Scholarlink

<p>The <i>Shivalik</i> foothills of Punjab are highly susceptible to soil erosion due to their uneven topography, fragile geological formations, and intense rainfall events. The land degradation is going on in alarming rate. A Google Earth Engine-based cloud computing framework was utilized to implement the Revised Universal Soil Loss Equation (RUSLE) model for spatial assessment of soil erosion. The present study aims to assess and map soil erosion risk in the Shivalik foothill watersheds using the Revised Universal Soil Loss Equation (RUSLE) integrated with cloud-based geospatial techniques. The watersheds were delineated using the ALOS PALSAR Digital Elevation Model (DEM). The RUSLE parameters were derived from multiple geospatial datasets: rainfall erosivity (R) was estimated from CHIRPS rainfall data, soil erodibility (K) from FAO ISRIC SoilGrids, slope length and steepness (LS) were generated from DEM-based flow accumulation equations, the cover management (C) factor was derived from Sentinel-2 imagery, and the conservation practice (P) factor was obtained from a land use/land cover (LULC) map, prepared using the Random Forest classification algorithm. The classification accuracy was evaluated using a confusion matrix, and the model performance was validated through the AUC/ROC curve method. The results showed that the R factor ranged from 600.24 to 824.35, with the highest values observed in watersheds 6 and 8. The K factor varied from 0.0082 to 0.027, while the LS factor ranged from 0 to 12.13, indicating the presence of steep slopes susceptible to erosion. The C and P factors varied from 0 to 1, reflecting differences in vegetation cover and conservation practices across the watersheds. Integration of all RUSLE parameters revealed soil loss rates ranging from 0 to 13.31 t ha⁻¹ yr⁻¹, with watershed 7 experiencing the highest soil erosion. Spatial analysis indicated that upstream areas exhibited greater erosion due to steep slopes and sparse vegetation, whereas downstream areas showed relatively lower erosion because of gentler slopes and denser vegetation cover. The study demonstrates that cloud-based geospatial modelling provides an efficient approach for identifying erosion-prone zones and prioritizing watershed management strategies. The true AUC/ROC curve shows the value 0.837 indicates good accuracy of model for erosion estimation. The findings highlight the need for site-specific soil and water conservation measures such as contour bunding, check dams, and afforestation to reduce soil loss and promote sustainable watershed management in the <i>Shivalik</i> foothills.</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

Cloud computing and geospatial approach for assessment of soil erosion in lower Shiwalik foothills of Punjab, India

  • 期刊论文

  • 发表日期：

    2026年6月17日

查看全文
  • Pravin Dahiphale,
  • Abrar Yousuf

摘要

The Shivalik foothills of Punjab are highly susceptible to soil erosion due to their uneven topography, fragile geological formations, and intense rainfall events. The land degradation is going on in alarming rate. A Google Earth Engine-based cloud computing framework was utilized to implement the Revised Universal Soil Loss Equation (RUSLE) model for spatial assessment of soil erosion. The present study aims to assess and map soil erosion risk in the Shivalik foothill watersheds using the Revised Universal Soil Loss Equation (RUSLE) integrated with cloud-based geospatial techniques. The watersheds were delineated using the ALOS PALSAR Digital Elevation Model (DEM). The RUSLE parameters were derived from multiple geospatial datasets: rainfall erosivity (R) was estimated from CHIRPS rainfall data, soil erodibility (K) from FAO ISRIC SoilGrids, slope length and steepness (LS) were generated from DEM-based flow accumulation equations, the cover management (C) factor was derived from Sentinel-2 imagery, and the conservation practice (P) factor was obtained from a land use/land cover (LULC) map, prepared using the Random Forest classification algorithm. The classification accuracy was evaluated using a confusion matrix, and the model performance was validated through the AUC/ROC curve method. The results showed that the R factor ranged from 600.24 to 824.35, with the highest values observed in watersheds 6 and 8. The K factor varied from 0.0082 to 0.027, while the LS factor ranged from 0 to 12.13, indicating the presence of steep slopes susceptible to erosion. The C and P factors varied from 0 to 1, reflecting differences in vegetation cover and conservation practices across the watersheds. Integration of all RUSLE parameters revealed soil loss rates ranging from 0 to 13.31 t ha⁻¹ yr⁻¹, with watershed 7 experiencing the highest soil erosion. Spatial analysis indicated that upstream areas exhibited greater erosion due to steep slopes and sparse vegetation, whereas downstream areas showed relatively lower erosion because of gentler slopes and denser vegetation cover. The study demonstrates that cloud-based geospatial modelling provides an efficient approach for identifying erosion-prone zones and prioritizing watershed management strategies. The true AUC/ROC curve shows the value 0.837 indicates good accuracy of model for erosion estimation. The findings highlight the need for site-specific soil and water conservation measures such as contour bunding, check dams, and afforestation to reduce soil loss and promote sustainable watershed management in the Shivalik foothills.