Jonas Ardö

This paper investigates how hyperspectral reflectance (between 350 and 1800 nm) can be used to infer ecosystem properties for a semi-arid savanna grassland in West Africa using a unique in situ-based multi-angular data set of hemispherical conical reflectance factor (HCRF) measurements. Relationships between seasonal dynamics in hyperspectral HCRF and ecosystem properties (biomass, gross primary productivity (GPP), light use efficiency (LUE), and fraction of photosynthetically active radiation absorbed by vegetation (FAPAR)) were analysed. HCRF data (rho) were used to study the relationship between normalised difference spectral indices (NDSIs) and the measured ecosystem properties. Finally, the effects of variable sun sensor viewing geometry on different NDSI wavelength combinations were analysed. The wavelengths with the strongest correlation to seasonal dynamics in ecosystem properties were shortwave infrared (biomass), the peak absorption band for chlorophyll a and b (at 682 nm) (GPP), the oxygen A band at 761 nm used for estimating chlorophyll fluorescence (GPP and LUE), and blue wavelengths (rho(412)) (FAPAR). The NDSI with the strongest correlation to (i) biomass combined rededge HCRF (rho(705)) with green HCRF (rho(587)), (ii) GPP combined wavelengths at the peak of green reflection (rho(518), rho(556)), (iii) LUE combined red (rho(688)) with blue HCRF (rho(436)), and (iv) FAPAR combined blue (rho(399)) and near-infrared (rho(1295)) wavelengths. NDSIs combining near infrared and shortwave infrared were strongly affected by solar zenith an-gles and sensor viewing geometry, as were many combinations of visible wavelengths. This study provides analyses based upon novel multi-angular hyperspectral data for validation of Earth-observation-based properties of semi-arid ecosystems, as well as insights for designing spectral characteristics of future sensors for ecosystem monitoring.