Modeling wind erosion flux and its seasonality from a cultivated sahelian surface: A case study in Niger

• Wind erosion is simulated for a bare and a cropped Sahelian plot.
• A 3-years set of data (wind, horizontal fluxes, surface roughness) is used.
• Existing modeling tools are suitable to estimate wind erosion from a millet field.
• Major events, interannual variability and seasonality are well represented.
• Difference between annual eroded mass of bare and cropped plots is well reproduced.

Wind erosion can strongly affect the cultivated areas in semi-arid regions, through soil losses and/or decrease in nutrient contents. Additionally, dust emitted by aeolian erosion affects both the biogeochemical cycles and the Earth radiation budget. Modeling wind erosion and dust emission remains complex especially in semi-arid regions where vegetation interacts with the wind field and may act as a protection of the soil. An existing and widely used wind erosion model is tested in the present study to check the capacity to reproduce observations collected over a millet field and a neighboring bare plot in southwestern Niger during a three-year period. Observations of sediment horizontal fluxes and of vegetation growth and decay show that most of the eroded mass is due to major events occurring at the end of the dry season and at the beginning of the rainy season for the millet field, while erosion also occurs during the dry season for the bare soil plot. Horizontal erosion fluxes were computed with and without obstacles and compared to the measurements. Simulations were found in a good agreement with erosion measurements for both bare and millet plots, in terms of temporal dynamics, reproduction of the major events and annual quantities.Accumulated horizontal fluxes over the millet plot were found to be much lower than over the bare plot for both observations and simulations (respectively 235 to 565 kg m− 1 y− 1 and 332 to 526 kg m− 1 y− 1 over the millet plot; and 773 to 2692 kg m− 1 y− 1 and 1003 to 1986 kg m− 1 y− 1 for the bare plot). Total values over the 3-year period also show a much stronger (by a factor 3 to 4) wind erosion on the bare plot (5365 kg m− 1 y− 1 for observations, 4548 kg m− 1 y− 1 for simulations) than on the cropped one (1357 kg m− 1 y− 1 for observations, 1398 kg m− 1 y− 1 for simulations). Modeling is therefore able to represent anthropogenic impacts on wind erosion over typical Sahelian surfaces.