Climate Etc.

Nature Unbound V – The elusive 1500-year Holocene cycle

by Javier

The existence of a 1500-year climatic cycle during the Holocene, related to the glacial Dansgaard-Oeschger cycle, is a matter of intense debate. However, by introducing precise timing requirements it can be shown that the 1500-year cycle displayed in Northern Hemisphere glacial records is also observed in Holocene records from all over the world.

The cycle is most prominently displayed in oceanic subsurface water temperatures, Arctic atmospheric circulation, wind deposits, Arctic drift ice, and storminess records. A lunisolar tidal hypothesis is uniquely capable of explaining the cycle’s timing, features, and effects. The hypothesis allows for the observed power in the 750-year harmonic, apparent lack of temperature signal in the cycle, and a phase shift during the Holocene Climatic Optimum observed in some proxies. While the Holocene 1500-year cycle is associated with cooling indicators in the Northern Hemisphere, its effect on global temperatures is undetermined. The next peak effect of the 1500-year cycle is expected in about 160 years.


Continuing the series of articles about abrupt climate change that started with the glacial-interglacial cycle, and the Dansgaard-Oeschger cycle, and followed with the nature of Holocene climate variability (Part A, and Part B), I last reviewed the ~ 2400-year Bray climate cycle (Part A), and its solar origin (Part B and Part C). In this article, I hunt through the literature for solid evidence on “The enigmatic 1,500-year cycle,” in the words of Raimund Muscheler (2012), and “one of the outstanding puzzles of climate variability,” according to Maxime Debret and colleagues. (2007).

The discovery of the 1500-year glacial Dansgaard-Oechsger cycle (DO) in Greenland ice cores, started a quest to identify this cycle in Holocene records, where its absence was puzzling. In 1995 Gerard Bond and Rusty Lotti demonstrated in deep sea cores that glacial-time DO events matched periods of increased iceberg activity and that they were linked to Heinrich events (periods of very high iceberg activity in the N. Atlantic). The records showed 3-4 DO periods taking place between Heinrich events. This oceanic sedimentary pattern from the glacial period became known as the “Bond cycle.” Since 1996 researchers started to report a 1500-year periodicity in Holocene proxies, and it became Bond’s goal to extend his cycle to the Holocene. He claimed to have achieved that in 2001 when he reported his famous drift-ice petrological record that revealed a series of cold events in the Holocene (Bond et al., 2001; figure 48). Bond’s report sparked a paradigm shift because at the time the Holocene was generally viewed as climatically stable, despite contrary evidence from glacier studies. We have seen that Bond was wrong (figure 48), and his misleading numbering of the cold events to try to fit them into a 1500-year series has caused unnecessary confusion in the field. Researchers all over the world have tried to match negative temperature anomalies and even precipitation fluctuations in their proxy records to Bond events, perpetuating the myth. As the evidence is contradictory among reports, the existence of a 1500-yr cycle in the Holocene has become more contentious with time. After an extensive research Wanner et al. (2011) concluded that “multicentury cold events were not strictly regular or cyclic, and one single process cannot explain their complex spatiotemporal pattern.”

Could Bond be wrong, Wanner be right, and still there be a manifestation of the 1500-yr DO cycle in the Holocene? That is what the evidence supports, and the nature of the evidence points once more to the process that is the likely cause of this climatic cycle.

What must we expect of a Holocene 1500-year cycle?

Over the last decade we have greatly increased our knowledge of the DO cycle. We know that DO events are abrupt warming events, that take place at the North Atlantic-Nordic seas area. They appear to take place when the water stratification is abruptly disturbed, allowing a surge of warm subsurface waters that have accumulated for long periods of time, to melt surface ice and release heat to the atmosphere (figure 29). They require high ice conditions and low sea level, between 45 and 90 m below the present level (figure 25). Their trigger is very precise, taking place every ~ 1500 years (figure 18) and maintaining this precise pacing through periods when the conditions are not met and the events do not take place, sometimes for over 10,000 years. That new events after such long attenuation periods still occur at the specified timing, without significant drift, strongly supports the existence of an external pacing control mechanism for DO events. The precision of this mechanism over time suggests an astronomical control. Wolfgang Berger and Ulrich von Rad (2002) proposed that the 1500-year cycle is a harmonic of the beat between the moon’s nodal and apsidal precessions, a hypothesis that fits not only the observed period, but also the required mechanism for vertical water mixing through tidal forcing.

As described, DO events cannot take place during the Holocene. It is too warm, there is too little ice, and sea level is too high. New DO events will have to wait until the next glacial period. However, the trigger clock is ticking all the time to keep its pace, and it is fair to ask if it might have other effects that could cause a 1500-year climate cycle during the Holocene.

