Modeling changes trend of time series of land surface temperature (LST) using satellite remote sensing productions (case study: Sistan plain in east of Iran)

To assess the long-term variations in the trend of the mean monthly land surface temperature, LST products from the MODIS sensor on Terra satellite (MOD11A1) were employed. A total of 1365 daily images of LST MODIS were provided from the Land Processes Distributed Active Archive Center for April (450 images), May (465 images), and June (450 images) months. Then the mean monthly land surface temperature (LST) was estimated separately using daily land surface temperature (LST) time series for 3 months, April, May, and June, as well as for the total statistical period studied. A total number of 30,080 pixels were investigated within the studied area according to spatial resolution 1 in 1 km each image. Finally, the long-term variations trend of the mean monthly land surface temperature was evaluated using Sen’s slope estimator and classic linear regression on a pixel basis. According to the results, there was no significant difference in both models about the estimation of long-term variations trend of the mean monthly LST and they had the same performance. The results of this research also showed that the highest increasing trends in LST was observed in the north of the plain, which may be due to the drying up of Hamoun triple wetlands. However, the highest decreasing trends in LST was observed in the northeast, southwest, and west of the study area, which may be attributed to the expansion of agricultural lands and gardens in the northeastern part of the plain (along the main branch of Hirmand River) and digging of small wells in the southwest and west of the plain by indigenous peoples as well as the use of supply water for farming.