Magnitude of hydroclimatic variance in paleoclimate records and in multi-millennial simulations of two Earth System Models

Thursday, 18 December 2014
Luke Alexander Parsons1, Jonathan T Overpeck1, Jianjun Yin1, John P Krasting2, Sergey Malyshev3 and Ronald J Stouffer4, (1)University of Arizona, Tucson, AZ, United States, (2)NOAA / Geophysical Fluid Dynamics Laboratory, Princeton, NJ, United States, (3)GFDL, Princeton, NJ, United States, (4)US Govt NOAA, Princeton, NJ, United States
We compare the magnitude of hydroclimatic variability in instrumental and paleoclimate records with long simulations from two state-of-the-art Earth System Models developed at GFDL (ESM2Mb and ESM2G). Instrumental and paleoclimate data indicate that hydroclimate varies more strongly at multidecadal to millennial time frequencies than at interannual frequencies. Multi-millennial control runs in both ESM2Mb and ESM2G reproduce strong temperature variance at multi-century to millennial scales in the Northern Hemisphere (20°N-90°N). However, ESM2Mb does not reproduce strong multi-decadal to millennial precipitation variance in the Northern Hemisphere, whereas ESM2G captures strong multi-century scale precipitation variance in the Northern Hemisphere. In the tropics (20°S-20°N), these models do not reproduce strong multi-decadal to millennial precipitation and temperature variance as compared to interannual variance. Our findings suggest that the latest IPCC AR5 climate models are likely to underestimate the background risk of multidecadal hydroclimatic variability (drought). Different representations of ocean processes in ESM2Mb and ESM2G can change the variance of hydroclimatic variability on interannual to millennial time scales.