Modelling and quantifying runoff and soil loss for a degraded and restored Jordan side wadi

Abstract

Water erosion is the most widespread land degrading process worldwide. Especially in drylands, this process is intensified by anthropogenic pressure on soil productivity, resulting in a self-reinforcing effect of topsoil removal, declining vegetation and increased runoff generation. This process is also ongoing in Jordan, where recurring droughts, years of overgrazing and mismanagement of agricultural land has led to removal of native vegetation and consequential land degradation through water erosion. Steep slopes of the side wadis that border the Jordan Valley exacerbate these problems, causing flash floods and clogging of water collection ponds. A SIDA-FAO-funded project of The International Centre for Agricultural Research in Dry Areas (ICARDA) arose to test rehabilitation implementation on small scale and thereby increase local stakeholders’ preparedness for uptake and out-scaling. Therefore, this research aimed to study and quantify surface runoff and soil loss in a Jordan side wadi during its degraded state and after modelled implementation of restoration measures. Simultaneously, erosion hotspots and restoration potential were identified for targeted erosion control. A hydrological pre-assessment was performed for model validation, using the curve number approach, runoff ratios and erosion feature field analyses. An erodibility map was created in GIS based on literature, datasets and field validation. Representative hotspots were selected for additional fieldwork to collect input for the Rangeland Hydrology and Erosion Model (RHEM). This process-based hillslope model was chosen for runoff and soil loss quantification as it has proven to be successful for the Jordan rangelands. It requires rainfall data, soil texture, slope- and cover characteristics. The CLimate GENeration model (CLIGEN) was bias corrected with a local rainfall data, resulting in a 300-year simulation of rainfall statistics. Soil texture and cover characteristics were collected through fieldwork at the hotspot classes and a stream network assessment was performed to obtain representative slope-length combinations, which were used to upscale from hillslope- to wadi level. The degraded scenarios yielded a runoff ratio of 10,7%, which was in agreement with the hydrological pre-assessment. Soil loss was estimated to be 2,31 ton/ha/y, with maximum hotspots of 4,99 ton/ha/y. This was lower than expected, which could likely be attributed to high rock content at the fieldwork sites. Restored equilibrium scenarios yielded an average reduction of 21,8% in runoff and 53,4% in soil loss for the entire wadi, emphasizing the impact of increased vegetation cover on the ecosystem’s resilience. Direct impact of intended WH structures was assessed by a script based on pit dimensions, infiltration rate and trapping efficiency. Runoff process description of the script lacked detail, underestimating runoff capturing capacity and overestimating sediment buffer capacity, yet providing a rough tool for determining optimal pit dimensions and spacing. This yielded an average decrease of 35% in runoff and 71% in soil loss for the restorable scenarios.