Remote sensing for environmental assessment in the context of humanitarian emergencies: a case study of Ethiopia, Gambella region, in Africa

The improvement in remote sensing techniques to obtain environmental information, coupled with a significant expansion in the number of refugee camps, in their size and longevity, and in their impacts on natural resources, has increased the need for better information about the nature of those resources and their resilience to human demands. Large and fast-growing refugee camps place heavy demands on areas to locate new arrivals and may have long-lasting impacts on land cover and land use, forest degradation, environmental services and social dynamics. In this context, the role of remote sensing is becoming increasingly popular in the field of humanitarian action, as it is an independent and reliable source of information and allows both a quick response to emergencies and the monitoring of gradual changes, especially when field observations in the area are not possible. This study was chosen to adhere to the “traditional” meaning of refugee and managed settlements in Africa, under the UNHCR mandate. As a result, certain groups, such as returnees or internally displaced people, as well as city slums and locations of dispersed or informal settlements, were left out of the selection process. The main goal of this study is to use remote sensing to analyse significant land cover changes relatable to refugee camp settlements. The Gambela region of Ethiopia was selected as a region of interest. Already in 2016, several sources reported Ethiopia to be within the top 6th refugee-hosting country worldwide, with the Gambela region having rich natural resources but hosting the country's largest refugee population. Most refugees are from South Sudan and live in seven refugee camps. The four biggest camps, Nguenyyiel, Tierkidi, Jewi, and Kule, were selected to conduct the study. A comprehensive and multidisciplinary literary review investigated existing RS application studies, historical developments, and social and state patterns. C-band synthetic aperture radar (SAR) from Copernicus Sentinel-1 and USGS Landsat-8, Level 2, Collection 2 imagery are used to compare and analyse vegetation cover and land use and land cover changes (LULCC) over a ten-year time series analysis period. Despite several limitations found in assessing optical imagery, SAR data was revealed to be a good alternative. These last have been pre-processed, including radiometric calibration, geocoding, and excluding speckle filtering. A composite time series raster image with a 10-meter resolution was created to display an RGB layer representing buildings and shelters’ presence over the years. As a result, this study found a correlation between the trend of SAR backscatter values and refugee population growth, paving the way for future investigations.