.

      Stealth behind closed door meetings (EPA, Obamacare, etc.), a law passed before it was written (Obamacare), and then the stonewalling. Another refused request for emails, at least this time they aren’t claiming the drive crashed or was recycled or that the emails are on on a scrubbed.private server, they are simply saying “I don’t wanna!”. Childish, simply frickin childish.

      http://www.frontpagemag.com/point/236776/lois-lerners-hard-drive-crashed-so-did-hard-drives-daniel-greenfield
      The administration has been having a hard drive epidemic. The crashes in just the Lerner case are said to be “less than 20” ,

      There is no justification for obstruction or noncompliance.

      There is no reason to take warmunists seriously until they learn what transparency is. If the government and scientists are not going to be transparent to information requests their claims of global warming should be ignored.

  55. Steven Mosher,

    You wrote – “You might want to read the literature on how bucket adjustments were decided.”

    What references would you suggest?

    Would you consider the following extract from a paper published in 2013 relevant?

    “Abstract. Discrepancies between historical sea surface temperature (SST) datasets have been partly ascribed to use of different adjustments to account for variable measurement methods. Until recently, adjustments had only been applied to bucket temperatures from the late 19th and early 20th centuries, with the aim of correcting their supposed coolness relative to engine cooling water intake temperatures. In the UK Met Office Hadley Centre SST 3 dataset (HadSST3), adjustments have been applied over its full duration to observations from buckets, buoys and engine intakes. Here we investigate uncertainties in the accuracy of such adjustments by direct field comparison of historical and modern methods of shipboard SST measurement.”

    The authors have performed field experiments and examined the problems associated with obtaining water temperatures from ships. Maybe you have something more current. Not unsurprisingly, it seems that there are many uncertainties as yet unresolved. It might be that for every assumption, there may be an equal and opposite assumption of equal validity.

    The paper seems reasonable in its methodology, research, field data, and physical component modelling. The conclusions and recommendations likewise appear reasonable.

    Would you support the authors, or dismiss it out of hand? What would you take issue with? I hope you find it useful.

    Cheers.

  56. Judith, Zeke and Kevin: The Karl et al paper didn’t conclusively demonstrate that a hiatus in warming hasn’t taken place. There is still plenty of evidence that some sort of a hiatus has occurred: OISST v2, RSS, the shallowest ARGO measurements, the southern hemisphere.

    What the Karl et al paper proves – to me – is that climate science doesn’t really know if it has been warming or not over the putative hiatus period. You can massage the data using different strategies and come up with different answers. Our COMPOSITE measurements of GMST are not robust: With all of the offsets needed to combine records of different types, changes on the order of 0.05-0.1 degC/decade are within the envelop of uncertainty. You can cherry-pick the composite GMST that suits your personal bias; but you probably can’t prove that someone else’s choice is inferior. And that weakens my faith that warming during the 20th century was 0.8 degC, not 0.4 degC or 1.2 degC.

    On the other hand, one can look at homogeneous records and make more definitive statements about what they say. There has be a hiatus in warming in the lower troposphere.

    • Hiatus in the lower troposphere? Well. Radiosondes perhaps?

      Not quite. If you want hiatus, go for the 1978 – 1998 hiatus. 20 years.

      A hiatus in the lower troposphere should show up in water vapor:

      Yes there are hiatuses/pauses. Use the pause 1988 – 1998 for making policy decisions.

      • Reply to ehak: I wish I could reply near where the posts occurred.
        I) In Fig. 1, the warming during the 30’s was at least was warm as now (+0.5, typo, sorry), but is missing from the plot.
        no answer, add: I’m still wondering how the warming of the 30’s is missing from Fig. 1 and other plots. 26 U.S. state record max temps set in the 30’s stand to this day. This is just one of many examples of data tampering, in my opinion. The Medieval Warming Period was at least as warm today but is missing from the plots after AR1, another example of data tampering. Are we not concerned about data integrity and scientific integrity?

        II) The surface record continues to diverge from the satellite/balloon record.
        your answer: Radiosondes diverges from the SST-indices. That is correct. Higher trend. They also diverge from RSS.
        add: I was referring to the satellite/balloon air temps, not SST’s. Your answer was not helpful. The satellite derived global temperatures match the balloon observations. Does the discrepancy (divergence) between these data sources suggest one of them is incorrect? I’m going with the balloon data matching satellite.

