Table 1 GWR-Derived coefficients and diagnostics for erosion Modeling.
Figure Ref. | Parameter | Role in erosion modeling | Coefficient range | Interpretation |
|---|---|---|---|---|
Figure 3j | Slope | Controls runoff velocity and sediment transport | –0.42 to + 0.68 | Positive values indicate steep slopes intensify erosion |
Figure 3k | Profile Curvature | Reflects vertical flow acceleration and detachment zones | –0.31 to + 0.47 | Convex profiles enhance erosion; concave reduce it |
Figure 3l | Plan Curvature | Captures lateral flow convergence and divergence | –0.29 to + 0.52 | Convergent zones concentrate runoff, increasing erosion risk |
Figure 4a | LS Factor | Combines slope length and steepness; key RUSLE component | + 0.15 to + 0.91 | Higher values correlate with greater erosion susceptibility |
Figure 4b | Topographic Wetness Index (TWI) | Indicates moisture accumulation and saturation potential | + 0.08 to + 0.63 | Saturated areas prone to soil weakening and detachment |
Figure 4c | Valley Depth | Measures incision and flow concentration | + 0.12 to + 0.74 | Deep valleys channel erosive energy |
Figure 4d | Channel Network Base Level | Elevation of stream outlets; influences upstream erosion dynamics | –0.21 to + 0.58 | Lower base levels intensify upstream erosion |
Figure 4e | Channel Network Distance | Proximity to drainage channels | –0.49 to + 0.36 | Closer proximity increases runoff concentration and erosion |
Figure 4f | Condition Numbers | Diagnostic for multicollinearity among predictors | 8.2 to 17.6 | Values < 30 indicate stable model; no multicollinearity detected |
Figure 4g | Local R² | Measures local explanatory power of the GWR model | 0.45 to 0.89 | Higher values reflect strong model fit in specific regions |
Figure 4h | Standard Error | Indicates uncertainty in coefficient estimates | 0.03 to 0.27 | Lower errors suggest reliable coefficient estimation |
