Estimation of Carbon Budgets for Croplands by Combining High Resolution Remote Sensing Data with a Crop Model and Validation Ground Data

Abstract:

Croplands occupy large areas of Earth’s land surface playing a key role in the terrestrial carbon cycle. Hence, it is essential to quantify and analyze the carbon fluxes from those agro-ecosystems, since they contribute to climate change and are impacted by the environmental conditions. In this study we propose a regional modeling approach that combines high spatial and temporal resolutions (HSTR) optical remote sensing data with a crop model and a large set of in-situ measurements for model calibration and validation. The study area is located in southwest France and the model that we evaluate, called SAFY-CO₂, is a semi-empirical one based on the Monteith’s light-use efficiency theory and adapted for simulating the components of the net ecosystem CO₂ fluxes (NEE) and of the annual net ecosystem carbon budgets (NECB) at a daily time step. The approach is based on the assimilation of satellite-derived green area index (GAI) maps for calibrating a number of the SAFY-CO₂ parameters linked to crop phenology. HSTR data from the Formosat-2 and SPOT satellites were used to produce the GAI maps. The experimental data set includes eddy covariance measurements of net CO₂ fluxes from two experimental sites and partitioned into gross primary production (GPP) and ecosystem respiration (R_eco). It also includes measurements of GAI, biomass and yield between 2005 and 2011, focusing on the winter wheat crop. The results showed that the SAFY-CO₂ model correctly reproduced the biomass production, its dynamic and the yield (relative errors about 24%) in contrasted climatic, environmental and management conditions. The net CO₂ flux components estimated with the model were overall in agreement with the ground data, presenting good correlations (R² about 0.93 for GPP, 0.77 for R_eco and 0.86 for NEE). The evaluation of the modelled NECB for the different site-years highlighted the importance of having accurate estimates of each component of the NECB. Future works aim at considering systematically post-harvest events (such as re-growths, weeds and intercrops) on NEE assessment and at assimilating radar remote sensing data for estimating GAI and biomass more accurately. This approach is currently being extended to summer crops and it could be applied to larger scales thanks to the recent satellite missions (Landsat-8, Sentinel-1 and 2…).