Monitoring Tree Cover Changes in Kenya’s Five Major Water Towers using Geospatial Technologies

Abstract

Water towers are the hearts that pump life into any ecosystem by providing fresh water recharge to sustain most living organisms that facilitate the very existence of humans. Numerous studies have documented the importance of these water towers in sustenance of fresh water that account for only about three percent of the Earth’s water. While the role water towers play in ensuring our survival is clear, they are continuously and gradually being degraded as a result of pressure from demographic factors in addition to uncontrollable natural factors. Inadequate information about the area and spatial extents of tree cover loss jeopardizes systematic monitoring of these forests. This has led to haphazard restoration efforts that tend to be unsuccessful in the long term. Earth observation technology enables efficient and effective mapping and monitoring of forest cover within water towers over an extensive range of temporal and spatial scales. This study applied remote sensing techniques to investigate the amount of tree cover lost between 2005 and 2019 in Kenya’s major water towers using Landsat 30-m resolution imagery. The images were classified into two classes, tree cover and non-tree cover, from which change detection maps were produced with their underlying statistics of tree cover loss in terms of area and location. The study established a trend in tree cover loss across all the water towers where Mt. Kenya experienced approximately 7% decline, Mt. Elgon 14%, Aberdare range 18%, Mau complex 24% and Cherangani hills 25% which all amount to an average of 19% throughout the study period. The government and NGOs should therefore apply geospatial techniques in monitoring tree cover. This will enable proper identification of degraded areas within forests and further make informed decisions in formulating policies and executing restoration programmes. Keywords: Water towers, tree cover, forests, fresh water, Landsat.