Can we use the Curve Number approach to predict gully head occurrence at the continental scale of Africa?

Gully erosion is an important land degradation process, threatening soil and water resources worldwide. However, in contrast to sheet and rill erosion, our ability to simulate and predict gully erosion remains limited, especially at the continental scale. Nevertheless, such models are essential for the development of suitable land management strategies, but also to better quantify the role of gully erosion in continental sediment budgets. We aim to bridge this gap by developing a first spatially explicit and process-oriented model that simulates average gully erosion rates at the continental scale of Africa.

We are developing a model that predicts the likelihood of gully head occurrence by means of the Curve Number (CN) method. This model will allow to simulate the spatial patterns of gully density at high resolution (30m) based on the physical principles that control the gully erosion process by using GIS and spatial data sources that are available at the continental scale. To calibrate and validate this model, we make use of an extensive database of 44 000 gully heads mapped over 1680 sites that are randomly distributed across Africa. The exact location of all gully heads was manually mapped by trained experts, using high resolution optical imagery available in Google Earth. This allows to extract very detailed information at the level of the gully head, such as the local slope and the area draining to the gully.

Based on an explorative analysis on a subset of this dataset we found that the CN method does not directly allow to make reliable predictions on gully head occurrence within a pixel. Although land use and land cover seem to play an important role (with gully heads being clearly located in erosion-prone land use classes), the hydrological soil groups (HSGs) based on soil texture do not provide a clear relation between soils with high runoff risk and gully occurrence. A potential cause for this is likely that compensating soil effects occur: i.e. HSGs that produce low runoff volumes may be characterized by a lower soil cohesion, making them nonetheless prone to gullying. This may then cause the combination of HSG and land use to be an insignificant predictor of gully occurrence. Also uncertainties on the input data likely play an important role in this.

Overall, our results indicate that modelling gully densities using a process-oriented and spatially explicit method offers opportunities to better quantify this important land degradation process at the global scale. Nevertheless, a key challenge lies in accurately quantifying the importance of soil characteristics and especially in better understanding their relative contribution to runoff production and soil cohesion.