We already know some characteristics that the 1500-year cycle should present: It should take place when the trigger indicates, and thus in phase with the DO cycle; its manifestation should be compatible with a mechanism that promotes vertical water mixing (mainly wind and tidal forces); and it might not be due to internal variability of the climate. Given that the warming nature of the DO events relies on the presence of abundant warm subsurface waters in a specific region, the 1500-year Holocene cycle doesn’t have to be a warming cycle. In fact, if the trigger promotes vertical water mixing, in most places this would mean sea surface cooling.

To my knowledge no author has set any requirements on the 1500-year cycle, except the periodicity, and since almost every proxy presents some periodicity, often for no reason, several of the reported 1500-year proxy periodicities might not correspond to the cycle that caused DO events during glacial periods.

The next question is where to look for the 1500-year cycle. As Bond identified the DO cycle in North Atlantic sediments, he set out to identify the 1500-year Holocene cycle there too. However, as we saw with the ~2400-year Bray cycle, the North Atlantic is especially sensitive to solar activity decreases (see figure 67), and therefore, most Bond events represent solar-induced cooling events. Some of the peaks in the Bond series cannot be assigned to solar activity and could represent cooling from the 1500-year cycle. In fact, it has been shown that during the Neoglacial the Bond series can be better fitted to a 1500-year periodicity (Debret et al., 2007; figure 48). However, given the strength of the solar variability signal in the North Atlantic, proxy records from that area usually present mixed periodicities more difficult to interpret. This has created another problem, as most authors have looked for the cycle in the North Atlantic area, following Bond’s steps. The result has been more confusion in the scientific literature and some claims that the 1500-year cycle was perhaps of solar origin.

Before reviewing the evidence for the 1500-year cycle let’s see where the DO cycle stands at the start of the Holocene. According to Rahmstorf (2003) the abrupt warming changes that started the Bølling period and the Holocene correspond to DO events 1 and 0 respectively, and they are separated by 3000 years (figure 69). He also places another event in the middle (DO-A) as possibly responsible for the warming after the Intra-Allerød cold period.

Figure 69. The Dansgaard-Oechsger cycle at the end of the last glacial period. Temperature curve derived from GISP2 ice core. DO events are indicated with continuous grey lines. Dashed and dotted grey lines represent harmonics of the DO cycle. Cold periods are indicated above the curve, and warm periods below. IACP: Intra-Allerød cold period. Sources: S. Rahmstorf 2003. Geo. Res. Let. 30, 10. R.B. Alley 2000. Quat. Sci. Rev. 19, 213-226.

To check the time relationship between proxy records and the DO clock, I have projected the DO periodicity observed in the 15,000-11,000 year BP interval to the rest of the Holocene.

The ~ 1500-year periodicity during the Holocene

For the last 20 years researchers have been reporting a ~ 1500-year periodicity in Holocene climate proxies. In 2007 Maxime Debret et al., carried out a wavelet analysis of some of these proxy records. The wavelet technique allows to determine two-dimensionally not only the periodicities present in the record, but also the times at which those periodicities are found. The results from their analysis are shown in figure 70.

Figure 70. Wavelet analysis of climate proxies over the Holocene period. Occurrence of the periods with respect to the time is given by the bright yellow-red colors. Black boxes indicate the position of the ~ 1500-year periodicity. a) Percentage of ice-rafted detrital petrology in marine sediments. b) Sediment grain size as a proxy for ocean current intensity. c) Emiliania Huxleyi concentration, a proxy of surface hydrology. d) Sodium flux in GISP2 ice core, a proxy for atmospheric circulation. e) Magnetic susceptibility from loess aeolian deposits, a proxy for wind activity. f) Marine ice core sediment lightness (color) as a proxy for changes in North Atlantic deep-water circulation. The spread in the ~ 1500-year period can be due to differences in the age model for each proxy. Source: M. Debret et al., 2007. Clim. Past, 3, 569–575.

The authors conclude from the wavelet analysis that the Holocene millennial variability is composed of three main periodicities, 1000, 1500, and 2500-year cycles. Based on the coincidence of the 1000 and 2500-year cycles with the wavelet analysis profile of 14C production rates they defend their solar origin. They assign an oceanic origin for the 1500-year periodicity because it is absent from the solar proxy and present in oceanic proxies. The wavelet analysis also shows that in some of the proxies the 1500-year cycle is not continuous through the Holocene, being absent or very attenuated during the early Holocene (figure 70, a, b, d, e), while in other, perhaps more sensitive, records the 1500-year periodic signal appears a few thousand years earlier (figure 70, c, f). It is possible that some of the climatic manifestations of the 1500-year cycle might have been masked by the conditions of the Holocene Climatic Optimum while becoming easier to detect during the Neoglacial.

