Summary
This study applied frequency ratio and random forest machine learning algorithms to develop gully erosion susceptibility maps for a small catchment in Ethiopia, analysing 56 surveyed gullies and 14 environmental variables. The research identified drainage density, elevation, land use, and groundwater table as the primary predictors of gully formation, with novel findings on groundwater table significance for watershed management planning. The models demonstrated utility in predicting erosion hotspots, though variable importance differed between soil types (land cover dominant for Nitisols; drainage density for Vertisols).
UK applicability
Whilst the specific environmental context is Ethiopian, the methodological approach and machine learning framework may be transferable to UK catchment management contexts. However, the relative importance of predictor variables (particularly groundwater table prominence) may differ substantially given UK soil types, climate, and hydrology.
Key measures
Area under the curve (AUC) for model prediction accuracy; variable importance rankings for 14 environmental factors; gully location and extent data from 56 surveyed gullies
Outcomes reported
The study developed gully erosion susceptibility maps using frequency ratio and random forest algorithms to identify hotspot areas. The analysis identified drainage density, elevation, land use, and groundwater table as the four most important predictive factors for gully formation.
Topic tags
Dig deeper with Pulse AI.
Pulse AI has read the whole catalogue. Ask about this record, its theme, or how the findings apply to UK farming and policy — every answer cites the underlying studies.