Sensitivity of BRDF, NDVI and Wind Speed to the Aerodynamic Roughness Length over Sparse Tamarix in the Downstream Heihe River Basin

The aerodynamic roughness length (z0m) is an important parameter that affects the momentum and energy exchange between the earth surface’s and the atmosphere. In this paper, the gradient wind speed data that were observed from May to October, 2014, at the Si Daoqiao station, which is located in an area of sparse Tamarix in the downstream region of the Heihe River Basin (HRB), are used to evaluate the sensitivity of the moderate-resolution imaging spectroradiometer (MODIS) near-infrared (NIR) bi-directional reflectance distribution function (BRDF) R, the MODIS/Landsat 8 normalized difference vegetation index (NDVI), and the wind speed at 5 m in comparison with the field-measured z0m. The results indicate that the NIR BRDF_R_MODIS and the NDVIMODIS/NDVILandsat8 are less sensitive indicators of the z0m over sparse Tamarix areas (R2: 0.0045 for NIR BRDF_RMODIS; R2: 0.0342 for NDVIMODIS; and, R2: 0.1646 for NDVILandsat8), which differs significantly from the results obtained by previous studies for farmlands and grasslands. However, there is a nearly linear correlation between the wind speed at 5 m and the z0m at the time scale of the NDVILandsat8 acquisitions (R2: 0.3696). Furthermore, the combination of the NDVI and wind speed at 5 m can significantly improve this correlation (R2: 0.7682 for NDVIMODIS; R2: 0.6304 for NDVILandsat8), whereas the combination of the NIR BRDF_RMODIS and wind speed at 5 m still has a low correlation (R2: 0.0886). Finally, the regional z0m of the oasis in the downstream region of the HRB was determined using Landsat 8 surface reflectance data and the wind speed data at the Si Daoqiao station, which properly reflect the temporal evolution of the z0m in that region. The parameterization scheme proposed in this paper has great potential to be applied to evapotranspiration, land surface, and hydrologic model simulations of sparse Tamarix at Si Daoqiao site in the future.