Fellowship Project Title: Carbon emissions and uptake from vegetation change in the tropics

Fellowship project summary:

Vegetation in the tropics is increasingly exposed to direct (e.g. land clearing) and indirect (e.g. rising atmospheric CO2 concentrations, increasing drought frequency and intensity) anthropogenic pressure. Thus far, land-use and land-cover changes, often tightly linked to fire, have been a major contributor to the pantropical carbon (C) budget. Fire and non-fire mediated emissions are poorly quantified though, in spite of the increasing availability of medium-to-high-resolution Earth Observation products of burned area and deforestation. Likewise, the extent of C uptake by the recovering vegetation is associated with large uncertainties. These uncertainties are likely contribute to the imbalance of the global carbon cycle, and initiatives aimed at closing the global and regional C budgets will need updated values of fire and non-fire mediated vegetation C emissions and uptake.

Several ongoing or upcoming spaceborne missions (e.g. GEDI, NISAR and ESA-BIOMASS) are expected to greatly improve our understanding of biomass spatial distribution and dynamics. Passive microwave systems like the Soil Moisture and Ocean Salinity (SMOS) sensor have shown interesting capabilities for C stock retrieval: the SMOS-derived L-VOD (L-band vegetation optical depth) vegetation index is correlated to aboveground biomass (AGB; of which ca. half is carbon), showing low signal saturation for AGB up to ca. 400 t/ha. In addition, its short revisit period and long time series makes it a strong candidate to study biomass dynamics.

This project aims to attribute observed changes in aboveground C stocks to processes such as vegetation clearing or regrowth across the tropics over the last decade. In particular, the project will separate fire from non-fire mediated emissions. To do this the project will use a locally-calibrated and spatially explicit “bookkeeping-like” approach. First, we will use an annual time series of aboveground C maps, obtained from the calibration of the SMOS-derived L-VOD vegetation index with reference biomass maps such as the biomass CCI map for 2017 and the first release of the GEDI-derived map of AGB for 2019, to compute changes in aboveground C stocks at 25km spatial resolution, the spatial resolution of the SMOS signal. Products with medium-to-high spatial resolution ranging from 300 down to 30m, such as land cover, fire and tree cover loss, to partition C emission and uptake within the SMOS pixels will be used. Higher-resolution products will also be mobilized where available, mostly for South America, to test the impact of sensor type and product resolution on carbon attribution. These products include canopy height models from repeated airborne LiDAR acquisitions, and Sentinel-1-derived products specifically designed to detect either fire or vegetation disturbance.

This project will work towards delivering maps of various emission and uptake components of the C budget in the tropics over 2010-2019, with possible extension to 2020-2021 as data become available; and produce updated estimates of fire and non-fire mediated C emissions. The project will work in close collaboration with members of the modeling community to ensure that project outputs meet their needs.

Research Fellow: Nicolas Labrière

Host Institution: Le Centre national de la recherche scientifique