8A.2
Fingerprints of the North American Great Lakes on simulated wintertime climate: beyond the regional impact
Xianglei Huang, University of Michigan, Ann Arbor, MI; and Y. Deng
Interactions between regional- and large-scale atmospheric circulations suggest that influences of large inland water bodies could extend beyond the regional scales. We first survey reanalysis datasets (ECMWF and NCEP NRAA)and several IPCC AR4 models with distinctly different treatments of the Laurentian Great Lakes to highlight the sensitivity of the wintertime climatology to the treatments of the Great Lakes. Then we use GFDL AM2 model as a case study to investigate the impact of the Great Lakes on its simulated wintertime climate. Two numerical experiments are carried out: a control run with the standard AM2 configuration and a lake run with six “Great-Lakes” gridboxes. Statistically significant differences exist in the simulated winter mean surface air temperature (SAT, ~1.2.-3.5K) over vast downstream land area. The probability distribution function of lake-run daily SAT agrees well with that of ECMWF ERA40 reanalysis while the control run has excessive occurrences of cold SAT. The SAT difference is consistent with lake-induced thermal low anomalies over the downstream area, of which the influence can be felt beyond the regional scale. It significantly increases the occurrences of strongly negative daily Arctic-Oscillation index. The source of the differences in the simulated AO between the control- and the lake-run is then further diagnosed through one-point lag-correlation and Rossby wave vorticity source anomaly analysis. The dynamical nature of such differences, including their potential interaction with winter planetary wave, is further examined in an idealized general circulation model (GCM). Implications of these findings on projecting future climate change, especially the regional climate changes, are further discussed.
Session 8A, Climate Change Modeling Part I
Wednesday, 20 January 2010, 1:30 PM-2:30 PM, B215
Previous paper Next paper