Modeling vegetation and wind erosion from a millet field and from a rangeland: two Sahelian case studies
Quantifying wind erosion and dust emissions in the semi-arid Sahel remains challenging because of the large seasonal and interannual dynamics of surface properties. The increasing conversion of rangelands into croplands raises issues for quantifying wind erosion over these two contrasted surfaces. Whereas wind erosion models have been so far applied to these two surface types separately, this study proposes a common modeling approach to represent the horizontal flux from Sahelian rangelands and croplands. Pair simulations of both typical Sahelian land surface types investigate the horizontal flux due to wind erosion over a 3-year period for two instrumented sites in Mali and Niger. Two different vegetation models simulate the specific phenology and growth of a rangeland grass and a millet crop. These models also account for the local cropping and pastoral practices. Compared to field measurements, the vegetation cover is satisfyingly simulated by the models, especially the strong seasonal dynamics. Specific parameterizations of the aerodynamic surface roughness length (Z0) as a function of vegetation variables are established using measurements from the two sites. The simulated horizontal flux turns out to be higher for a cropland than for a rangeland by approximately a factor 1.5, implying that increasing Sahelian cropped areas would increase dust emissions from the Sahel. This difference is mainly due to the time shift between grass and millet growth: the latter starts growing about 2–3 weeks later than annual grass. The amount of dry vegetation remaining during late dry season is also important for Sahelian wind erosion.