Toward the Use of the MODIS ET Product to Estimate Terrestrial GPP for Nonforest Ecosystems

Moderate Resolution Imaging Spectroradiometer (MODIS) gross primary production (GPP) data (MOD17), based on the light-use-efficiency algorithm, have been widely used to assess large-scale carbon budgets. However, systemic errors of this product have been reported, particularly for nonforest ecosystems. Here, we test a simple and operational way to estimate GPP in nonforest ecosystems by inverting the MODIS evapotranspiration (ET) product (MOD16) using ecosystem water use efficiency (WUE = GPP/ET) . Field measurements from 17 nonforest AmeriFlux sites of GPP were used for validation. Results show that the inverted GPP from MOD16 (MOD16 GPP) agrees better with the observed GPP than MOD17 does. The overall root-mean-square error (RMSE) and mean bias of MOD16 GPP are 19.63 g C/m²/8-day and -4.06 g C/m² /8-day, respectively, which are lower than the corresponding values of MOD17 GPP ( RMSE = 23.82 g C/ m²/8-day and mean bias = -9.07 g C/m²/8-day). This finding suggests the potential to achieve a better assessment of GPP for nonforest ecosystems with a fine resolution.