DEVELOPING URBAN ENVIRONMENTAL CLIMATE MAP FOR SELECTED NORTHERN NIGERIAN CITIES

Abstract

Rapid urbanisation in developing countries is becoming issue of great concern, where high-density or mega-cities becoming the order of the day. Public concern and climate change when high-quality urban living is involved, one of the most important task is sustainable climatic spatial planning and planners and policymakers have challenges entering an era of mega-urbanisation. This study provided a social science-based approach for developing an UECM for heterogeneous cities, using three states (Kano, Kaduna and FCT) in Nigeria. Landsat 8 (OLI) images of 2018 and Landsat 4 (TM) images of 1988 covering the entire study area were collected, clipped with the shapefiles of each states and classified using the nearest neigbourhood analysis and forest classifiers to generate the LULC and LCZ maps of the study areas respectively. Vegetation indices (NDVI), built-up indices (NDBI) and land surface temperature (LST) were also calculated and daily temperature data form 1988 to 2018 (30 year period) were analysed using R software package. On the final note, a total of seven input layers were developed along with a final evaluation UCAM, together with URM, and were implemented into a system of local urban planning. The result of two classification schemes showed the morphology characteristics of rural and urban area in the three states and also detected the potential Urban Heat Island (UHI) distribution pattern across the states. The results of the accuracy assessment in each state also showed satisfying result as LCZ Kappa/overall accuracy includes 0.96/84.69%, 0.99/88.76& and 0.99/78.12% for Kaduna, FCT and Kano respectively, while the LULC Kappa/overall accuracy includes 0.81/89.85%; 0.79/89.90% and 0.86/88.57% respectively for Kaduna, FCT and Kano. The results of Pearson correlation coefficient analysed in SPSS software measures strength and direction of association that existed between the variables and indicated a perfect inverse relationship between LST and NDVI with r value of -1, while indicating a perfect relationship between LST and NDBI with r value of +1. Time plots of temperature observations showed indications of trend, seasonality and randomness in the patterns; hence decomposition of additive time series analysis were carried out and these revealed the pattern of average monthly temperature over the years under consideration (1988 – 2018). One can therefore deduce especially right from analysis of variance and standard deviation that the nature of temperature over the study period in Kaduna and FCT followed similar trend, this has been substantiated also with the time plot analysis and decomposition of additive time series analysis and eventually corroborated with seasonal index plot; in all situations, Kano was showed a different trend. In addition, the UECM of these three Nigerian States were developed for implementation at the state level. Further work at the city, district and local planning levels are needed. In practice, the findings through the URM and corresponding climatic planning recommendations that has to do with wind, vegetation, water bodies, urban morphology and topography can make available to governments and local planners with a useful reference which could guide them to better climatic spatial planning at the any level depending on the study focus.