Historical Hazard Assessment of Climate and Land Use–Land Cover Effects on Soil Erosion Using Remote Sensing: Case Study of Oman

Abstract

Human activities, climate change, and land-use alterations accelerated soil erosion in recent decades and imposed significant threats to soil fertility and stability worldwide. Understanding and quantifying the spatiotemporal variation of soil erosion risks is crucial for adopting the best management practices for surface soils conservation. Here, we present a novel high-resolution (30 m) soil erosion framework based on the G2 erosion model by integrating satellite and reanalysis datasets and Machine Learning (ML) models to assess soil erosion risks and hazards spatiotemporally. The proposed method reflects the impacts of climate change in 1 h time resolutions and land use in 30 m scales on soil erosion risks for almost 4 decades (between 1985 and 2017). The soil erosion hazardous maps were generated/evaluated using Extreme Value Analysis (EVA), utilizing long-term annual soil erosion estimations/projections to aid policymakers in developing management strategies to protect lands against extreme erosion. The proposed framework is evaluated in the Sultanate of Oman, which lacks soil erosion estimation/assessment studies due to data scarcity. Results indicate that soil erosion has increasing perilous trends in high altitudes of the Sultanate of Oman that may cause substantial risks to soil health and stability.