The oceanic 1500-year cycle

Among the possible explanations for the 1500-year cycle proposed by different authors, most appear to attribute it to an oceanic oscillation, whether externally forced or due to internal variability. However most oceanic proxies that contain a 1500-year frequency signal display a very complex pattern, indicating that the proxy is affected by other climate cycles and changes, and precluding a clear identification of the 1500-year cycle.

Sea-surface temperature (SST), is affected by changes in wind patterns, wind strength, insolation, cloud cover, pressure, and precipitation patterns, among other factors. Thus, most SST proxy records, especially those from the North Atlantic, do not display a recognizable 1500-year cycle. One exception is the northwestern Pacific SST alkenone-based proxy record from the coast off Japan reported by Isono et al. (2009; figure 71 a). Although the temperature reconstruction does present a 1500-year periodicity that matches the DO pattern, the match is not very good and presents a clear 180° phase shift at 7 kyr BP (figure 71 a, black bar), when higher temperatures went from taking place at mid-cycle before the shift, to lower temperatures taking place at mid-cycle afterwards.

Figure 71. Oceanic proxy records displaying the 1500-year cycle. a) Variations in detrended alkenone-derived SST from a core off the coast of central Japan in the northwestern Pacific. Source: D. Isono et al. 2009. Geol. 37, 591–594. b) Variations of oxygen isotope signature from Pulleniatina obliquiloculata calcite in a sediment core at the Okinawa Trough at the southeastern part of the East China Sea, as a proxy for top of the thermocline temperatures. Source: L. Wang et al. 2016. Geo-Mar. Lett. 36, 281-291. c) Marine sediment core 5-18 μm siliciclastic fraction originated from Taiwan rivers and transported by the Kuroshio Current to the Okinawa Trough as a proxy for the strength of the Kuroshio Current. Source: X. Zheng et al. 2016. Earth Planet. Sci. Lett. 452, 247-257. d) δ18O measurements from Globigerina bulloides from a Murray Canyon (Great Australian Bight) marine sediment core as a proxy for intermediate water temperatures. Source: M. Moros et al. 2009. Quat. Sci. Rev. 28, 1932–1940. DO periodicity is indicated with continuous grey lines. Dashed and dotted grey lines represent harmonics of the DO periodicity. Black bars: Periods with an apparent 180° phase shift in data periodicity. Thick olive green curve: 1500-yr periodicity.

Water temperatures right above the thermocline (usually 50-100 m. depth) are less affected by precipitation and insolation changes, and can be determined by the oxygen isotopic composition, or the Mg/Ca ratio, of sedimented shells from foraminifera that inhabit that zone. Wang et al. (2016) determined the 18O variations from the foram Pulleniatina obliquiloculata, a thermocline dweller, in a marine sediment core from the Okinawa Trough, where the Kuroshio Current transports warm waters from the tropics to higher latitudes. The sea subsurface temperature proxy displays a very clear 1500-year cycle in phase with the DO cycle, where for the past 7000 years higher temperatures at the thermocline took place at the times determined by the DO periodicity (figure 71 b). Curiously this proxy also seems to present a 180° phase shift around 8000 years ago, as previously it was lower temperatures and not higher ones that coincided with the DO periodicity (figure 71 b, black bar). And it is not only the temperature, but also the strength of the Kuroshio Current that appears to be affected by the 1500-year cycle, as this periodicity is found also in the 5-18 µm particle fraction that originates from Taiwan rivers and is transported north by the current into the Okinawa Trough (Zheng et al., 2016; figure 71 c). At some of the times that coincide with the DO periodicity the presence of this fraction collapses, indicating a sudden drop in the current strength, probably due to the Kuroshio Current not entering the Okinawa Trough.

The effect of the DO cycle on thermocline water temperatures was already reviewed in the DO cycle article (see figure 29; Dokken et al., 2013), and was one of the arguments used to support its possible tidal nature. It is therefore very interesting that the Holocene 1500-year cycle displays the same manifestation, not only in the Northwest Pacific, but also in the Southern Ocean. Moros et al. (2009) determined thermocline water temperature changes during the Holocene at Murray Canyon (Great Australian Bight) using the oxygen isotopic signature of another thermocline foram, Globigerina bulloides. Interestingly the record not only displays a clear 1500-year periodicity in phase with the DO cycle (figure 71 d), but it also has a characteristic saw-tooth aspect quite similar to the reported proxy record from the last glacial period in the Norwegian Sea (figure 29). And again, we find a period between 9-8,000 years BP when the record displays a 180° phase shift (figure 71 d, black bar).