        III) Where are the uncertainty estimates?
        no answer, add: I read of some uncertainty estimates in Karl’s paper (good) but we need to see more of that (+- 0.1), less of plots like Fig. 1.

        IV) What about the Argo data trends, pretty flat except for El Ninos of 2010 and 2015.
        your answer: And ARGO is not flat. Surface ARGO matches ERSST4.
        add: cite please. I think that Bob Tisdale may disagree with your assertion. The NOAA ERSST.v4 data may be suspect.

        V) The USCRN data show the same trend for last 10 years, close to zero. I know it’s U.S. only but that brings into question the veracity of the global surface record.
        your answer: https://tamino.files.wordpress.com/2015/09/rssminusrat.jpeg?w=750&h=498
        add: another good example of a plot with no uncertainty estimates. BTW, Tamino is on my no-fly list along with Mann and Gleick. You ignore the USCRN data, no surprise. Here’s a good WUWT link for you:
        http://wattsupwiththat.com/2015/07/21/tamino-grant-foster-is-back-at-his-old-tricksthat-everyone-but-his-followers-can-see-through/

        I trust the satellite and USCRN, the “adjustments” in the measured surface data are counter to everything I learned about respect for observations as ground truth. Here’s a recent article (opinion) by Lamar Smith. I agree, “NOAA’s habit of picking and choosing data raises serious questions about the agency’s independence. In fact, it shreds NOAA’s credibility.”
        http://www.washingtontimes.com/news/2015/nov/26/lamar-smith-noaas-climate-change-science-fiction/

        JerryG∞

      • Jerry Gorline:

        Hard to focus on the subject of this post it seems. Anyhow: You keep saying radiosondes track MSU TLT. Try this. Right after you have seen a plot with radiosondes showing no slowdown after 1998. On the contrary. RSS vs Ratpac A:

        And yes. The divergence suggests one of them is wrong. In this case RSS. RSS also diverges from RSS water vapor. Steady increase in RSS water vapor but not in RSS TLT.

        More repeating goes on for the ARGO showing divergence with ERSST4. Not true. Try this:

        http://wattsupwiththat.com/2015/01/22/learning-from-the-argonauts/

        And this:

      • ehak- the new graph from the authors, why didn’t they include OISST weekly: the original series?

      • Well… it is an El Nino.

        During the 2017 La Nina, Chak et. al. are going to get a real beating on the “no hiatus”, which they deserve.

        It takes 260 W-Years/m2 on the surface to warm the top 2000 meters 1K. The 1940 warming is still being incorporated.

        As far as the USCRN:

        It is obvious that all UHI should be removed from the temperature trend. Tweaking CO2 has no effect on UHI. The USCRN should have had no UHI. “Rural” sites have UHI.

        There is some psychosis or mental syndrome that precludes decent siting of these stations.

        NOAA should be forced by law to create a USRCRN (United States Real Climate Reference Network) They should be compelled by law to purchase 114 1/4 square mile plots of land in pristine areas upwind of urban areas (outside of any heat plume). The stations by law should be located in the exact center of the 1/2 x 1/2 mile plot. The stations will be 1/4 mile from any conceivable human influence. It should be required that any other artifact of humanity be removed from the land plot.

      • JCH:

        I don’t know. Ask them. Anyhow that version is close to Ersst3b. Not bias adjusted.

      • PA – La Nina is not your friend. She’s my friend. ENSO neutral is not your friend. It’s my friend. El Nino is not your friend. He’s my friend. Heat moving around within the ocean is not your friend, it’s my friend. If you think a La Nina in 2017 is going to to save your horrific interpretations of science, well, that’s why you have no friends. The ACO2 knob is turned to the right a little bit each day, and a little more water leaks into the sinking ship of the people who have fooled themselves. Now, go save some civilization from the liberal barbarians.

      • They are happy clowns when it’s warming. They are sad clowns making up excuses when it ain’t. We are supposed to take them seriously.

      • I just as happy during a warmest La Nina in the record as I am in the hottest El Nino in the record. The earth gets hotter, in most cases, in both. So sorry you’re so confused.

      • And yes. The divergence [ between RATPAC and MSU ] suggests one of them is wrong.

        No it doesn’t!

        Consider the coverage of RAOB versus the coverage of MSU.

        Of course there will be variation (+ and -, which is what your graph indicates )!

        Looking in the vertical over the full MSU period of record, however, gives one a much better indication of just how well the two forms of measurement agree ( and how much the models diverge ):

      • And one of the giggling clowns comes along to prove my point.

