Drought Monitoring in a Data-Sparse Region Based on In-situ and Satellite Imagery within Google Earth Engine: A Case Study of Awash River Basin, Ethiopia

Abstract

Drought is a significant climatic event that adversely affects agricultural productivity, water resources, and livelihoods, particularly in regions such as the Awash River Basin in Ethiopia. This study aims to monitor drought conditions from 2000 to 2021 by integrating remote sensing data and in-situ observations using the Google Earth Engine (GEE) platform. The study utilized the Vegetation Condition Index (VCI), Temperature Condition Index (TCI), and Vegetation Health Index (VHI) derived from MODIS satellite imagery. These indices were categorized into drought severity levels (No Drought, Mild, Moderate, Severe, and Extreme) and were used to monitor the spatial and temporal patterns of drought. SPI3 data were derived from rainfall records at 23 meteorological stations in the basin, and Pearson correlation analysis was conducted to assess the relationship between SPI3 and satellite-derived indices. The results show that mild and moderate drought conditions were predominant in the Awash River Basin over a 22-year period. The VCI results indicated that mild drought affected over 50% of the area in many years, with notable peaks in 2001, 2010, 2011, and 2021, where mild drought extended over 70% of the basin. Moderate drought conditions were prevalent in 2008, 2009, and 2016, affecting up to 50% of the basin area. Severe and extreme droughts, although less frequent, occurred in specific years such as 2009, 2015, and 2017, particularly in the middle and lower parts of the basin. The TCI results revealed that temperature-related drought stress was significant in 2009 and 2017, with severe and extreme droughts affecting the northern and central regions. The VHI, which combines the VCI and TCI, showed that moderate droughts were most common across the study period, especially in 2009 and 2017. The highest VHI values, which indicate healthier vegetation conditions, were observed in 2020 and 2021.Pearson correlation analysis between SPI3 and the satellite-based indices showed strong positive correlations. The VCI and SPI3 exhibited moderate to strong correlations across most stations, with the highest correlation in Asebe Teferi (0.57). TCI correlation with SPI3 ranged from -0.26 to 0.53, while VHI demonstrated strong correlations, particularly in Asebe Teferi (0.63), validating the use of vegetation-based indices for drought monitoring. This study underscores the broader significance of effective drought monitoring in the Awash River Basin because droughts in this region have profound impacts on local agriculture, water availability, and livelihoods. The ability to monitor and assess drought conditions supports more resilient agricultural practices and informs water resource management strategies that are crucial for sustaining the regions economy and food security. The findings of this study have practical applications such as enhancing early warning systems and guiding resource allocation for drought mitigation efforts. By leveraging GEE's capabilities of GEE, policy