Part I introduced the challenges of climate prediction on decadal scales, specifically in the context of global climate model simulations. On the Part I thread, Paul_K writes: