Using remote sensing technology to detect, model and map desertification: A review

Abstract

Desertification is a serious global environmental problem that affects many people directly in countries with semi-arid or arid climates. The causes of desertification are diverse and complicated, ranging from international economic activities to unsustainable land use practices of local communities. The consequence of desertification reduces the ability of the land to support life and decreases biodiversity. Natural processes and artificial forces initiate this process. Natural causes of desertification include climatic factors drought, and water or wind erosion. Human induced activities that can cause desertification include over-cultivation, overgrazing, deforestation, and poor water management.The aim of this paper is to analyse how remote sensing and GIS have been used to monitor desertification globally. Remote sensing has proven to be efficient in detecting desertification processes including changes in natural vegetation, land use, and soil. Multi-temporal coverage provided by satellite data facilitates the use of remote sensing imagery to monitor changes in land coverage and usage over time. Remote sensing data and GIS are critical for extracting reliable information important for assessing environmental changes and land quality in any given region. Various techniques in remote sensing combined with analytical methods such as NDVI or classification provide primary data that can be used to assess desertification processes. Several key indicators of the processes of desertification are identified where remote sensing can be used to detect, monitor, and map affected areas. Changing vegetation and land use, drought, soil, erosion, and urbanization are the most common indicators of desertification used by researchers. Results of studies can be used to make important management, environmental, and political decisions. Therefore, care, must be taken to use recent imagery, select the appropriate technique, and to include as much additional data as possible so that reliable and robust results are obtained.