ABSTRACT
The study was carried out to investigate the prevalence of Urban Heat Island in Kaduna metropolis, Kaduna State, Nigeria. Long-term point-based periodic mean annual temperature and total precipitation was collected from the Nigerian Meteorological Agency (NiMet), Abuja. To streamline the analysis, the monthly data for precipitation and temperature were collapsed into annual bases (total annual precipitation and mean annual temperature). Data was collected for the period of 1950-2018 and Land Use/Land Cover (LULC) was derived from Landsat imageries for the relevant dates. Landsat imageries were chosen based on the fact that its moderate spatial (30 meters) and temporal (16 days) resolution sufficed for the study. The required Landsat data sets were downloaded from USGS EROS Data Center for 1990, 2001, 2010 and 2018 respectively. Radiometric and geometric corrections were applied to initially uncorrected Landsat images. Dark Object Subtraction-1 (DOS-1) technique designed for atmospheric correction which ensured that water body and shades are referred to as dark body when they have indices of reflectance that is almost zero was used for all the images. GIS-based digital image classification was deployed in order to examine the various classes of land use/land cover in Kaduna metropolis (1990-2018). Land use/cover was categorized into four classes based on the acquired Landsat imageries. Images were classified using the supervised classification and unsupervised classification. The validation classified image was used to produce an error matrix with which the accuracy assessment was carried out in other to unveil the extent to which the dataset relates to the validation data. The Normalized difference vegetation index was used to establish the difference between areas with healthy vegetation and the non-vegetated areas. Mann-Kendall trend test was used in the trend analysis of the study. Spearman’s rho test was used for the determination of significance in the correlation between variables. Land surface temperature was estimated using GIS-base Zonal Statistics. Land cover land use classification were processed and estimated from the relevant bands of the Landsat scenes. To take advantage of the information in all the available bands, the different bands were composited into a single image to become a multi band imagery which allowed both visualizing and analyzing the bands simultaneously. Forest, grass/sparse vegetation, water and built-up/bare surfaces were the four (4) land use/land cover classes of land use/land cover that were classified using the Landsat TM 1990, 2001, 2010 and 2018 imageries. The overall image classification accuracy (ICA) and Kappa coefficients (KC) were above 85% and 0.85 showing that all the remotely-sense products were satisfactorily classified.
The spatial pattern of Land Surface Temperature (LST) exhibited a typical Urban Heat Island (UHI) pattern with the highest mean LST recorded in the built- up/bare surface areas. LST decreased with increasing distance to the urban periphery and to the hinterland. The first research hypothesis which states that “the spatial pattern of LST in the study area is random within the years (1990, 2001, 2010 and 2018)” was rejected given that all the z-scores and p-values were positive and significant at p<0.001 respectively. Also, the second, third, fourth and fifth research hypotheses were rejected at 95% confidence level.This research concluded that land cover type primarily affects land surface temperature in Kaduna metropolis. The study also confirms that surface temperature rises as built-up spaces increases in the metropolis. The absence of vegetation preponderance of impervious non-evaporative surfaces explains the large expanses of high temperatures as observed especially in the built-up part of the study area.
Therefore, attitudinal shift, appropriate urban planning strategies, legislation, public enlightenment, to make for deliberate increase in vegetal cover, use of urban infrastructures with high reflectivity and low heat absorption capacity and the funding of cutting-edged researches on urban thermal environment are recommended.
