Impact of LAI assimilation by LDAS-Monde on modelled photosynthesis and respiration in the ISBA land surface model

Land Data Assimilation Systems (LDASs) aim to monitor the evolution of land surface variables (LSVs). They were initially designed with a focus on soil moisture and temperature. Since then, the focus has increasingly expanded towards vegetation monitoring with the assimilation of Leaf Area Index (LAI) or other vegetation-related data. LDAS-Monde, the offline land data assimilation system (LDAS) developed by Météo-France’s research centre (CNRM), has been a pioneer in that domain, as it can assimilate LAI while updating directly soil moisture especially in the root-zone area. This approach has also demonstrated several times its ability to improve the simulation of Gross Primary Production (GPP) by the ISBA (Interactions Soil-Biosphere-Atmosphere) land surface model, included in LDAS-Monde.

In this work, the impact of assimilating LAI on GPP and net ecosystem exchange (NEE) is assessed with two versions of the ISBA land surface model: the classical ISBA A-gs involved currently in LDAS-Monde and the more sophisticated ISBA-Carbon Cycle (ISBA-CC) version. ISBA A-gs simulates the assimilation of carbon by photosynthesis following the work of Goudriaan et al. (1985) and Jacobs et al. (1996) while the ecosystem respiration is emulated with a simple Q₁₀ formulation (Rivalland et al., 2005). ISBA-CC uses the same approach for photosynthesis but by modelling all biomass reservoirs (such as roots or wood) can calculate more accurately autotrophic respiration. ISBA-CC also involves a heterotrophic respiration calculated by a soil carbon module. The comparison and the impact assessments are carried out at site levels using the PLUMBER2 datasets and at larger spatial scales using a new version of FLUXCOM GPP and NEE products.