Evaluation and improvement of the daily boreal ecosystem productivity simulator in simulating gross primary productivity at 41 flux sites across Europe.

Vegetation gross primary productivity (GPP) is an important component in the global carbon cycle and its accurate estimation is essential in ecosystem monitoring and simulation. Previous studies show that ecosystem models usually overestimate GPP under drought and during spring, late fall and winter. In this study, these issues are addressed in the daily boreal ecosystem productivity simulator (BEPSd) by introducing a new water stress factor ( f w ) to replace the old one and a designed fraction in term of the normalised difference vegetation index (NDVI) ( f ndvi ) to indicate the effect of chlorophyll on photosynthesis. GPP simulations are conducted at 41 flux sites across Europe to test BEPSd with the new f w and f ndvi . The new f w captures drought conditions well and f ndvi expresses the chlorophyll constraint on photosynthesis. Although BEPSd with the old f w performs well for some plant function types (PFTs), it is unsatisfactory for others. BEPSd incorporating both the new f w and f ndvi gives better simulations than the old version, particularly for evergreen broadleaf forest, deciduous broadleaf forest and closed shrub with R (RMSE) value increasing (decreasing) from 0.69 (3.20 gCm −2 d −1 ) to 0.74 (1.65 gCm −2 d −1 ), 0.72 (4.01 gCm −2 d −1 ) to 0.82 (2.91 gCm −2 d −1 ), 0.54 (1.82 gCm −2 d −1 ) to 0.75 (1.59 gCm −2 d −1 ), respectively. Furthermore, the new f w effectively mitigates GPP overestimates under drought, and f ndvi counteracts GPP overestimates during spring, late fall and winter. Overall, the improved BEPSd shows a satisfactory performance at flux sites over Europe. [ABSTRACT FROM AUTHOR]