Estimation of the gross primary production of an old‐growth temperate mixed forest using eddy covariance and remote sensing

Abstract

Continuous flux data from CO2 flux sites can be used to improve our understanding of leaf phenology and validate the algorithms of satellite‐ based carbon cycling models. In this study, we conducted a simulation of the Vegetation Photosynthesis Model (VPM) using the Enhanced Vegetation Index (EVI) and the Land Surface Water Index (LSWI) derived from the 8‐day Moderate Resolution Imaging Spectroradiometer (MODIS) surface reflectance product, as well as site‐specific air temperature, biological temperature, and photosynthetically active radiation (PAR) data. Gross primary production (GPP) estimates derived from the VPM were compared with field observations of a flux tower in an old temperate mixed forest in northeastern China during 2003–2005. Time series data for the EVI have a stronger exponential relationship with the GPP (R 2 = 0.74, n = 67, p<0.01) than those for the Normalized Difference Vegetation Index (NDVI) (R 2 = 0.62, n = 67, p<0.01), indicating a different light use efficiency during the different stages of foliage development. In comparison to the flux tower GPP, the VPM‐predicted GPP captured the onset of the growing season well, and their seasonal dynamics were generally consistent in terms of phase in the peak growing season, while the end date of the growing season was 8–16 days earlier than that of field measurements. The annual forest GPP estimated from the flux tower observations varied from 1312 g C m−2 (grams of carbon per metre squared) to 1490 g C m−2 in the three observation years from 2003 to 2005, which is less 10% different from the VPM‐based annual GPP. These results demonstrate the potential of the satellite‐driven VPM for scaling up the GPP of forests at the CO2 flux tower site, a key issue for the study of the carbon budget at regional scales.