摘要

We simulated changes in annual maximum 5-day rainfall (RX5D) and annual maximum number of consecutive dry days (CDD) in Central America, Mexico, and the Caribbean with three different horizontal resolution atmospheric global general circulation models (AGCMs) and quantified the uncertainty of the projections. The RX5Ds and CDDs were projected to increase in most areas in response to global warming. However, consistent changes were confined to small areas: for RX5D, both coastal zones of northern Mexico and the Yucatan Peninsula; for CDD, the Pacific coastal zone of Mexico, the Yucatan Peninsula, and Guatemala. All three AGCMs projected that RX5Ds and CDDs averaged over only the land area and over the entire area (land and ocean) would increase. The dependence of RX5D probability density functions on the horizontal resolutions was complex. Precipitation unrelated to tropical cyclones was primarily responsible for the projected increases in the frequency of RX5Ds greater than 300 mm.