Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya
This study aimed at (i) developing, evaluating and comparing the performance of support vector machines (SVM), boosted regression trees (BRT), random forest (RF) and logistic regression (LR) models in mapping gully erosion susceptibility, and (ii) determining the important gully erosion conditioning...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-04-01
|
| Series: | Land |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-445X/12/4/890 |
| _version_ | 1797604647035207680 |
|---|---|
| author | Kennedy Were Syphyline Kebeney Harrison Churu James Mumo Mutio Ruth Njoroge Denis Mugaa Boniface Alkamoi Wilson Ng’etich Bal Ram Singh |
| author_facet | Kennedy Were Syphyline Kebeney Harrison Churu James Mumo Mutio Ruth Njoroge Denis Mugaa Boniface Alkamoi Wilson Ng’etich Bal Ram Singh |
| author_sort | Kennedy Were |
| collection | DOAJ |
| description | This study aimed at (i) developing, evaluating and comparing the performance of support vector machines (SVM), boosted regression trees (BRT), random forest (RF) and logistic regression (LR) models in mapping gully erosion susceptibility, and (ii) determining the important gully erosion conditioning factors (GECFs) in a Kenyan semi-arid landscape. A total of 431 geo-referenced gully erosion points were gathered through a field survey and visual interpretation of high-resolution satellite imagery on Google Earth, while 24 raster-based GECFs were retrieved from the existing geodatabases for spatial modeling and prediction. The resultant models exhibited excellent performance, although the machine learners outperformed the benchmark LR technique. Specifically, the RF and BRT models returned the highest area under the receiver operating characteristic curve (AUC = 0.89 each) and overall accuracy (OA = 80.2%; 79.7%, respectively), followed by the SVM and LR models (AUC = 0.86; 0.85 & OA = 79.1%; 79.6%, respectively). In addition, the importance of the GECFs varied among the models. The best-performing RF model ranked the distance to a stream, drainage density and valley depth as the three most important GECFs in the region. The output gully erosion susceptibility maps can support the efficient allocation of resources for sustainable land management in the area. |
| first_indexed | 2024-03-11T04:49:43Z |
| format | Article |
| id | doaj.art-1608945bcf8f46ce914b659b8f08bf93 |
| institution | Directory Open Access Journal |
| issn | 2073-445X |
| language | English |
| last_indexed | 2024-03-11T04:49:43Z |
| publishDate | 2023-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Land |
| spelling | doaj.art-1608945bcf8f46ce914b659b8f08bf932023-11-17T20:03:42ZengMDPI AGLand2073-445X2023-04-0112489010.3390/land12040890Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of KenyaKennedy Were0Syphyline Kebeney1Harrison Churu2James Mumo Mutio3Ruth Njoroge4Denis Mugaa5Boniface Alkamoi6Wilson Ng’etich7Bal Ram Singh8Kenya Agricultural and Livestock Research Organization, Kenya Soil Survey, P.O. Box 14733, Nairobi 00800, KenyaSchool of Agriculture and Biotechnology, University of Eldoret, P.O. Box 1125, Eldoret 30100, KenyaSchool of Agriculture and Biotechnology, University of Eldoret, P.O. Box 1125, Eldoret 30100, KenyaSchool of Agriculture and Biotechnology, University of Eldoret, P.O. Box 1125, Eldoret 30100, KenyaSchool of Agriculture and Biotechnology, University of Eldoret, P.O. Box 1125, Eldoret 30100, KenyaSchool of Agriculture and Biotechnology, University of Eldoret, P.O. Box 1125, Eldoret 30100, KenyaSchool of Agriculture and Biotechnology, University of Eldoret, P.O. Box 1125, Eldoret 30100, KenyaSchool of Agriculture and Biotechnology, University of Eldoret, P.O. Box 1125, Eldoret 30100, KenyaFaculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, P.O. Box 5003, 1432 Ås, NorwayThis study aimed at (i) developing, evaluating and comparing the performance of support vector machines (SVM), boosted regression trees (BRT), random forest (RF) and logistic regression (LR) models in mapping gully erosion susceptibility, and (ii) determining the important gully erosion conditioning factors (GECFs) in a Kenyan semi-arid landscape. A total of 431 geo-referenced gully erosion points were gathered through a field survey and visual interpretation of high-resolution satellite imagery on Google Earth, while 24 raster-based GECFs were retrieved from the existing geodatabases for spatial modeling and prediction. The resultant models exhibited excellent performance, although the machine learners outperformed the benchmark LR technique. Specifically, the RF and BRT models returned the highest area under the receiver operating characteristic curve (AUC = 0.89 each) and overall accuracy (OA = 80.2%; 79.7%, respectively), followed by the SVM and LR models (AUC = 0.86; 0.85 & OA = 79.1%; 79.6%, respectively). In addition, the importance of the GECFs varied among the models. The best-performing RF model ranked the distance to a stream, drainage density and valley depth as the three most important GECFs in the region. The output gully erosion susceptibility maps can support the efficient allocation of resources for sustainable land management in the area.https://www.mdpi.com/2073-445X/12/4/890soil erosionland degradationsustainable land managementlandscape restorationspatial predictionmachine learning |
| spellingShingle | Kennedy Were Syphyline Kebeney Harrison Churu James Mumo Mutio Ruth Njoroge Denis Mugaa Boniface Alkamoi Wilson Ng’etich Bal Ram Singh Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya Land soil erosion land degradation sustainable land management landscape restoration spatial prediction machine learning |
| title | Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya |
| title_full | Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya |
| title_fullStr | Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya |
| title_full_unstemmed | Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya |
| title_short | Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya |
| title_sort | spatial prediction and mapping of gully erosion susceptibility using machine learning techniques in a degraded semi arid region of kenya |
| topic | soil erosion land degradation sustainable land management landscape restoration spatial prediction machine learning |
| url | https://www.mdpi.com/2073-445X/12/4/890 |
| work_keys_str_mv | AT kennedywere spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT syphylinekebeney spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT harrisonchuru spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT jamesmumomutio spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT ruthnjoroge spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT denismugaa spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT bonifacealkamoi spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT wilsonngetich spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya AT balramsingh spatialpredictionandmappingofgullyerosionsusceptibilityusingmachinelearningtechniquesinadegradedsemiaridregionofkenya |