Although these oceanic proxy records are all very consistent in displaying an in-phase 1500-year periodicity, they don’t give a consistent temperature signal. The NW Pacific SST proxy and the Kuroshio Current subsurface temperature proxy display a warming signal at the specified DO periodicity, while the Southern Ocean proxy displays the opposite signal. The warming nature of the glacial DO cycle appears to respond to the specific conditions of warm water accumulation below a fresh cold thermocline layer in the Nordic Seas and is not intrinsic to the cycle nature. At the position where the SST were determined by Isono et al. (2009; figure 71 a) the warm Kuroshio Current and the cold Oyashio Current mix, while the Great Australian Bight record site is close to the present northern limit of sub-Antarctic water. Therefore, the temperature response to the 1500-year pacing might be determined by regional or hemispheric climatic conditions and thus the cycle does not appear to convey a temperature signal by itself.

The atmospheric 1500-year cycle

In 1997 Paul Mayewski and colleagues reported that the chemical ions found in GISP2 ice core presented a 1450-year periodicity during the last glacial period, that extended into the Holocene. The chemical species that precipitate on polar snow are introduced into the atmosphere by sea salt aerosols and continental dust, and their abundance depends on atmospheric conditions. Using the relative abundance of these chemical tracers, Mayewski et al. (1997) reconstructed a Polar Circulation Index (PCI) that provides a relative measure of the average size and intensity of polar atmospheric circulation. In general terms, PCI values increase (e.g., more continental dusts and marine contributions) during colder portions of the record (stadials) and decrease during warmer periods (interstadials and interglacials). Although the amplitude of the PCI changes decreases markedly during the Holocene, due to its much tamer climate variability, the 1450-year cycle persists in the record. This indicates that the nature of the periodicity is the same, and that the increase in PCI that accompanies the cycle is associated with more active atmospheric circulation, due to colder winter conditions. A periodogram for the last 11,500 years shows that not only the 1450-year peak is significant at the >99% level, but also the 725 and 2900-year harmonics, and that neither of these peaks can be assigned to 14C production variability indicative of a solar origin (Mayewski et al., 1997; figure 72).

Figure 72. Power spectra for the Polar Circulation Index during the Holocene. PCI series covering the last 11,500 years (continuous line) and the 14C residual series (dashed line) for the same period. Spectral peaks above the respective horizontal lines are significant at the >99% significance level. Source: P.A. Mayewski et al. 1997. J. Geophys. Res. 102, 26345-26366.

Another atmospheric-linked proxy record that displays a very clear, in phase, 1500-year periodicity comes from the Arabian Sea, where dust from the Arabian desert containing rare-earth elements is deposited after being transported by northwesterlies (Sirocko et al., 1996). Determination of the variability of a group of these rare-earth elements in a marine core efficiently removes analytical errors for each of them, and the resulting REE-score shows a very significant 1500-year periodicity that is in phase with the DO cycle (figure 73 a). In addition, this proxy record displays a 1500-year cycle continuously for the past 19,000 years (only 11,700 shown in figure 73 a) supporting that the pacing mechanism is always ticking and not affected by the drastic climatic changes that took place between 19-11 kyr ago. Furthermore, for the first 10,000 years (19-9 kyr BP) the increase in dust took place at mid-cycle, while for the last 8,000 years the mid-cycle showed a decrease in dust. Between 9 and 8 kyr BP the cycle displayed a 180° phase shift (figure 73 a, black bar) as in previous cases. This phase shift is probably affecting the cycle periodicity determination, reported as 1450-year (Sirocko et al., 1996). When plotted, the oscillations are clearly 1500-year long on both sides of the shift, but the introduction of a half length oscillation, to produce the 180° phase shift, changes the average to ~ 1450 years.

Figure 73. Atmospheric proxies for the 1500-year cycle. a) Changes in Arabian dust rare-earth elements abundance in an Arabian Sea sediment core. Black bar: Period with an apparent 180° phase shift in data periodicity. Source: F. Sirocko et al. 1996. Science 272, 526-529. b) Isothermal Remanent Magnetization as a moisture proxy in two lake cores (WL 03-1 continuous, WL 03-2 dotted) from White Lake, New Jersey, USA. Source: Y-X. Li et al. 2007. The Holocene 17, 3-8. Purple bar: Position of the 4.2 kyr arid event.

The cyclic increase in Arabian rare-earth element containing dust probably reflects an increase in northwesterlies strength, although a cyclic increase in aridity reflected in a higher dust production cannot be ruled out. In fact, evidence of the association between the 1500-year cycle and precipitation has been found in the Mid-Atlantic region of the US, where two lake sediment cores show that periods of low lake levels took place at a 1500-year periodicity (Li et al., 2007; figure 73 b). Exposure of marls due to low lake levels led to their oxidization and magnetic intensity increase, followed by their transport and re-sedimentation. Despite the age-depth model imprecision of this proxy and both cores showing different peaks, the coincidence of the peaks with the DO periodicity appears clearly within dating error. The question of the possible 1500-yr cycle association with a precipitation cycle in certain regions is an interesting one that deserves more research, as we approach a new cycle peak (~ 2180 AD) and precipitation is so critical to our society.

