Modeling regional cropland GPP by empirically incorporating sun-induced chlorophyll fluorescence into a coupled photosynthesis-fluorescence model

Abstract:

Global sun-induced chlorophyll fluorescence (SIF) retrievals are currently available from several satellites. SIF is intrinsically linked to photosynthesis, so the new data sets allow to link remotely-sensed vegetation parameters and the actual photosynthetic activity of plants. In this study, we used space measurements of SIF together with the Soil-Canopy Observation of Photosynthesis and Energy (SCOPE) balance model in order to simulate regional photosynthetic uptake of croplands in the US corn belt. SCOPE couples fluorescence and photosynthesis at leaf and canopy levels.

To do this, we first retrieved a key parameter of photosynthesis model, the maximum rate of carboxylation (Vcmax), from field measurements of CO2 and water flux during 2007-2012 at some crop eddy covariance flux sites in the Midwestern US. Then we empirically calibrated Vcmax with apparent fluorescence yield which is SIF divided by PAR. SIF retrievals are from the European GOME-2 instrument onboard the MetOp-A platform. The resulting apparent fluorescence yield shows a stronger relationship with Vcmax during the growing season than widely-used vegetation index, EVI and NDVI. New seasonal and regional Vcmax maps were derived based on the calibration model for the cropland of the corn belt. The uncertainties of Vcmax were also estimated through Gaussian error propagation. With the newly derived Vcmax maps, we modeled regional cropland GPP during the growing season for the Midwestern USA, with meteorological data from MERRA reanalysis data and LAI from MODIS product (MCD15A2). The results show the improvement in the seasonal and spatial patterns of cropland productivity in comparisons with both flux tower and agricultural inventory data.