Drought Monitoring Method Based On MODIS And Uncertainty Data

Relatively, drought is one of the largest natural disasters in China and the world. Mainly because of its multi-frequency, long duration, wide scope, and besides, the heavy losses to economy sectors such as agricultural, the potential impact on water resources, land resources and the community. Therefore, drought gains more and more attentions from governments, scientists, and the public of all over the world. Dynamic monitoring of the drought, timely and accurate reflection of the drought, and development trends to provide a scientific basis for the anti-drought measures to governments at all levels, is of great significance.Traditional Drought Monitoring is to use soil moisture of sparsely scattered point to monitor the extent of the drought. The representation of this method is poor and can not realize large-scale drought monitoring. Satellite remote sensing data because of the macro, dynamic, objective, timely, good features, provides an efficient and convenient platform for large-scale drought monitoring. Among them, MODIS data, with its low-cost, multi-spectral resolution, high time resolution and moderate spatial resolution, has outstanding advantages in dynamic large-scale drought monitoring.China's arid northwest, the semi-arid area pay special attention to water resources, because its basic crops rely on irrigation to grow. Thus, Drought monitoring the irrigated areas, is an effective way to implement irrigation reasonably, improve water use efficiently in irrigated areas and to save water in agriculture. This thesis, based on the analysis of existing theories and technologies, using Shule River Irrigation District in Gansu for the study area, implement drought monitoring in two aspects: the extract drought indicators and the relatively arid ratings.In this thesis, a lot of MODIS data preprocessing work to ensure that data availability and effectiveness. Then, extract three drought indicators, that does not need to rely on the ground data, to study a ten-day relatively arid extend of the region. Extract max NDVI and a ten-day max Brightness Temperature of a ten-day, using the largest-synthesis-algorithm, and then casual-point temperature correlation analysis in order to determine Vegetation Temperature Condition Index, and the drought index level evaluation; To facilitate business, using brightness temperature instead of land surface temperature, to calculate the max difference temperatures between day and night, and other drought indicator- Vegetation Support Water Index. Because of the applicable limitations of itself, single indicator of drought ratings is not ideal, and even mutually contradictory to each other. In light of this uncertainty of this evaluation, this thesis provides a fuzzy comprehensive method to evaluate these two indicators of drought index. Comparison and analysis of three different drought assessment shows that fuzzy comprehensive approach makes satisfactory results, with practical value.