Estimation of land surface temperature using NDVI in MODIS and Landsat ETM+ imageries

Land surface temperature is one of the main parameters required for hydrological elements estimations in regional scale. This variable can be measured using thermal band of satellite sensors. In this research ETM+ and MODIS imageries of downstream end of Qazvin irrigation network of Qazvin were used. Four different scenarios were proposed and the obtained results were evaluated by volumetric lysimeter measurements. Results indicated that if water bodies are removed from satellite imageries، a reverse relation can be observed between NDVI crop index and land surface temperature in both ETM+ and MODIS imageries. Although، non-linear relation has a good result in compare to linear، no significant differences were observed between them. Using simple aggregation method produced acceptable result in compare to the aggregate nearest neighbour method. Finally، cells with dimensions following f (n)  602، n 1،2،3،4،… n rule after integration، produced the best results. Among recommended scenarios، initial segregation of lands and application of proposed algorithm in lands with vegetation cover produced the best coincidence between land surface temperature data and vegetation index (NDVI).