Assessment of Soil erosion parameters in Costa Rica using reflectance hyperspectral and simulated EnMAP imagery

Abstract

Soil erosion can be linked to relative fractional cover of photosynthetic-active vegetation, non-photosynthetic-active vegetation and bare soil, and terrain characteristics such as the C and LS factors. This study investigates the capacity of EnMAP imagery to map areas prone to soil erosion in a region near San Jose, Costa Rica, characterized by spatially extensive coffee plantations and grazing in a mountainous terrain. Simulated EnMAP imagery is based on airborne hyperspectral HyMap data. Fractional cover estimates are derived using a Multiple End-member Spectral Mixture Analysis approach. The C and LS factors are derived from a digital elevation model. The analysis was done independently for EnMAP and HyMap, the latter treated as reference data as no ground truth is available. Results demonstrate that with EnMAP imagery it is possible to extract quality endmember classes with important spectral features related to PV, NPV and soil, and be able to estimate relative cover fractions. This spectral information is critical to separate soil and NPV which greatly can impact integrated modelling with terrain characteristics. From a regional perspective EnMAP can be used to highlight specific areas that may be prone to erosion where this information can be extract directly from the imagery using automated processes.