      • T Eddie:

        So sampling is a problem in comparing the MSU with radiosonde?

        Really. Tell Christy that. Who presented graphs to the President’s Council on Environmental Quality with no subsampling when comparing radiosondes and MSU.

        Also tell Gorline that. Who thinks those graphs validates the MSU series.

      • Tell Christy that. Who presented graphs to the President’s Council on Environmental Quality with no subsampling when comparing radiosondes and MSU.

        Are you thinking about his comparison of co-located sonde and MSU data? Yes, co-located sonde data would be better for comparing MSU nearest neighbors.

        So sampling is a problem in comparing the MSU with radiosonde?

        I didn’t say it was a problem so much as a reason that the sonde data will be more variable because it represents a much smaller portion of the globe than MSU, which is what your chart demonstrates.

        The models/raobs/msu all indicate cooling stratosphere and warming Arctic.
        Good.

        Neither raob nor MSU exhibit the very large modelled hot spot.
        Bad.

      • T Eddie:

        Christy did not “co-locate” MSU and radiosonde. No subsampling. In addition of course to his famous aligning of temperature to av very short span of years in the beginning of the series.

        And how does this have anything to do with the adjustments in ERSST4 and how those match unadjusted buoys? Nothing. Trying to avoid inconvenient facts is very telling.

    • All this talk of hiatus seems off the point. The real issue is not whether it is 0.00 or 0.02 but the overall trend of 0.11 over 70 years since 1945 is simply not enough to produce a gain that is close to the models predictions especially as you carry the predictions out into the future. The models showed a 0.5C higher temperature now. With that 0.5C a 2.0C gain by 2100 would be possible but with the current rate 1.0C is the outside edge of possibility or less from 1945 to 2100.

    • Ehak, you do realize your graph demonstrating no pause for water vapor, is flat until 2015. It starts increasing after the wasn’t El Nino and current El Nino. Not a very wise choice of data to demonstrate there was no pause. Especially when there clearly was a pause until a naturally occurring event in the Pacific.

      • Flat until 2015?

        Something wrong with your eyes Brian?

      • Gorline:

        I don’t know. Ask them. But it would not make much difference. OISSTv2 monthly corresponds closely to the non-corrected ERSST3b. So all that would prove would be that OISSTv2 weekly/monthly is not bias adjusted. But then again: we already know that. We also know ERSSTv4 does a good job in adjusting the ship-buoy bias. The comparison between buoy only temperature and ERSSTv4 (aslo ARGO) clearly demonstrates that fact.

        If you think the right thing to do is to ignore known biases, please explain why. If you think that is the right thing to do then you cannot claim the 30ies were so warm either:

  57. Does it seem to anyone else that this is part of a possibly unconscious effort to find every scrap of heating possible? Why did they initiate this analysis? If the result had come back with a positive bias would the article have been written? Would the data be “homogenized?” It’s a probalistic thing. If we look at 100 of these types of studies of errors in the record will we find that 95 of them find a cooling bias in the record? If so, then one has to at least suspect that a bias on the part of researchers possibly unconscious is going on that possibly we should either cancel some of the adjustments or somehow start a conscious effort to find positive bias’s in the data. I personally have never read a news story or article that ever said the researchers themselves ever found a positive bias. I wonder if anyone ever remember reading such a story?

    I know the adjustments are not ALWAYS to the upside. Biases in the past tend to be positive more which is in effect a negative bias found recently because it enhances the desired effect. I would love to see a meta-study of all these adjustments made. I guess we are getting some of that with the recent effort to reanalyze the land record adjustments by some group.

    My feeling is that there are so many adjustments made that many may overlap and produce an overall result that is meaningless.

  58. The graph in figure 3 is fairly convincing. I thought part of the problem though was the offset relative to the long term trend so I’m wondering if the zero baselining in the figure tells the whole story.

  59. Zeke and Kevin,

    Interesting way to attack this question, I also was wondering like Bob Tisdale, why you would remove measured data from your purpose built instruments. Anyway enough on that dead horse.

