Spatiotemporal Variability of Surface Urban Heat Island Intensity in Kumasi from 1986 to 2022
Description
Research Hypothesis. The study hypothesised that biophysical variables such as normalised difference vegetation index (NDVI) normalised difference vegetation index (NDVI), normalised built-up index (NDBI), dry barrenness index (DBSI), and elevation affect the intensity of urban heat island effect. Data Representation. The study was divided into two using a 15km buffer from the urban core. The division of the study area into urban and rural helps with the segregation of the extracted data. This enables the study to analyse the temperature difference between urban and rural areas. Data was collected from United States Geological Survey: Landsat and Modis respectively for three years. The digital value numbers of the satellite data were extracted using Arc GIS tools. Thus, the processed values of the satellite data depict the pixel values of the data extracted. These values give a representation of temperature variance between urban and rural areas. Findings. Notable findings include the determination of progressive increase in urban surface heat intensity from 1984 to 2022 for both Landsat data and Modis data. Also, biophysical variables: NDVI, DBSI, and NDBI resulted high positive variation with land surface temperature.
Files
Steps to reproduce
1. Acquisition of Data from United States Geological Survey 2. Processing Landsat data for Land Surface Temperature (LST) , NDVI, DBSI, NDBI Arc GIS 3. Processing Landsat data for Land Cover and Land Use analysis using Arc GIS 4. Processing Modis data for Land surface temperature using Arc GIS 5. Set buffer from urban core using Arc GIS buffer tool to aid creating urban and rural regions. 6. Using Fish Net tool in Arc GIS to create random points for the extraction of values from the processed LST, NDVI, DBSI, NDBI and UHI. 7. Analysing the extracted data with excel and tableau for correlation analysis. A critical guide is provided in the workflow in the main paper.