The 4.2 kyr event

At about 4,200 yr BP an abrupt climate change took place that had a strong aridity effect at middle and low latitudes in Africa, the Middle East and southern Asia. The intense drought reduced precipitation by about 30% for about 100-200 years likely causing the end of the Egyptian Old Kingdom, the collapse of the Akkadian Empire in Mesopotamia, and initiated the dispersion of the urban Harrapan civilization in the Indus Valley. The 4.2 kyr event is also seen throughout the Northern hemisphere but in a more complex and irregular manner, unlike most Holocene cold events. Although intense cooling is detected in Iceland lake sediments at 4.2 kyr BP (Geirsdottir et al., 2013), it is brief and completely reversed in about 100 years. Glacier advances are also recorded at the time in Central Asia, the Southern Hemisphere and North America (Mayewski et al., 2004). Interestingly, the 4.2 kyr event is also seen in the GISP2 Greenland ice core. It shows as a significant drop in chlorides (sea salt) concentration (a sea ice proxy; Mayewski et al., 2002), unlike most cold events of the Holocene, suggesting that the cold might have been accompanied by reduced precipitation.

Proxies indicate that the 4.2 kyr event is centered in the Arabian sea region, affecting the East African and Asian monsoons, the Mediterranean and Southern Europe, with a smaller effect on the North Atlantic region and South America, while the cooling appears global. A Kilimanjaro (East Africa) ice core presents a 200-fold increase in dust particles at the time (Thompson et al., 2002; figure 74 a), while a marine sediment core in the Gulf of Oman presents a 10-fold increase in wind transported dolomite from the Mesopotamian region (Cullen et al., 2000; figure 74 b).

Figure 74. The 4.2 kyr event. a) 50-year average of the Holocene dust history from Kilimanjaro ice core NIF3. Dust concentration measured as 0.63-16.0 µm diameter particles per ml sample. Source: L.G. Thompson et al. 2002. Science 298, 589-593. b) Gulf of Oman core M5-422 changes in dolomite which reflect eolian mineral supply from Mesopotamian sources. Source: H.M. Cullen et al. 2000. Geology 28, 379-382. c) Deuterium changes in sedimentary plant leaf wax δDwax measured as ‰ vs. Vienna standard, as a proxy for East African monsoon strength at Lake Challa (Kenya). Source: J.E. Tierney et al. 2011. Quat. Sci. Rev. 30, 798-807.

Tierney at al. (2011) analyzed in detail the hydrology of Lake Challa, close to Kilimanjaro. One of the proxies they used was the proportion of deuterium in lake sediment plant leaf waxes, interpreted as a proxy for the strength of the East African monsoon. While other proxies indicate Lake Challa did not have low lake levels at the time, δDwax indicates the monsoon decoupled at the time from the total rainfall amount in the local basin. The East African monsoon showed at the time its weakest values in the entire Holocene (Tierney et al., 2011; figure 74 c). The 4.2 kyr event coincides also with a period of great weakness of the Asian monsoon (figure 54 f). The general monsoonal weakness during the 4.2 kyr event must have contributed to its unusual aridity.

We must conclude that the 4.2 kyr event is a uniquely abrupt regional arid event that also caused global cooling. The proxies that show it best do not display a clear periodicity (figure 74), indicating that Holocene climate cycles were not the cause. The very strong monsoon weakening and severe aridification are different to the rest of Holocene cooling events and underscore a primary atmospheric manifestation. Its cause is a complete mystery. Most authors talk about shifts and thresholds in oceanic/atmospheric systems. No big volcanic eruption or asteroid impact capable of such global effect has been convincingly linked to the event, although the abruptness, nature and development of the arid-cold event is compatible with a big tropical volcanic eruption or asteroid impact. Since 1998 soil scientist Marie-Agnès Courty has been defending that soil micro-fabrics bear the signature of a cosmic impact at the time (Courty et al., 2008). However, the lack of more substantive evidence, like iridium, nickel or platinum spikes, or a well-dated crater, has made her research largely ignored.

Whatever its cause, the 4.2 kyr event had a brutal impact on human societies, wiping out the most advanced civilizations at the time and changing the course of history. The world is now 100 times more populated and, despite civilization advances, no less vulnerable to the effects of the changes described. The success of the Akkadian empire was partly due to the sophisticated measures (at the time) they implemented to cope with recurrent droughts in the region. They were just unprepared for the unimaginable scale of what came their way.

