Please login first
Estimation of Crop Production, CO2 Fluxes and Carbon Budget Using Remote Sensing: Application to Winter Wheat/Sunflower Rotations

A recent study has proposed and validated an approach based on the use of an agro-meteorological model assimilating remote sensing products, to simulate development, production, CO2 fluxes and thus carbon budget for winter wheat in southwest France. This model (SAFY-CO2) was validated thanks to data acquired at two instrumented ICOS sites and close to 16 fields in our area of study. It was able to accurately reproduce biomass and yield (rRMSE of 27% and 21%, respectively) as well as the daily net CO2 flux (RMSE of 1.29 gC.m-2.d-1) allowing to estimate accurate net annual ecosystem carbon budget (NECB). The use of remote sensing products (green area index) drastically limits the need for ground measurement or information about agricultural practices. This specificity allows the model to be upscale easily and to simulate crop rotations without any in situ information other than that provided by satellites.

Based on these results, we propose a further analysis allowing to estimate the carbon budget of cropland over several years. In this study, the model has been adapted to run in multi-year and applied at the regional scale in southwest France. Besides winter wheat, sunflower has also been implemented in the model in order to simulate winter wheat/sunflower rotations, which are the two main cultivated crops in the study area.

The model showed good performances at estimating sunflower biomass and yield (RMSE of 66 and 55 g.m-2, respectively) as well as the net CO2 flux (RMSE of 0.97gC.m-2.d-1), which allows to simulate crop rotations.

In addition to remote sensing products from Sentinel 2 and Venμs satellites, we used the French Land Parcel Identification System which is the delimitation of French agricultural lands as well as the SAFRAN re-analysis meteorological data. This work allows to quantify the biogeochemical impact of certain crop rotation as well as agricultural practices. Indeed, the proposed approach accounts for potential vegetation development during fallow periods (weeds, cover crops) and scenarii with and without export of wheat straws as well as their impact on the NECB will be discussed.

Keywords: crop modelling; remote sensing; carbon budget; croplands