Monitoring And Estimating The2010Drought Impact In Southwestern China Based On Remote Sensing

Along with global warming, natural disasters, especially drought, increased bothfrequency and intensity. As one of the most devastating disasters, drought not only have adirect effect on human’s production and living, bringing immense damage to social economy,but also greatly impact the terrestrial ecosystem. Drought will decrease vegetationphotosynthetically activity, increase the possibility of fire occurrence, induce disease andinsect eruption, and finally lead to tree mortality. The study on drought will not only bebeneficial to the human’s production and living standard, but also promote ecosystems todevelop healthily.Southwestern China has experienced several severe drought in autumn and winter eversince2000, however, the drought happened in2009to2010reached a peak due to both spatialspan and duration. This drought not only caused a dry up in many river and lakes, which leadto a water shortage for people, but also induced large area of vegetation growth decline insouthwestern China region, which had a significant impact on terrestrial ecosystem carboncycle.In our research, we applied various kinds of data including remotely sensed data fromsatellites, and gauge data from weather station. After standardized, we analyzed droughtspatial distribution and vegetation response to the drought. These majorly expressed as theanomaly in Enhanced Vegetation Index (EVI), Gross Primary Production (GPP), fireoccurrence, land surface temperature, and precipitation from both radar and weather station.After analyzing these data comprehensively, we find that anomalies in precipitation, landsurface temperature, and fire occurrence coincide with each other, majorly concentrate insouth Sichuan, Yunnan, Guizhou, and northwestern Guangxi. However, for the vegetationresponse, EVI and GPP declined significantly in Sichuan Basin and Guangxi besides theabove region. The spatial inconsistency may be caused by the poor quality of the remotelysensed data. After evaluating the quality of EVI and GPP data in southwestern China, we findthat most region in Sichuan Basin, Guizhou, and Guangxi did not have a good data quality, majorlycausedby aerosol in these regions. Thicker aerosol load cannot be removed throughthe traditional radiation transmission model. Therefore, the corrupted data need to be removed.After these most affected data are eliminated, most of the western part of our study area didnot have sufficient data to characterize the change. Therefore, the previous vegetation growthdecline may be unreliable. However, even after these corrupted data are excluded, spatialinconsistency still exist in different satellite’s EVI and different versions of GPP data, and thereason may be numerous. After a correlation and path analysis for MODIS GPP with otherfactors, we find that MODIS GPP C55data has the least correlation with other droughtindicators and therefore may not be good to explain the drought impact for vegetation.