The importance of using a high-resolution model to study the climate change on small islands: the Lesser Antilles case

Abstract

High-resolution climate change simulations over the Lesser Antilles are performed using the ALADIN-Climate regional climate model nested within the global model ARPEGE (Météo-France). Three sets of simulations are conducted at 10 km grid spacing for reference (1971–2000) and future climate (2071–2100) under two CMIP5 scenarios (RCP4.5 and RCP8.5). With the dynamical downscaling, islands of Lesser Antilles are considered as land by the model, whereas, for the driving model, there is only sea over the domain. Temperature and precipitation change are analysed on land and on sea separately. For temperature, the warming is greater on land than on sea, especially for the minimum daily temperature (3.2°C vs. 2.3°C for the RCP85 scenario). For precipitation, projections are less reliable because the seasonality is not well reproduced by the model. Nevertheless, simulations exhibit the fact that projections on land differ from one island to the other and disagree with those on sea notably during the wet season. This underlines the importance of the dynamical downscaling to study the climate on small islands. Statistical downscaling has been performed on the Guadeloupe Island to study changes in extreme precipitation indices. The projections provided by the regional climate model suggest an increase in extreme rainfall events: longer dry periods, a bigger annual total precipitation, more frequent very heavy daily precipitation and a stronger 1 d maximum precipitation, whereas for the driving Global Climate Model, these trends are less intense.

Keywords

• climate change
• regional climate model
• RCP scenarios
• Lesser Antilles
• temperature
• precipitation