How good are the simulations of tropical SST–rainfall relationship by IPCC AR4 atmospheric and coupled models?

Journal of Earth system science

K Rajendran1²  Sulochana Gadgil3¹  J Srinivasan2 3³  Ravi S Nanjundiah2 3³ 

[1] Centre for Atmospheric and Oceanic Sciences, Indian Institute of Science, Bangalore 560 012, India.$$;Centre for Mathematical Modelling and Computer Simulation, Bangalore 560 037, India.$$;Divecha Center for Climate Change, Indian Institute of Science, Bangalore 560 012, India.$$

关键词: SST;    rainfall;    AGCMs;    CGCMs;    local and remote impacts.;   

DOI  :  

学科分类：天文学(综合)

来源: Indian Academy of Sciences

【 摘 要 】

The failure of atmospheric general circulation models (AGCMs) forced by prescribed SST to simulate and predict the interannual variability of Indian/Asian monsoon has been widely attributed to their inability to reproduce the actual sea surface temperature (SST)–rainfall relationship in the warm Indo-Pacific oceans. This assessment is based on a comparison of the observed and simulated correlation between the rainfall and local SST. However, the observed SSTconvection/rainfall relationship is nonlinear and for this a linear measure such as the correlation is not an appropriate measure. We show that the SST–rainfall relationship simulated by atmospheric and coupled general circulation models in IPCC AR4 is nonlinear, as observed, and realistic over the tropical West Pacific (WPO) and the Indian Ocean (IO). The SST–rainfall pattern simulated by the coupled versions of these models is rather similar to that from the corresponding atmospheric one, except for a shift of the entire pattern to colder/warmer SSTs when there is a cold/warm bias in the coupled version.

【 授权许可】

Unknown