Storminess, drift ice and tidal effects

The proposed lunisolar tidal basis for the 1500-year cycle has an outstanding prediction. One of the most salient effects of high tides is that they multiply the water rise due to storms (storm surge). Hurricane Sandy in New York City, 2012, had a storm surge of 4.2 m (14 ft) due to high tides at the time. By analyzing past storminess records we should be able to detect the effect of the 1500-year cycle if indeed it is a tidal cycle. The difficulty is that storms have a random nature and it is necessary to combine multiple records from different locations. Sorrel et al. (2012) analyzed high-energy estuarine and coastal sedimentary records from the macrotidal Seine Estuary and Mont-Saint-Michel Bay in the southern coast of the English Channel and defined five Holocene storm periods that also reflected periods of high storm activity at other northern European locations (Sorrel et al., 2012; figure 75 a). According to the authors, these periods of high storm activity occurred periodically with a frequency of about 1,500 years, closely related to cold and windy periods identified by Bond et al. (2001), and Wanner et al. (2011).

Figure 75. The 1500-year storminess cycle. a) Five Holocene widespread storm intervals defined on the basis of nine independently dated records of northern coastal Europe storm activity. Source: P. Sorrel et al. 2012. Nature Geosci. 5, 892-896. b) Score of the seventeen independent storminess studies for the North Atlantic and Western Mediterranean compiled from the literature in S. Costas et al. 2016. Earth Planet. Sci. Lett. 436, 82–92. c) Figure 10 from Costas et al., 2016 displaying the periods of high storm activity as black boxes for the seventeen studies that are the basis for the score analysis in b. For the identification of the individual studies see the source. DO periodicity is indicated with continuous grey lines. Dashed and dotted grey lines represent harmonics of the DO periodicity. Arrows indicate storminess power at the 750-year harmonic.

The temporal resolution of the Sorrel et al. (2012) findings can be improved if we use more records. Costas et al. (2016), in their SW Europe Holocene windiness study, cite 17 storminess studies from Iceland and Scandinavia to the western Mediterranean (Costas et al., 2016, their figure 10; figure 75 c). By simply scoring these 17 studies I have produced figure 75 b, showing that the storminess cycle has not only a 1500-year periodicity, but the periodicity is coherent with the DO cycle. This semi-quantitative reconstruction shows an increase in storminess levels with time. This increase could simply be due to a better preservation of more recent records or alternatively reflect the effect of the increasingly colder Neoglacial period on cyclone frequency. This interpretation is supported by the high frequency of storms during the Little Ice Age, the coldest period of the Holocene.

The tidal hypothesis for the 1500-year cycle is strengthened by the storminess evidence. As expected, the cycle displays a clear association with storminess intensity. In the tidal hypothesis, as proposed by Berger and von Rad (2002), the 1500-year periodicity is a harmonic of the beat between the moon’s nodal and apsidal precession. This configuration allows the manifestation of power at the other harmonic periodicities, and indeed that is the case, as storminess also displays an increase at the 750-year periodicity (figure 75 b arrows), and the LIA storminess maximum coincides with this half cycle periodicity.

Wolfgang Berger proposed the tidal hypothesis of the 1500-year cycle after studying the varved sediments in an oxygen-minimum zone of the continental slope off the coast of Pakistan in the Arabian Sea. Varve thickness and the presence of turbidites (sedimentary storm deposits) display a large proportion of multiples of the basic tidal cycles of the lunar perigee and the lunar half-nodal. Three of the four longest cycles detected in the 5000-year sedimentary record are 366, 490 and 750 years which are one fourth, one third, and half the 1500-year cycle (Berger & von Rad, 2002). The authors link the tidal cycle to Bond events of increased drift ice through a periodical removal of marine-based glacial ice from the shelves by unusually high tidal waves.

This conjecture might have support from drift ice data off the coast of Alaska obtained by Darby et al. (2012), although the authors appear unaware of Berger’s work. Darby and colleagues assessed patterns of sea-ice drift in the Arctic Ocean over the past 8,000 years by geochemically determining the source of ice-rafted iron grains in a sediment core off the coast of Alaska. They identified pulses of sediment carried by sea ice from the Kara Sea, that display a 1500-year periodicity (Darby et al., 2012; figure 76). The periodicity is coherent and in phase with the DO periodicity (figure 76 b) supporting the same causality. Furthermore, spectral analysis of the data shows not only the 1500-year peak, but also the two lower harmonics (figure 76 a, red ovals). This composite nature of the 1500-year cycle also agrees well with the tidal hypothesis.

