Simulation and Sensitivity in a Nested Modeling System for South America. Part II: GCM Boundary Forcing

Abstract

A regional climate model driven at the lateral boundaries by ensemble integrations of a general circulation model (GCM) is used to investigate 1) the large-scale circulation anomalies associated with tropical sea surface temperatures (SSTs) that lead to extreme rainfall anomalies during January?May of 1983 (dry) and 1985 (wet) in tropical South America, and 2) the sensitivity of the nested model results to the choice of domain. The nested model is composed of the Regional Climate Model (RegCM) and the Community Climate Model version 3 (CCM3), both developed at the National Center for Atmospheric Research. In Part I of this study, the RegCM's ability to simulate circulation and rainfall observed in the two extreme seasons was demonstrated when driven at the lateral boundaries by reanalyzed forcing. Seasonal integrations with the RegCM driven by GCM ensemble?derived lateral boundary forcing demonstrate that the nested model responds well to the SST forcing, by capturing the major features of the circulation and rainfall differences between the two years. The GCM-driven model also improves upon the monthly evolution of rainfall compared with that from the GCM. However, the nested model rainfall simulations for the two seasons are degraded compared with those from the reanalyses-driven RegCM integrations. The poor location of the Atlantic intertropical convergence zone (ITCZ) in the GCM leads to excess rainfall in Nordeste in the nested model. An expanded domain was tested, wherein the RegCM was permitted more internal freedom to respond to SST and regional orographic forcing. Results show that the RegCM is able to improve the location of the ITCZ, and the seasonal evolution of rainfall in Nordeste, the Amazon region, and the southeastern region of Brazil. However, it remains that the limiting factor in the skill of the nested modeling system is the quality of the lateral boundary forcing provided by the global model.