Gully erosion zonation mapping using integrated geographically weighted regression with certainty factor and random...

Area of Science:

• Environmental Science
• Geosciences
• Soil Science

Background:

• Gully erosion poses a significant threat to agricultural land, infrastructure, and residential areas annually.
• Effective gully erosion assessment and mapping are crucial for informed environmental management and soil conservation strategies.

Purpose of the Study:

• To develop and evaluate a novel integrated model for gully erosion susceptibility zonation mapping.
• To combine Geographically Weighted Regression (GWR) with Certainty Factor (CF) and Random Forest (RF) models for enhanced prediction accuracy.

Main Methods:

• A new GWR-CF-RF integrated model was developed and applied to the Mahabia watershed, Iran.
• Gully erosion inventory map (GEIM) and 12 gully-related causal factors (GRCFs) were utilized for modeling.
• The watershed was segmented into eight homogeneous units using GWR coefficients before applying the RF-CF model.

Main Results:

• The integrated GWR-CF-RF model achieved a high predictive accuracy of 96.7%, outperforming individual CF (76.3%), RF (77.6%), and CF-RF (89.7%) models.
• Distance to stream, distance to road, and land use/land cover were identified as key factors influencing gully formation.
• The model demonstrated superior accuracy in gully erosion susceptibility zonation mapping.

Conclusions:

• The proposed GWR-CF-RF approach offers a significant advancement in gully erosion susceptibility mapping.
• The findings provide valuable tools for local managers and planners to implement effective environmental management and soil conservation measures.