جستجوی مقالات مرتبط با کلیدواژه « جزایر حرارتی شهری » در نشریات گروه « منابع طبیعی »

The purpose of this study was to assess the impact of the land use changes in the Minab city on the variations in the urban heat islands (UHI), using the random forest classification method and spatial statistics on TM and OLI Landsat images in 1988, 1998, 2008 and 2018. First, land surface temperature (LST) was calculated using the split-window and the single-channel algorithms. Land use map was generated using supervised classification random forest method and its performance was assessed by two criteria of overall accuracy and kappa coefficient. In order to survey spatial autocorrelation and clustering, pattern of hot spots, the two statistics of Anselin Local Moran's I and Getis-Ord GI were applied. In 1988, land uses of vegetated, barren, and urban built-up lands were occupied 30.1, 32.53 and 37.37 percent of the city area, respectively; in 2018, the areas of these land uses were 16.36, 9.56 and 74.08 percents, respectively. A threefold and twice-fold decrease in the area was observed for barren and vegetated lands, respectively; while the area of urban built-up lands had more than doubled. The calculated values for urban-heat-island ratio index (URI) in 1988, 1998, 2008, and 2018 were 0.45, 0.34, 0.11, and 0.22, respectively. The outcomes of two considered spatial statistics indicated the clustering, pattern for UHI of the Minab city. In addition, there was a good agreement between the results of Getis-Ord GI statistic (hotspots spatial analysis) and the Local Moran's I statistic (spatial autocorrelation) on the spatial pattern of heat and cool clusters.

Land surface temperature is a key indicator of energy balance. Besides, it serves as input data for models of climate change, agriculture, meteorology, urban heat islands, choosing the best time to agricultural activities, study of volcanic and geothermal activity, and fire detection. In this study land surface temperature has been extracted by available methods using 4 images of TM and ETM+ sensors of Landsat in span years of 1985 to 2003. The methods of land lurface temperature extraction included landsat project science office, mono window, SEBAL, Stefan-Boltzmann and single channel. Because of the multiplicity of methods and the number of images used in this study using a statistical method is required. It is required to determine the most efficient extraction method of land surface temperature, which is close to the existing field data. The statistical indicator used in this study was a mean absolute error (MAE). The results indicated that Stefan-Boltzmann method was the best method for both TM and ETM+ sensors. The MAE values for TM and ETM+ were 4.3 and 6.8 respectively, which showed a minimum value among other results.