reservoir water temperature modeling by geostatistical analysis of aster images (case study: dez dam, iran)

satellite data have been used for temperature modeling both in urban areas and water. this paperstudies viability of aster satellite images that provide high spatial and spectral resolution to model watersurface temperature, as a fundamental water quality parameter. this study is focused on dez dam reservoirresiding in khuzestan province, south west of iran. after the image corrections required for the asterimage, the ndwi was determined. water pixels extraction was done using the ndwi equal to 0.88. usingmultiple linear regression analysis (mlr), water temperature model was retrieved by a recursive approach.the suggested model estimates the water surface temperature measured at ground control stations data.a high r-square of 0.87 was observed between the control station data and modeled temperature. aftervalidation test of the temperature model at ground control station, the validity of the interpolated data in longdistance was assessed. the spatial auto-correlation (moran-i) and clustering analysis (hot spot) of the resultalso show that while long distance interpolation in the inner area of the dam reservoir seems acceptable,different interpolations are required at shores of the dam reservoir and at river outlets to model water surfacetemperature of the dam reservoir.