The Effects of Land Cover Changes on Land Surface Temperature in Amman; an Urban Climate Change Study

Abstract

In this study, remote sensing techniques were used to examine the relationship between land cover change and land surface temperature in Amman city (the area of study). Landsat TM5 datasets for the years 1987 and 2003 , landsat 8 for the year 2014 were used in the analysis using PCI Geomatica 9.1 to perform supervised classification, MS Excel to calculate the changes in areas, ENVI 4.5 to perform land surface temperature estimation and ArcGIS 9.3 for overlap analysis of the results. Four classes were classified within the study area: Built-up area, Vacant Land, Vegetation and Quarries & Restructured Areas. An accuracy assessment was conducted for each data set and the statistics show that the overall accuracies are 92% and 86% and 88%, and the overall Kappa statistics were 0.872 % and 0.783% and 0.823 for the images of 1987 and 2003 and 2014, respectively. The overall land surface temperature ranged from 19°C to 29°C with a mean of 23.9°C and with a standard deviation of 3.0°C for the 1987 image and it ranged from 20°C to 32°C with a mean of 26.2°C and with a standard deviation of 4.1°C for the 2003 image. where it ranged from 20°C to 33°C with a mean of 26.5°C and with a standard deviation of 4.4°C for the 2014 image. The overall mean temperature was also increased by 3.2°C. LST values were increased for each land cover type in the 2014 image compared to the 1987 image, except for the Vegetation. The mean of temperature values for the built up area was increased by 4°C, and by 1.7 °C for Vacant Land and by 0.1°C for Quarries & Restructured Areas. . The mean for temperature values for the Vegetation areas were decreased by -1.7°C. These results showed strong evidence and relationship between the changes of land cover types on the local temperature of the city.