The use of remote sensing for desertification studies: A review

The study and assessment of desertification and/or the advance or retreat of arid areas as a function of natural and anthropogenic causes is necessary for the prediction of future risks from climate change, and to support policymaking, action plans, and mitigation measures that can be taken at local and global scales. Remote sensing enables modelling, monitoring, and prediction of the behaviour of several elements of desertification. There have been numerous approaches to study desertification using remote sensing over the years. This research explored the timeline and global distribution of studies using remote sensing in studying desertification. Additionally, the review evaluated the key methods and variables that have been used to study desertification from remote sensing data. The use of remote sensing for desertification studies can be trace back to 1991. From 2015 to 2020, more than 40 articles were published per year, showing that there has been a recent increase in the use of remote sensing techniques and its availability for monitoring desertification. Most regions of the world affected by desertifi-cation are being studied using remote sensing, however, there is a marked geographical variation between the number of studies in various regions, with Asia having disproportionately high number of studies compared to America or Africa. The country with most studies of desertification using remote sensing is China. In terms of satellite data, Landsat images provide the bulk of data used to study desertification, especially the Thematic Mapper (TM) sensor. Classification and change detection are the most used methods to study desert-ification from remote sensing data. Additionally, land cover/land use change and vegetation and its attributes (e. g., Normalized Difference Vegetation Index -NDVI) are the most used variables to study desertification using remote sensing techniques. Finally, the review found major differences in terms of the ranges or thresholds applied to these variables when determining the presence or risk of desertification. Therefore, there is a need to develop thresholds and ranges of changes of key selected variables, which can be used to determine the presence of desertification.