Figure 76. The 1500-year cycle in Holocene Arctic sea ice drift. a). Time series analysis of the iron grains, by the maximum entropy method, originated in the Kara Sea and deposited on the coast off Alaska by sea-ice drift (black line) and the Total Solar Irradiance reconstruction (blue line). The dashed curves are the 0.99 confidence limit for both records. A prominent 1.5 kyr cycle is present in the Kara dataset but absent from the TSI. Its harmonics are indicated by red ovals. b). 100-year average of the Kara Sea iron grain-weighted percentage in core JPC16. DO periodicity is indicated with continuous grey lines. Dashed and dotted grey lines represent harmonics of the DO periodicity. Source: D.A. Darby et al., 2012. Nature Geosci. 5, 897-900.

Ending the confusion about the 1500-year cycle

For historical reasons and through researchers mistakes the Holocene 1500-year cycle has been mired in confusion, however as highlighted in this article, the evidence to clarify its timing, cause and effects is already available in the published scientific literature. It is simply awaiting critical reading and integration.

As the glacial DO cycle and the Holocene 1500-year cycle display phase coherence and similar manifestations (subsurface water stratification disruption at the halocline, polar atmospheric circulation intensification, and enhanced Arctic drift ice), it can be assumed that they represent different states of the same cycle, due to the very different climatic conditions at those two periods. Over the last decade the DO cycle has become increasingly questioned (Ditlevsen et al., 2007; Obrochta et al., 2012). The observed DO 1470-year periodicity is based mainly on the original GISP2 timescale. When using the more recent GICC05 age model that is considered superior, the strong 1470-peak weakens considerable and shifts to ~ 1500 years. However, this is consistent with what is observed in the more recent Holocene proxies. The 1500-year cycle has undergone 8 oscillations in the past 11,700 years. If the cycle had a periodicity of 1470 years, instead of the observed ~ 1500 years, the accumulated drift would be of 240 years which is larger than what is observed in the proxy data presented. As the Holocene data is also in agreement with the new GICC05 age model, the DO cycle must actually be closer to 1500 than 1470 years.

While some proxies show the 1500-year cycle taking place regardless of climatic conditions (Sirocko et al., 1996, for the past 19 kyr), other proxies indicate that some manifestations of the cycle became muted or absent during the warmer Holocene Climatic Optimum (figure 70). Although this behavior is also mirrored by the DO cycle, that only displays DO events when climatic conditions (sea levels, obliquity, ice) allow it, some researchers have problems accepting that a real cycle might be responsible. To them such behavior is better explained by the chaotic response to internal variability (Ditlevsen & Ditlevsen, 2009). However, we must accept that the final output of a cycle depends as much on the forcing behind the cycle as on the response to the forcing by a climate system that can be in such different conditions as in a glacial maximum, an interglacial optimum, or a number of intermediate conditions.

When discussing the cause of the 1500-year cycle we must take into account the nature and distribution of the proxies that display it. These proxies are spread all over the world (figure 77), making it very difficult to argue that the cycle might be caused by internal variability or specific oceanic currents.

Figure 77. Global distribution of proxies displaying the 1500-year cycle. Evidence from proxies reflecting oceanic (blue), wind (red), storm (green), tidal (purple) and drift ice (black) manifestations. References for these proxies are given in figures 71-76.

Any hypothesis for the cycle must explain the exact timing and global synchroneity of these events. Authors of the studies on the Northwest Pacific proxies have proposed teleconnections linking the North Pacific Gyre with the North Atlantic through the westerlies, but this explanation falls short since this also requires teleconnections with the Southern Ocean. As there are no inter-hemisphere tropospheric winds, and currents would have a delay of decades to centuries, this seems implausible. The only consistent explanation is an external pacer. This pacer cannot be the sun for multiple reasons. There is no 1500-year solar cycle; the effect on subsurface waters takes place above the halocline, at 50-100 m depth, and even below sea ice during the glacial period. These characteristics are difficult to explain with a solar cause. Further, the cycle can cause either cooling or warming depending on location and conditions.

The lunisolar tidal hypothesis, however, can explain a global synchronous effect and the variety of manifestations, as it enhances both atmospheric and oceanic tides. Unlike a solar cycle, a tidal cycle does not carry a temperature signal, and temperature changes are determined within the climate system depending on conditions. Furthermore, the tidal hypothesis is built over shorter harmonics, and some of those harmonics described by Berger & von Rad in 2002 have been observed (figures 72, 75 & 76; Mayewski et al., 1997; Darby et al., 2012).

One of the problems of the tidal hypothesis is that the cycle is composed of a basic unit of ~ 375 years, while most of the power is displayed at the 1500-year harmonic. Berger & von Rad, 2002, advanced a possible explanation:

…these tidal maxima [have to] occur in the correct seasonal window. The [1500-year cycle] would thus have a plausible mechanism, that is, lunar forcing tied to season.

