Abstract
The apparent optimum air temperature for vegetation photosynthesis (T(opt)) is a key temperature parameter in terrestrial ecosystem models estimating daily photosynthesis or gross primary production (GPP, g C/m(2)/day). To date, most models use biome-specific T(opt) (T(opt-biome)) parameter values. Given vegetation acclimation and adaptation to local climate, site-specific T(opt) (T(opt-site)) is needed to reduce uncertainties in estimating daily GPP across the scales from site to region and the globe. Previous studies have demonstrated using the Enhanced Vegetation Index (EVI) derived from Moderate Resolution Imaging Spectroradiometer (MODIS) images and daytime air temperature data to estimate the T(opt-site) at the eddy covariance tower sites. This study used MODIS-derived EVI and ERA5 climate data to estimate and generate global T(opt-site) data products from 2000 to 2019. The T(opt-site) of individual pixels within a biome has large variation, which clearly cannot be represented accurately by the widely used T(opt-biome). Therefore, using this global dataset of T(opt-site) estimates might significantly affect GPP simulation in current ecosystem models.