Abstract
Urbanization and human activity within an urban system produce many destructive and irreversible effects on natural environment such as air pollution and climate changes. One of the important effects of climate change is the formation of surface urban heat island (SUHI) which is an area with higher temperature than surroundings. It is important to study the surface urban heat islands to understand the complexity of the climate systems and to lessen their impact on the environment. In this paper, an approach for detecting SUHIs based on the combination between a set of Landsat 8's Thermal Infrared Sensor (TIRS) night vision images and Spot5 data was proposed. To accurately detect SUHIs over Jeddah City, it is important to determine the land surface temperature (LST). To achieve this goal, pixel values of Landsat images were converted to represent at sensor temperature. The spot image was classified using supervised classification techniques to determine feature types in the scene, the emissivity value for each pixel was assigned using classification-based emissivity and NDVI-based emissivity. Then, the two values of at sensor temperature and feature emissivity were linked together to retrieve an accurate LST. Based on the results of this study, the SUHIs over Jeddah City appeared as small boundaries in the South area of the city, as a result of the land use patterns. The difference between urban and non-urban areas ranges from 4 to 7 degrees C. The SUHIs over Petromin neighborhood and Almohajer neighborhood were presented. Night vision Landsat 8 offers an effective framework to delineate and monitor behavior, movement, and size of SUHIs. The early detection of SUHIs by remote sensing data contributes in discovering environmental imbalance and helps to identify problems and developing solutions.