Influence of changes of underlying surface on the output of sand-dust model

Underlying surface data used in a sand and dust model is critical to its output results. In this paper, investigation of effects of the used underlying data on its outputs was carried out by using GRAPES_SDM to simulate a severe dust storm process occurred from 23 to 25 April, 2010 in Minqin, China with latest and original underlying data for sensitive and control test, respectively. The results show that changes of land cover results in changes of model outputs in extent and intensity of surface dust emission and dust concentration and its spatial distribution. However, the effects on dust concentration are primeval and ambulant. It also displays that the stimulation top dust concentration value obtained by the updated data is 5200 µg/m3, which corresponds better to the observations (6000 µg/m3) than that of the original one (7700 µg/m3), indicating higher precision.