    I do have some concerns, you mention how to homogenize the data, the boy data was increased .1C. I read a few months ago it was something like .094C. So given both your number and the one I read months ago are quite similar its probably close. Heres my problem, when you made your graph putting V4 along side only buy data, I’m not buying it, period. The Y axis on your graph is in tenths of a degree, so there should be at least 1 increment difference between bouy and V4 data. Everyone agrees that too homogenize the boy data, it had .094C added to it and it accounts for roughly 50% of the homogenized data after 2000. Further towards the end of the data, there is approx 70% spacial coverage of the oceans, in your words. Reading Karl, et al, it is clear, boys account for .026C of the entire global increase of .064C. Almost 50%!

    Seeing that it is impossible that there is no discernible difference between buy only and total V4 data.

    Finally, both you and Karl, say the purpose of the corrections is to correct for cooling biases in the bouys, that lets remind ourselves, you said are purpose-built instruments. Why on earth would we remove a ‘cooling bias’ from data coming from instruments that are purpose built, when we know for a fact that the ERI data is biased warm. Why would we not remove the warming bias and leave the high quality data alone. Unless of course it was intended for the sole purpose of increasing the last decade and a half of the global temp record. I for one agree with you, the trend would have been the same had they removed the .1C from the known biased warm data, however the warming would have happened in the part of the data set that most scientists agree was natural warming. Instead of happening during the part of the data when 40% of all manmade carbon was added to the atmosphere.

    • The crux, Mr. brian:

      “I for one agree with you, the trend would have been the same had they removed the .1C from the known biased warm data, however the warming would have happened in the part of the data set that most scientists agree was natural warming. Instead of happening during the part of the data when 40% of all manmade carbon was added to the atmosphere.”

      Science skewed by politics.

  60. I have scanned the Supplementary Material (SI) for Karl et al and found some interesting and puzzling handling of the observed temperature trends, trend differences and methods applied to these estimations.

    First I find Karl’s use of 90% Confidence Intervals (CIs) for statistical significance of a trend being different than 0 an unusual approach. The standard in papers of this kind is using 95% CIs and then noting, if the test fails significance at the standard level but is close, that significance is approached closely in the estimation. That there are 2 different calculations and values of CIs and along with the use of a non standard significance level could allow for some (mis)quoting of differences that are not well detailed or explained in the quote. This situation is particularly problematic when used by the lesser informed like politicians and journalists.

    It appears to me that Karl et al did not use an ARMA simulation – that would be derived from the best scoring aic model – to determine the CIs for the trend. The Monte Carlo approach would have better included the effects of autocorrelation on the regression residuals. The authors instead used the less efficient method of Santer et al 2000 and Quenouille (1952) which adjusts the degrees of freedom for ar1 autocorrelation when determining CIs.

    I also note that I could not find a statistical significance test of the difference in trends derived between using the Karl data set and the older data set(s). Looking at the trends and the corresponding CIs in the Karl SI would indicate that there is no significant differences in the derived trends between data sets.

    I also do not know where the trend comparisons between from 1950 to the late 1990s and the late 1990s to near present time comes from but that has to be wrong-headed when considering the ramping up of the GHG levels from 1970 onward.

    Zeke Hausfather and Kevin Cowtan or anyone here: can you provide a convenient link to the Karl et al data series so that I can do some of my own testing and analyses? I would want to determine the best ARMA fit and determine regression CIs using Monte Carlo simulation with that model. I would also like to look at linear and non linear trends and, of course, I would want to calculate 95% CIs.

    • kenfritsch:

      I asked about confidence intervals in previous CE posts including 2 or 3 exchanges on the generic use of p-hacking with dikranmarsupial (who is apparently Dr. Gavin Cawley). https://judithcurry.com/2015/11/06/hiatus-controversy-show-me-the-data/#comment-742386

      Given some barely significant results at 90% (and some failures to achieve “significance”) I’ve been curious as to whether the Karl hiatus paper was reworked after failing initial tests at 0.05. Although there seems to be no hard-and-fast rule for climate science, I wonder why a single agency would be allowed to use both 0.05 and 0.10, depending on the specific analysis being conducted.

      Perhaps you can unravel their process. I hope you will post the results.

      Kent

  61. From Karl et al (2015) we have the following:

    “In essence, the bias correction involved calculating the average difference between collocated buoy and ship SSTs. The average difference globally was −0.12°C, a correction that is applied to the buoy SSTs at every grid cell in ERSST version 4.”

    Does anyone know what the CIs are on that -0.12 degree C difference or what the distribution of differences looks like or better where might that data be linked?

  62. I like that this post says:

    As an aside, the decision to adjust buoys up to ERIs or ERIs down to buoys should nominally be trend neutral. Indeed, in their work on HadSST3 Kennedy and colleagues explicitly tested this, and found “no appreciable difference” on trends.

