Estimation of instantaneous air temperature using remote sensing data

In arid areas, the variation of air temperature can be considerable, so instantaneous air temperature (T-ai) estimation is needed in different environmental researches. In this research, two different remote sensing data are used for estimating T-ai for clear sky days in 2009 in Fars Province, Iran, including atmospheric temperature profile and land surface temperature (LST) data from Moderate Resolution Imaging Spectroradiometer. The T-ai from a number of surface weather sites is used to judge the best T-ai estimation. Stations' elevation, latitude, and land cover type are considered to show their effect on T-ai estimation. The estimated T-ai evaluation focuses on daily and seasonal timescales in the daytime and night time separately. Both LST and vertical temperature profile data produced relatively high coefficient of determination values and small root mean square error value for T-ai estimation, especially during the night time. Land cover and elevation vary the error values in T-ai estimation more, when LST data is used. In comparison atmospheric temperature profile indicates a smaller error in T-ai estimation in spring and summer and in urban land cover type, while using LST data presents a better result in fall and winter especially at night time.