An assessment of future Caribbean climate change using 'business as usual' scenario by coupling GCM data and RAMS.

摘要

The Caribbean rainfall season has a bimodal nature, which is divided in the Early Rainfall Season (ERS) and Late Rainfall Season (LRS). To carry out the long-term average conditions Caribbean season analysis, the National Center for Environmental Prediction (NCEP) reanalysis data, the Xie-Arkin precipitation and the Reynolds-Smith Sea Surface Temperature (SST) observed data were used. The Parallel Climate Model (PCM) is evaluated to determine its ability to predict the Caribbean climatology. As a result, PCM under predicts the SSTs, which along a cold advection cause a lower rain production than the observed climatology. The future Caribbean climatological condition simulated by PCM shows a future warming of up to -1°C along with an increase of the rain production during the Caribbean seasons. RAMS was coupled with PCM to asses the dynamical downscale technique. The PCM data used in RAMS as initial conditions has very low SSTs and a stable atmosphere cutting off the vertical convection. To avoid the deviations generated by the PCM output when it is used in RAMS, the 1998 observed data plus the PCM atmospheric variables difference between the years 2048 and 1998 is taken as initial conditions to RAMS. The synoptic scale simulated by RAMS shows a close behavior to the simulated by PCM during the dry season and LRS, while the mesoscale rainfall is strongly influenced by the land dry areas computed in the parent grid.