Late Quaternary Climate Change in Eastern North America:
A Comparison of Pollen-Derived Estimates
with Climate Model Results

Webb III, T., K. H. Anderson, P. J. Bartlein and R. S. Webb

Quaternary Science Reviews Volume 17 Issues 6-7, pages 587-606 (1998)

Abstract      Figures (unavailable)     Full Text (unavailable)

Abstract

Late Quaternary pollen data from eastern North America and pollen-climate response surfaces provide tests of the climate simulations from Version 1 of the National Center for Atmospheric Research (NCAR) Community Climate Model (CCM1) for 21, 16, 14, 11, and 6 ka. The model results are also compared to those from Version 0 of the NCAR model (CCM0). In contrast to CCM0, CCM1 used a slab ocean model to compute sea surface temperatures, included a seasonal cycle, and computed soil moisture. It also used an improved set of boundary conditions from 21 to 6 ka. In eastern North America, CCM1 simulated lower temperatures at the last glacial maximum (LGM) 21,000 years ago than did CCM0 and therefore was in better agreement with the pollen data than the CCM0 simulations were. The simulations by CCM1 for mean July temperatures from 16 to 11 ka, however, were much higher than those by CCM0 and were higher than present for much of the area south of the Laurentide ice sheet. These simulated conditions differ markedly from those reconstructed from the pollen data, which indicate July temperatures significantly lower than present from 16 to 11 ka. At 6 ka, CCM1 simulated climate conditions similar to today and therefore not too different from those simulated by CCM0 or those inferred from the pollen data. For 6 ka in the upper Midwest, however, CCM1 simulated moisture conditions similar to present, which is an improvement over CCM0. However, the new simulation still does not match the drier-than-present conditions inferred from the pollen data. Detailed analysis of the circulation variables and the surface-energy-budget terms simulated by CCM1 show that both dynamic and thermodynamic factors contributed to the major discrepancy in July temperatures from 16 to 11 ka. The simulated glacial anticyclone created less cloudiness and rainfall south of the ice sheet, and these conditions allowed the increased summertime insolation from 16 to 11 ka to overheat the surface. The impact of the ice sheet on model dynamics therefore led to overestimation of summer temperatures south of the ice for this time period.

Figures

Click below to enlarge this figure.

Figure 1
72dpi (21K), 300dpi (163K)

Figure 1. The...