Mapping and Modelling of Urban Landscape of Osogbo Metropolis, Osun State Nigeria, Using Artificial Neural Network

Continuous Geospatial studies of the transitions in Landuse and landcover are very important especially as it relates to baseline assessment as an approach for advising in policy formulations concerning the natural resources sector. This study aimed at mapping and modeling the urban landscape of Osogbo metropolis, Osun state Nigeria, using an artificial neural network with the view of providing a framework for sustainable development and as well as generating data on Landuse and landcover change transitions and maps for planning purposes. Its objectives are to; model and analyze Landuse and landcover changes in Osogbo metropolis for the last 30 years (1990 – 2020) using an artificial neural network; ascertain the trend, and characteristics of Landuse and landcover changes in Osogbo metropolis in the last 30 years; assess the urban landscape change across various terrain configurations with Osogbo Metropolis over the last 30 years, and predict the future urban landscape of Osogbo Metropolis in 2040 using artificial neural network. The methodology involved data acquisition of Landsat, Sentinel-2, and ALOS Palsar images, image preprocessing to correct the scan line error in Landsat 7 ETM+, development of classification scheme, identification of class features and image classification, trend analysis, land cover/land use transition, and prediction to 2040. The assessment of landcover/landuse change revealed significant LULC changes in the studied area. Over 30 years (1990–2020), the built-up area classes increased significantly by 111.97 km2, while vegetation, open space, and water body decreased by 189.33 km2, 7.26 km2, and 3.46 km2 respectively. In terms of increased built-up area, this is largely seen in flat and undulating terrains between 281m and 341m. According to the prediction, by 2040, built up area is expected to grow from 35.89 % to 64.48 % covering an area of 201.2 km2, water body is expected to decrease from 1.11 % to 1.07 % with an area of 3.33 km2, vegetation is expected to decrease from 60.68 % to 32.42 % with an area of 101.15 km2, open space is expected to decrease from 2.33 % to 2.03 % to an area of 6.34 km2. The study´s annual rate of change results is recommended as it reveals the annual decline vegetation within the study area, as a direct consequence can lead to an increase in urban heat islands within the study area.