    Because it shows Zeke Hausfather has developed a better understanding of the issue since a tweet he posted earlier this month where he responded to our hostess to say:

    @curryja you do realize that adjusting buoys up to engine intakes or engines down to buoys is by definition trend-neutral, right?

    Which is false. The choice of which data set to adjust when adjusting two data sets to bring them into alignment is not, by definition, trend neutral. There are cases where the choice can have a non-neutral effect, particularly on regional scales. They may not be particularly common or expected, but the possibility is not something which can simply be dismissed as impossible. That’s particularly true given there are a wide variety of possible ways to create temperature records, and what’s true for one is not always true for others.

    I am, however, a bit confused as to why Zeke chose to cite a tweet mentioning tests and not the paper which actually discussed those tests and showed their results. I find it more convincing to actually be able to see the details and results of tests rather than just a 140 character summary of them. Plus, it is kind of interesting to see the (very) minor differences not adjusting the buoys would result in for that data set.

  63. The Obama political minion heading the NOAA has responded to the House Science oversight committee. This is the least one-sided account I could find. Pretty atrocious reporting:

    http://www.csmonitor.com/Science/2015/1124/Scientists-hold-their-ground-against-Rep.-Smith-s-NOAA-subpoena

    They say they ain’t giving up no stinking emails. They can get away with stonewalling Congress, for as long as their lawless boss Obama controls the DOJ. A citizen FOIA will get results sooner.

    • The story plays up the stinkly “rushed” red herring and fails to mention the substantive allegations of the NOAA scientist whistleblowers: politicization of the science and failure of Karl and his gang to follow established NOAA scientific procedures and methods.

      And the dispute is framed as an attack on the scientific community by Lamar Smith “prominent climate denier”. But not to worry good folks “the scientific community doesn’t show any signs of backing down.” They think it’s the alleged climate science community that gets to decide how and when the Congress can exercise it’s oversight powers, rather than the U.S. Constitution and the Supreme Court.

    • I’ve yet to see a legal justification for withholding the requested information. If a legal defense against a Congressional subpoena of publically funded and publically owned materials existed, I assume NOAA would have mentioned it by now.

  64. This is interesting. Maybe the numerate can tell us if it’s a plausible analysis;

    https://ecologicallyoriented.wordpress.com/2015/06/06/karl-et-al-2015-updated-analysis/

    “And there are some serious discrepancies in my trend values vs theirs. For example, Karl et al report the trend from 2000 to 2014 as 0.116; I get 0.0649. For 1998-2012 they report 0.086 and I get 0.034. The numbers I’m getting are much closer to their reported “old” (bias uncorrected) values, than to their “new” ones, but they don’t exactly match those either. And the data link names provided by NOAA at their ERSSTv4 page do not seem to refer to version 4. So, some problems here.”

    • Also, there is this:

      ” As mentioned in the piece below, I had suspicions that the data at the NOAA ERSST v4 webpage were not the correct (version 4) data, and it turns out that this is in fact the case. I learned this after Gavin Schmidt of NASA forwarded me data he had been sent directly from NOAA personnel.”

      This was written in June. I wonder if the correct data is on the NOAA ERSST v4 webpage, now. Does this have any effect on the analysis in this post.? Let the numerate step up and inform myself and other avoiders of the math.

  65. In is good that Hausfather and Cowtan have shown up on this thread to report their work and leave comments. Unfortunately they evidently have no better sources than I do for the land data used in Karl et al (2015).

    The SST portion of Karl can be downloaded from KNMI as the ERSST v4 data set. I have downloaded that data and the data from the earlier version ERSST vb3. From the Karl supplemental material I see the land part of their new data set comes from the unadjusted ISTI v1.0.0 databank and then adjusted by the methods of Menne and Williams followed by placing the adjusted data into 5×5 degree grid boxes. From the gridded anomalies a global analysis for each year-month using Empirical Orthogonal Teleconnections as describe by Smith is performed and finally cosine weighted to produce anomalies 1880 to present. I cannot find any link to that final adjusted land data set. If that is the case it presents for me a major negative view of the Karl paper..

    To Steve Moshner: Please no inane remarks about doing these adjustments myself.

  66. I keep hearing the adjustments are trend neutral. This may be true. The real question is, “Are they anomaly neutral?”

