SIF-based hourly biosphere carbon fluxes for the Amazon from 2010-2020

The terrestrial tropics dominate the inter-annual variability of the global carbon cycle. As a significant component of the South American terrestrial tropics, the Amazon is a major contributor to both the long-term global terrestrial carbon sink and the inter-annual variations in the terrestrial biosphere’s sequestration of anthropogenic carbon dioxide (CO2) emissions. Accurately modeling the Amazon’s carbon response to anthropogenic disturbance at multiple timescales requires a comprehensive effort that includes developing, evaluating, and improving estimates of CO2 surface fluxes. We estimate hourly 12km×12km ecosystem carbon fluxes (partitioned into photosynthesis and respiration, in micromole CO₂ m^-2s^-1) over the Amazon, from 2010–2020. To quantify spatial and temporal variations in biospheric carbon activity, we have adapted the Vegetation, Photosynthesis, and Respiration Model (VPRM) to use Solar Induced Fluorescence (SIF), a parameter highly correlated with photosynthesis. SIF-based VPRM formulations (VPRM-SIF) entirely replace and show promise over the traditional VPRM indices such as the Enhanced Vegetation Index (EVI) and the Land Surface Water Index (LSWI). Where available, each ecosystem type in the model domain is calibrated with representative ground-based carbon flux measurements. The VPRM is scaled up to the region with ECMWF-ERA5 temperature and solar radiation. We compare our VPRM-SIF performance with the traditional VPRM as well as with a selection of recent biosphere flux models. We present the results from a case study focusing on March 2015 – the tail-end of the 2015/2016 severe El Niño – where we conduct a top-down optimization using OCO-2 CO₂ observations. The case study demonstrates the relative strengths and weaknesses of our Amazonia VPRM-SIF, while highlighting future opportunities to use the model in partitioning the region’s biogenic carbon fluxes and fire-related carbon emissions.