Information Systems for Water Resources
Fusion of Less Revisit but High-resolution Images and Many Revisit but Low-resolution Images for a Both Spatially and Temporally High-resolution Areal Crop Evapotranspiration
Wednesday, January 4
1:30 PM - 3:00 PM
Location: 5th Meeting Room
The per capita water resources of China are only about 28% of the world average while the spatio-temporal distribution of water resources is extremely uneven in China. Therefore, water scarcity is an extremely critical issue in many regions of China. Monitoring of regional crop evapotranspiration (ET) is very important for improving the efficiency of agricultural water resources. Traditional monitoring of crop evapotranspiration based on ground stations provides accurate, temporally continuous but spatially sparse observations, whereas remote sensing images obtain less accurate, due to hybrid pixels, temporally non-continuous but spatially relatively continuous observations. The major research work will be fusion of less revisit but high-resolution images and many revisit but low-resolution images for creating a both spatially and temporally high-resolution areal reversion model of areal crop evapotranspiration.
In this paper, we will makes full use of the advantages of the temporal resolutions of low-spatial-resulution images and spatial resolutions of low-temporal-resolution images to construct a regional evapotranspiration model. The model parameters in mixed pixels become confusing in estimates of ET using lower spatial resolution data. For example, land cover (land use or land class) is an important parameter in the ET modeling, but in mixed pixels, the land cover of the whole pixel is usually the one that has the maximum area ratio in that pixel. This could result in error in the other parameters related with it and in the estimated ET. In this paper, the high resolution land cover parameter is derived from Landsat TM data, and used within the algorithm of ET by MODIS data. In this paper, we extract high-resolution crop classification and vegetation index information from remote sensing images, and design a method processing subpixels for low spatial resolution images, which can improve the spatial resolution of crop evapotranspiration.