    Going from 10°C to 11°C in one year is he same trend as going from 30°C to 31°C in one year.

  67. Ehak, curious minds would want to know why the series that Springer pointed to track well from 1981 to around 2005 and then rapidly diverge.

    A breakpoint analysis might be the proper analysis to determine whether a regime change occurred, but visual motivates to go back and lok at what Karl presents in the paper and the SM. How about you?

    • I believe 2005 was when the buoy deployment reached the targeted coverage. Since prior to that coverage was spotty, it could be that the anomaly period didn’t have enough coverage for a reasonable seasonal signal compared to a more global coverage.

      The absolute values just show the impact of the 0.12 C adjustment bringing ERSSTv4 up to the other products. Changing the anomaly baseline wouldn’t vary the trend normally, but since it does impact warmest years EVAH, it could in this case.

    • Of course the effect of the ERI – buoy adjustment increases when the fraction of buoys increases. Agree? If you prefer to call that a regime change, please do.

    • David Springer

      Karl 2015 is SO busted. The pause is evident in every time series except theirs. They pencil whipped it into oblivion. These are not honest brokers. Of course we knew that already.

  68. David Springer

    I can hardly wait to see what the congressional inquiry reveals. Lamar Smith you go boy! Makes you proud to be a Texan, eh, ehak?

    • David Springer said:
      I can hardly wait to see what the congressional inquiry reveals. Lamar Smith you go boy! Makes you proud to be a Texan, eh, ehak?

      Thanks David, I think these adjustments in observations have given the Gov’t data stewards a blank check to alter the ground truth to fit a political agenda. I hope this investigation into the Karl et al. paper (including the emails) exposes any attempts to present advocacy masquerading as science. It’s about time.

      I have tried to point out that the satellite record continues to diverge from the surface record and that the satellite derived temperatures match the balloon data. I don’t know how to insert an image here but I have a link (see Fig. 1). Ehak has posted ratpac A vs. RSS. I like Fig. 1 in the link, UAH vs. Balloon. It’s funny, ratpac is “adjusted”, I think Argo is being “adjusted” too. The powers that be even tried to “adjust” the USCRN. I’m having trouble believing anything from NOAA (credibility problems). There is a lot of hand waving with the buoy data adjustments, other data sets including the ERSST, etc. I think that the “hiatus busting” paper may have busted the entire bank and these so-called adjustments. Bye for now.

      https://www.masterresource.org/debate-issues/john-christy-climate-scientist-on-nepa/

      JerryG∞

      • Gorline: The most adjusted series of all are satellite MSU/AMSU. This is Roy Spencer:

        “All data adjustments required to correct for these changes involve decisions regarding methodology, and different methodologies will lead to somewhat different results. This is the unavoidable situation when dealing with less than perfect data.”

        http://www.drroyspencer.com/2015/04/version-6-0-of-the-uah-temperature-dataset-released-new-lt-trend-0-11-cdecade/

        But at last, this time you had to realize that radiosonde tropospheric temperature diverge from RSS TLT. RSS TLT even diverges from RSS water vapor.

        USCRN is something quite different than ocean temperature. In case you did not know that. The adjusted USHCN matches USCRN almost perfectly. That might be a problem for you.

    • Gorline:

      Data fiddling in the surface records? Those are small in comparison to the UAH TLT/TMLT record.

      And we know there are biases in SST. Measurements form buoys are lower than from shipengine intakes. Easy to find out. Take measurements in the same area with ship intake measurements and bouys. Known bias. Correct for that. And as we see in this post: Unadjusted buoys matches adjusted ERSTTv4 much better than unadjusted ERSTTv3b. Shows the validity of that adjustment.

      USCRN might very well change an OLS trend from positive to flat in just one year. That just proves the point that a series with this kind of variability and short timespan will never give a statistically significant trend.

      The point here however was that the adjusted non-USRN stations matches USCRN. Shows validity of that data fiddling as you prefer to call it.

    • Thanks David for leaving a hook open for my response to ehak@. Mr. Ehak, there is this concept called credibility and Roy Spencer’s adjustments to the UAH MSU data are reasonable and easily explained. The methods and product results are reliable. I trust Dr. Spencer and Dr. Christy because these scientists are careful with the data. I don’t trust NOAA any more. Dr. Christy’s paper on Sierra Snowfall over the last 130 years – no trend, no effect from CO2, is stellar. Nobody questioned his methods during his talk at the 2009 AMS meeting. Somebody asked about his funding from oil companies. Somebody always does. The answer was “no.”

