Tracking the seasonal and inter-annual variations of global gross primary production during last four decades using satellite near-infrared reflectance data

Terrestrial vegetation absorbs approximately 30% of the anthropogenic carbon dioxide (CO2) emitted into the atmosphere through photosynthesis (represented by gross primary productivity, GPP) and thus effectively mitigates global warming. However, large uncertainties still remain in the global GPP estimations and their long-termtrends. Here we used the satellite-based near-infrared reflectance (NIRv) as the proxy of GPP and generated a global long-term (1982-2018) GPP datasets (hereafter GPP(NIRv)). Analysis at the site-level showed that NIRv could accurately capture both the monthly and annual variations in GPP (R-2 = 0.71 and 0.74 respectively) at 104 flux sites. Upscaling the relationships between NIRv and GPP to the global scale, the global annual GPP was estimated to be 128.3 +/- 4.0 Pg C yr(-1) during the last four decades, which fell between the estimations from the machine-learning upscaling approach, light-use-efficiency (LUE) models and processed-based models. The seasonal variation of GPP(NIRv) was also consistent with those from flux sites and models. More importantly, the inter-annual trends in GPP(NIRv) during the last four decades were consistent with those from processed-based models across latitudes, which outperformed the other GPP products. This evidence suggested that the long-term GPP datasets derived from NIRv have better abilities to capture the seasonal and inter-annual variations of terrestrial GPP at the global scale. The long-term GPP(NIRv) product could be beneficial for the estimation of terrestrial carbon fluxes and for the projection of future climates. (C) 2020 Elsevier B.V. All rights reserved.

Automated summary

Researchers used satellite data to generate a global long-term GPP dataset and found that NIRv can accurately capture the seasonal and annual variations of GPP. Analysis at 104 flux sites showed that this dataset can effectively estimate global vegetation primary productivity.