Scholarlink

<p>The Mannarkad Watershed is a subtropical mountain catchment in the Western Ghats, Kerala, facing critical soil erosion issues due to climate change and anthropogenic activities, which cause adverse impacts on soil infertility and agricultural productivity. This study investigates the spatio-temporal variability of soil erosion rate (SE) and erosion hotspots for 2000, 2010, and 2023 using the GIS-based Revised Universal Soil Loss Equation (RUSLE) model. Sediment yield estimates were derived at detailed spatial scales using the Sediment Delivery Ratio (SDR) method. The results show that annual soil erosion rates have grown from 14.9 t/ha in 2000 to 23.02 t/ha in 2023, occurring in the sloped upland valleys and foothills across the watershed have seen the most severe erosion. Wherein, rainfall erosivity is at a higher rate of 1129–1669&#xa0;MJ·mm/ha·hr·yr in 2000, and 2344–3076&#xa0;MJ·mm/ha·hr·yr in 2023, causing sediment transport towards downslope and deposited along the valley fills and channel levees. Over the last two decades, the findings revealed that the erosion rate has progressively increased in various hotspots owing to rainfall variability and land-use changes. The RUSLE model exhibits that the erosion hotspots, mainly found in middle laterite plateaus, and sparsely vegetated uplands, consist of steep-sloped valleys, upland stream orders, barren lands, down-sloped fallows and proximity of built-up areas. Significantly, the sediment yield is at a higher rate (0.27–23.81 t/ha/yr) in the valley fills, downpour streams of the foothills, with an average rate of 6.75 t/ha/yr, contributing to alluvial plains and channel levees. The RSULE model calculates the AUC value of 0.823 based on the spatial correlation of eroded sites of GPS survey and map-derived data, indicating the model’s reliable predictability. The sediment yield was validated with rain gauge-based sedimentation data from the Kanjirapuzha Dam station, showing close agreement with the derived SDR rate (predicted 2.79 t/ha/yr vs. observed 5.8 t/ha/yr). GIS-based RUSLE–SDR is widely used for soil erosion rate and sediment yield at the pixel scale, and in heterogenic landscapes, and lacks this in other conventional approaches. However, the output of this model is subject to uncertainties that depend on the quality of input parameters, scale, and resolution. Despite these limitations, the outcomes provide a spatio-temporal pattern of soil erosion and sediment yield and are a valuable database for soil and water conservation planning and management activities.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Unveiling decadal variability of soil erosion and sediment yield using GIS-based RUSLE and SDR techniques – a case study of mountain watershed of the Western Ghats, India

  • 期刊论文

  • 发表日期：

    2025年12月31日

查看全文
  • S. Kaliraj,
  • S. Richard Abishek,
  • P. S. Preethy,
  • Sruthy Manokaran,
  • Reji Srinivas,
  • Subbarayan Saravanan

摘要

The Mannarkad Watershed is a subtropical mountain catchment in the Western Ghats, Kerala, facing critical soil erosion issues due to climate change and anthropogenic activities, which cause adverse impacts on soil infertility and agricultural productivity. This study investigates the spatio-temporal variability of soil erosion rate (SE) and erosion hotspots for 2000, 2010, and 2023 using the GIS-based Revised Universal Soil Loss Equation (RUSLE) model. Sediment yield estimates were derived at detailed spatial scales using the Sediment Delivery Ratio (SDR) method. The results show that annual soil erosion rates have grown from 14.9 t/ha in 2000 to 23.02 t/ha in 2023, occurring in the sloped upland valleys and foothills across the watershed have seen the most severe erosion. Wherein, rainfall erosivity is at a higher rate of 1129–1669 MJ·mm/ha·hr·yr in 2000, and 2344–3076 MJ·mm/ha·hr·yr in 2023, causing sediment transport towards downslope and deposited along the valley fills and channel levees. Over the last two decades, the findings revealed that the erosion rate has progressively increased in various hotspots owing to rainfall variability and land-use changes. The RUSLE model exhibits that the erosion hotspots, mainly found in middle laterite plateaus, and sparsely vegetated uplands, consist of steep-sloped valleys, upland stream orders, barren lands, down-sloped fallows and proximity of built-up areas. Significantly, the sediment yield is at a higher rate (0.27–23.81 t/ha/yr) in the valley fills, downpour streams of the foothills, with an average rate of 6.75 t/ha/yr, contributing to alluvial plains and channel levees. The RSULE model calculates the AUC value of 0.823 based on the spatial correlation of eroded sites of GPS survey and map-derived data, indicating the model’s reliable predictability. The sediment yield was validated with rain gauge-based sedimentation data from the Kanjirapuzha Dam station, showing close agreement with the derived SDR rate (predicted 2.79 t/ha/yr vs. observed 5.8 t/ha/yr). GIS-based RUSLE–SDR is widely used for soil erosion rate and sediment yield at the pixel scale, and in heterogenic landscapes, and lacks this in other conventional approaches. However, the output of this model is subject to uncertainties that depend on the quality of input parameters, scale, and resolution. Despite these limitations, the outcomes provide a spatio-temporal pattern of soil erosion and sediment yield and are a valuable database for soil and water conservation planning and management activities.