Landuse Landcover Analysis and Prediction using Marcov Chain in Parts of Rivers State, Nigeria

Abstract

Landuse / Landcover changes and its effects on our environment has been a phenomenon of great concern. This study focuses on the need to identify and monitor the Landuse changes within the study area with a view to detect the land consumption rate and the changes that have taken place within a temporal scale of 1989 - 2019. ArcMap 10.1 and Erdas Imagine were deployed for change detection analysis. The prediction of Landuse changes was carried out using Markov Chain analysis. Seven Landuse / Landcover classes were developed. Supervised and post-classification algorithms were employed. Landcover maps were generated and change detection analyses were performed using ArcMap 10.1 and Erdas Imagine software. The statistic evaluation of Landuse Landcover change reveal that built-up areas between 1989 and 1999 increased by 27.44%, the increase was 29.41% between 1999 and 2009 then the increase observed between 2009 and 2019 is 179.4%. The evaluations from the first and last dates reveal that built-up areas, thick forest, bare land and water bodies increased by 360.8%, 127%, 128% and 123.9% respectively while farmland, light forest and swamp decreased by 93%, 33.3% and 20.5% respectively. The overall classification accuracies for 1989, 1999, 2009 and 2019 are 87.00%, 90.00%, 94.53% and 94.14% respectively. The transition probability grid from Markov Chain Analysis reveals that Farm Land and bare Land would be the highest contributors in the Landuse classes to the future increase that would be experienced in the built-up areas in 2029 and 2039.Classified maps from the spatio-temporal Landuse/Landcover changes in the study area would be used as a tool for Land administration, urban planning and environmental management.