preprint
The rapid and typically uncontrolled growth of many African cities leads to a plethora of problems, including the formation and expansion of large urban gullies (UGs). These UGs often result in the destruction of homes and infrastructure, displacement of people, and loss of life. In many ways, the formation mechanisms of UGs are similar to those of gullies in other environments. Yet, urban land cover and tropical rainfall conditions, as well as their location in densely populated areas typically make them much more severe. Furthermore, the problems associated with UGs are likely to worsen in the near future as a result of continued urban expansion and climate change. However, this newly emerging geo-hydrological hazard received hitherto very little research attention. Several studies report on the occurrence and impacts of UGs but they remain limited to specific local case studies. A clear understanding of the patterns, impacts and driving factors of UGs at larger scales is currently lacking. To address this gap, we aim to better understand the spatial patterns and UG susceptibility at the scale of Africa. Through the visual analysis of satellite imagery, we documented more than 4,000 cases of UG occurrence, significantly affecting 12 countries across Africa. These UGs are mainly spread over (sub-)tropical areas with D.R. Congo, Angola, Republic of Congo, Nigeria, and Mozambique being the most impacted countries. Using this database, we trained a random forest model that accurately simulates UG occurrence in (peri-)urban areas across Africa, with AUC greater than 0.9. Our results demonstrate that a combination of topography, rainfall characteristics, soil type, and variables describing the urban context (e.g. built-up area, road density) can explain variations in susceptibility to UG occurrence within and across cities. This dataset and model represent critical initial steps toward understanding, mitigating and preventing the risks of UGs in Africa, both now and in the future.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.5194/egusphere-egu25-8975
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.