Evaluation of the Effect of Vegetation on Urban Heat Islands Using ‎Remote Sensing Techniques in Hillah City

Abstract

Many cities in Iraq are currently exposed a rise in temperatures. Most of the urban forms observed in these cities are not suitable for such harsh conditions, as residential units, buildings and high-rise facilities have been planned and designed in a way that hinders the movement and transfer of air masses between different locations. In addition, the lack of vegetation cover, water bodies, and the use of building materials with high thermal retention capacity exacerbate the problem. In order to investigate one of the factors contributing to the aggravation of the urban heat island phenomenon, it is very necessary to study the role of vegetation cover. This element is of great importance due to its multifaceted benefits, including increasing oxygen levels in the atmosphere and its positive impacts on economic and recreational activities. To identify the most effective vegetation cover area to mitigate air temperature in urban areas, we conducted this study in Babil Governorate in Iraq. We used remote sensing applications to create thermal, vegetation, and image maps within the study area, where we identified several types of land cover, such as vegetation, barren lands, rural buildings, urban buildings, and streets. We excluded water bodies from our analysis because they can affect the relationship between vegetation cover and surface temperature. It is important to note that low vegetation usually leads to high surface temperature in dry areas, and on the other hand, low vegetation leads to low surface temperature of water bodies, which leads to a decrease in the correlation coefficient between vegetation index and surface temperature. A map of normalized difference vegetation index (NDVI) was created for a selected model within the geographical boundaries of Babylon Governorate. In addition, a map of surface temperature (LST) was created for the same area. A model consisting of (386) points was created by integrating heat maps and vegetation data. Then, a table was carefully prepared to include the above values, and correlations were created. Linear, logarithmic and polynomial regression models were evaluated, and the model with the highest correlation coefficient was chosen. We also found an approximate relationship between NDVI and the percentage of green areas, then the two relationships were linked by imposing the target temperature and extracting the approximate green area required to bring the surface temperature to the target level.