During the glacial cycle that season would be the summer, when sea ice is at a minimum, while during the Holocene the season could be the winter, when the cooling effect would be maximal. If this explanation is correct it should have two important predictions. The first one is that in the transition from the glacial period to the Holocene, the change to the opposite season should cause a 180° phase shift in the cycle. This is exactly what we find in several proxies (figure 71 a, b & d; figure 73 a; black bars). The other prediction is that if the effect of the 1500-year cycle is tied to the season, it should be opposite in each hemisphere, as the seasons are inverted. Again, that is what is found, as the cycle tied to subsurface water temperatures is inverted between the Southern Ocean and the Northwest Pacific (compare figure 71 b & d).

Regardless of its cause, by knowing its timing, we can analyze the effects of the 1500-year cycle during the Holocene. The Polar Circulation Index reconstruction (Mayewski et al., 1997) suggests in the Arctic region the cycle is associated with increased cooling and more winter-like conditions. North Atlantic proxies, however, support that the cooling effect was muted during the Holocene Climatic Optimum (Debret et al., 2007; figure 70). Comparison of the 1500-year cycle with the Bond series of ice-drift activity in the North Atlantic further confirms the lack of effect during this period (figure 78, downward arrow). It is only during the Neoglacial cooling of the past 6000 years when the cycle is consistent with an increase in drift ice (figure 78, upward arrows) responsible for giving the record its 1500-year apparent periodicity, but indicative that the 1500-year cycle is only contributing to the drift ice record, and not its primary driver.

Figure 78. The 1500-year cycle and Bond events. Holocene record of North Atlantic iceberg activity (black curve) determined by the presence of drift-ice petrological tracers. Source: G. Bond et al., 2001. Science 294, 2130-2136. The 1500-year cycle is represented by grey lines indicating the DO periodicity (continuous lines) or its harmonics (dashed and dotted lines), and by the fitted 1500-year cyclic periodicity (olive green curve). A downward arrow marks the lack of correlation during the first half of the Holocene, while upward arrows suggest some correlation during the last 6000 years.

Unlike in the case of the 2400-year cycle, a Holocene temperature reconstruction (Marcott et al., 2013; figure 37) does not show a clear effect of the 1500-year cycle on global temperatures. If Berger’s tidal, season-linked, hypothesis is correct, global temperatures would see less effect from the cycle than hemispheric temperatures. This is clearly not the “Earth’s unstoppable 1,500-year climate cycle” proposed by Fred Singer (Singer & Avery, 2005).

The 1500-year climate cycle is the millennial cycle whose peak effect is scheduled to take place next, at ~ 2180 AD. What should we reasonably expect from such cyclic occurrence? If our knowledge of the cycle is correct we should see bigger tides and an increase in storm flooding events. There should be an increase in Arctic sea ice and iceberg activity. We should also see an increase in zonal wind circulation and associated precipitation changes. Most of the effects could be smaller than in previous instances of the cycle if global temperatures continue at the current level or increase in the next 150 years. Any decrease in Northern Hemisphere surface temperatures caused by the cycle should be limited, and global average temperatures should not be reduced by much, probably not more than 0.2°C, as suggested by the standard deviation in a global reconstruction where its effect cannot be detected (Marcott et al., 2013; figure 37). It will be however an outstanding opportunity to study the 1500-year cycle and establish the reality of climate cycles at the millennial level, if any observed effect is correctly attributed.


1) The 1500-year cycle displayed in Northern Hemisphere glacial records is also observed in Holocene records from all over the world.

2) The cycle is most prominently displayed in oceanic subsurface water temperatures, Arctic atmospheric circulation, wind deposits, Arctic drift ice, and storminess records.

3) Proxies indicate the cycle also displays power at the 750-year harmonic and might have undergone a phase shift during the Holocene Climatic Optimum.

4) A lunisolar tidal hypothesis currently best explains the cycle’s timing, features, and effects. The hypothesis proposes that the cycle is season linked and thus has different manifestations in each hemisphere.

5) The tidal hypothesis also provides an explanation for the cycle’s lack of a temperature signal, a phase shift observed during the Holocene Climatic Optimum, as well as the apparent opposite response from a Southern Hemisphere proxy.

6) Although the 1500-year cycle is associated with cooling indicators in the Northern Hemisphere, its effect on global temperatures remains to be determined.

7) The next peak effect of the 1500-year cycle is expected in about 160 years, and will provide a rare opportunity to clarify its causes and effects.


I thank Andy May for reading the manuscript and improving its English.

[Bibliography ]

Moderation note:  As with all guest posts, please keep your comments civil and relevant.