Eleventh International Conference on Climate Change: Impacts & Responses

Abstract

The aim of this paper is to present a more critical evidence of man’s (modifying) activities in an urban landscape. Hence, the study deployed geospatial techniques to model the LST of Onitsha Metropolis in south-east Nigeria based on the dynamics of urban biophysical descriptors via a geospatial perspective. Data acquired for the study comprised of Landsat-8 satellite imagery of 2016 and the Advanced Spaceborne Thermal Emissions and Reflections Radiometer’s Global Digital Elevation Model (ASTER GDEM) data from appropriate authorities. LST was extracted from Landsat thermal bands while eighteen (18) land biophysical descriptors which comprised of Latitude and Longitude; aspect, slope and elevation; Linear Spectral Mixture Analysis (LSMA) of LULC categories of vegetation, impervious surface, soil and water endmember fractions; and nine (9) other land cover indices computed via image thresholding: UI, NDBI, NDVI, SAVI2, EVI2, NDWI, MNDWI, NDMI and MSAVI. Principal Component Analysis (PCA), Pearson Correlation test and step-wise multiple regression were used. The study was carried out using ArcGIS 10.5, Idrisi Terrset and the Statistical Package for Social Sciences (SPSS 25) environments were presented in tables, graphs and maps. NDBI, Unmixed soil fraction, slope, elevation and locational descriptors were found to be most significant and thereby impactful in modelling LST in the study area, with a coefficient of determination (R) of 0.86 and adjusted R2 greater than 0.73. Meanwhile, the standard residual map of LST suggests other variables (e.g. proximity to road, waterbody, urban geometry, etc.) that could be included to further increase the efficiency of the model.