      I don’t know why you persist with your argument about radiosondes and RSS divergence. The link I provided (Fig. 1) shows the average of two satellite datasets and 4 balloon datasets that are in agreement, compared with CMIP-5 predictions. The models are clearly over-predicting and it seems that the observed surface record data adjustments have produced an unjustified warm bias. It’s clear to me why, politics, and that is unfortunate. Why perform adjustments on the USCRN data? The sites are in pristine locations. You don’t have to be snarky, it’s clear that the USCRN data are from land-based sites. Here’s a better plot for the 10-year USCRN trend:

      http://www.ncdc.noaa.gov/temp-and-precip/national-temperature-index/time-series?datasets%5B%5D=uscrn&parameter=anom-tavg&time_scale=p12&begyear=2005&endyear=2015&month=12

      The point is that the USCRN is showing a 10-year trend not significantly different from zero. Meanwhile we have this global surface record of land + ocean data from Karl et al. Are you familiar with the acronym FUBAR?

      Cheers,
      JerryG∞

      • Gorlin has reached the stage of gishgalloping. Trying to avoid the subject of this post. The ship-buoy bias and how to correct that bias.

        Anyhow, here we go: Suddenly some adjustments are reasonable. Like the Spencerized new TMLT. I spite of the fact than Spencer for years argued that RSS had a wrong adjustment for the diurnal drift.

        Now we know. They cannot even agree with themselves. But the adjustments are ok. It is ok to do adjustments. Then we must judge other adjustments accordingly. Like this post where we see that the adjusted record matches non-adjusted buoys clearly better than a record where the ship-buoy difference is not corrected for.

        The Christy Fig. 1 shows the same divergence between radiosondes and TLT as in my graph. The MSU record running higher in the late 90ies and lower after 2005 than the radiosondes. In addition In Christy’s used an average of UAH LTL 5.6 and RSS. So he is actually using a version that is wrong according to you and Spencer and Christy. If they had used the version 6 betasomething the divergence between the radiosondes and the satellites would have been even greater. I guess you know by now that there is not much sign of “pause” in version 5.6. Version 5.6 agrees much better with the radiosondes. But that is hidden when using an average of RSS and version 5.6.

        Anyhow. Thank you for drawing attention to the divergence between radiosondes and TLT.

        Whoever said USCRN is adjusted? Stop dreaming. The point was that a record consisting of adjusted non USCRN stations matches almost perfectly USCRN.

        https://www.ncdc.noaa.gov/temp-and-precip/national-temperature-index/time-series?datasets%5B%5D=uscrn&datasets%5B%5D=climdiv&parameter=anom-tavg&time_scale=p12&begyear=2004&endyear=2015&month=10

        So please tell me: Does that tell us something of the validity of the adjustments of the non-USCRN stations?

        Also you should know that a 10 year period in this kind of series will never show a significant trend. It is impossible. The variability is far to large for that.

      • Hi Mr. Ehak. Gishgalloping? I think you are projecting. Are you Gavin? I am talking to a brick wall but here goes, I have said that Spencer and Christy’s credibility has remained intact. I can’t say the same for NOAA, Karl in particular. I am tired of the unreasonable data fiddling with surface observations. The UAH temps are derived and adjustments are necessary for systematic biases and are clearly identified. The divergence in the 90’s between the balloons and the UAH/RSS data are probably from the strong El Nino of ’97 – ’98. The balloons did not pick up that warming, sparse launching site locations I would speculate. I never said that the RSS data were wrong. It looks like your gish gallop skills exceed mine, not that I’m trying. As to USCRN adjustments, I have seen earlier comparisons where the USCRN seemed to show a slight positive trend. Later versions don’t show that. I don’t have those earlier plots but I did see them. I ran a by-month average for the conterminous U.S (1948 – 1998), TD-3200 data. I found no trend for any month. Since my analysis, the 10,000 active COOP sites have been reduced to about 1200, I think. Something strange is going on and I hope Congressional oversight can find out what it is. You can continue defending a weakening position if you like. It’s a free country. The pause is real and the next five years will confirm that fact.

        JerryG∞

  69. Pingback: A Buoy-Only Sea Surface Temperature Record Supports NOAA’s Adjustments | Enjeux énergies et environnement

  70. Pingback: Any doubts about Climate Change? - Page 124 - TeakDoor.com - The Thailand Forum