redicting Land Use Land Cover Change of Jos-Bukuru Metropolis Plateau State through integrated CA-Markov approach

Abstract

The dynamics of Land use Land cover (LULC) is an essential concern that affects the global environment, climate change, and sustainable development and Jos-Bukuru Metropolis is not an exception. Hence, detecting and predicting these changes will help in deriving active land use and planning policies that fits the local conditions. Landsat images of the years 1986, 2000 and 2019 were used to study the spatio-temporal distributions pattern of LULC. The study is aimed at modelling these changes using the integrated CA-Markov process. The maximum likelihood algorithm and cross tabulation technique (CTT) was performed to analyze these spatial changes. Validation and prediction using CA-Markov was performed. The accuracy of classification obtained are: 0.9659, 0.9510, and 0.9537 for 1986, 2000 and 2019 images respectively. Change detection from 1986 to 2000 reveals that built-up areas and rocks have positive image difference while vegetation, water body, mining area and open space with negative image difference. Built-up area continuously dominate with annual rate of change (ARC) by 6.27% from 1986 to 2000 and 2.17% from 2000 to 2019 while rock has 5.80% and 0.45%. – 2.82%, -0.87%, -8.00% and -5.265 for vegetation, water body, mining area and open space from 1986 to 2000. Likewise -0.48%, -0.88%, -2.78% and -2.65% from 2000 to 2019. CA-Markov analysis was used to predict LULC of 2019 where a Kappa statistics of validation was 0.8083. This shows a substantial level agreement. Based on the analysis of LULC change, this has revealed that there is rapid conversion of LULC classes most especially, vegetation cover, water body, mining area and open space to built-up. The worst hit LULC is open space while built up continually shows significant increase based on the predictive LULC.