ASSIMILATION OF CYCLOPES LEAF AREA INDEX INTO THE ISBA-A-GS LAND SURFACE MODEL OVER FRANCE

摘要

In our study the SURFEX modeling platform is used in off-line mode. It contains the land surface model ISBAA- gs that simulates photosynthesis and plant growth. A version of the Extended Kalman Filter scheme is used for the assimilation in situ and CYCLOPES LAI product derived from the SPOT4/VEGETATION sensor within the ISBA-A-gs model. The LAI data is incorporated into the model every 10 days to analyse its impact on vegetation biomass. The experiment is performed over France at a spatial resolution of 8 km. A model grid box is divided in a number of patches each having its own set of prognostic variables. The filter algorithm is designed to provide the analysis for each patches independently by using one observation per grid box. When needed, the updated values are aggregated by computing a weighted average. In this study we demonstrate that the assimilation scheme works effectively within the multi-patch version of our model. An important reduction of the LAI bias between the model and measurements is achieved.