干旱区生态环境知识资源中心

The present study focuses on the evaluation of dust in on-line chemistry-climate simulations with the regional climate model RegCM4. The simulations cover the period from December 2006 to November 2014 and use the new dust scheme implemented in RegCM4 that splits 12 dust size groups according to the changes of dry deposition velocity with size (isogradient method). In the model, dust is emitted over desert and semi-desert grid cells when a wind speed threshold value is exceeded. This critical wind speed value varies and depends on soil particle size distribution, surface roughness and soil moisture. The simulated dust optical depth is evaluated over the Sahara and the Mediterranean with the pure dust product LIVAS derived from CALIPSO satellite observations (http://lidar.space.noa.gr:8080/livas/). Furthermore the evaluation results are discussed with respect to the comparison of the simulated wind speed and precipitation fields in our domain with the ERA-interim reanalysis and CRU observational gridded datasets. We combine the simulated emission and deposition fluxes and present their annual cycle to explain the biases we observe in the dust optical depth. Finally we discuss thelimitations of climate models to accurately simulate the production/destruction processes of dust and propose possible ways to improve RegCM4 simulated results.