Multi-Temporal land use change detection analysis using landsat data: The case of Lilongwe City

By: Brino Kapatamoyo, Vincent Katonda, Steven Gondwe,

Abstract

Land use land cover (LULC) studies have played a significant role in sustainable urban planning and development. In this study, Remote Sensing and GIS was used to detect and analyse LULC change for the years 1998, 2008 and 2018 in Lilongwe city. The maximum-likelihood algorithm was adopted for supervised classification Landsat TM and Landsat OLI images. Classification results showed that built-up area increased substantially whereas bare land and vegetation declined, while water remained constant throughout the study period. Post-classification comparison of the classified images based on the transition matrix revealed that for the period of 1998 to 2008, vegetation had the highest transition with 47.35 km² (47.20%) of its total area in 1998, the majority being converted to bare land (40.24 km²), built-up area (7.05 km²) and water (0.056 km²). The results further showed that built-up area increased at annual rate of 7.8%, 14.5% &, and 16.81% for the period of 1998 to 2008, 2008 to 2018, and 1998 to 2018 respectively. The overall accuracies for 1998, 2008 and 2018 images, were 95.46%, 94.23%, and 89.73% respectively. The study unveiled LULC change dynamics for Lilongwe city and the rate at which changes were taking place. With an ever-increasing urbanization in this City, these findings present a call for the City to strengthen measures for restoring the deteriorated environment in order to ensure a sustainable environment for all city dwellers.