OCO-2 Satellite-imposed Constraints on Terrestrial Biospheric CO2 Flux over South Asia

The spatiotemporal variability of terrestrial biospheric carbon dioxide (CO₂) flux over South Asia is poorly understood. The inverse model "top-down" CO₂ flux estimates which rely on atmospheric CO₂ observational data are impeded by sparse in situ measurements over this region. The Orbiting Carbon Observatory 2 (OCO-2) provides much needed global retrievals of column-average CO₂ dry-air mole fraction (XCO₂) with the finest spatial and temporal resolution and highest sensitivity to surface fluxes available to-date. This study conducted global inverse model simulations as part of the second OCO-2 Multi-model Intercomparison Project assimilating version 9 of the OCO-2 XCO₂ retrievals in land nadir (LN) and land glint (LG) observing modes and global in situ (IS) measurements. The four-dimensional variational assimilation system with the GEOS-Chem global chemical transport model was used to estimate CO₂ fluxes from 2015 to 2018. We then assessed the spatiotemporal variability of optimized CO₂ Net ecosystem Exchange (NEE) fluxes aggregated over the South Asia region. The most robust result found by assimilating OCO-2 observations was the constraints imposed on the seasonal cycle of terrestrial biospheric fluxes over South Asia. The seasonality of South Asian NEE estimated by assimilating OCO-2 or global IS data showed a larger seasonal cycle compared to the current understanding of NEE in this region (represented by the prior NEE used in the model). The satellite-in situ joint inversion (IS + LN + LG) led to land flux seasonal amplitude (absolute magnitude of the difference between peak and trough of monthly mean values over a year) of 4.0 PgC yr^-1, compared to the prior model (1.5 PgC yr^-1). Moreover, OCO-2 data imposed a phase shift in the seasonal cycle, resulting in a large CO₂ source in April and a large uptake in September. Most of the features in the seasonal cycle imposed by OCO-2 data are in agreement with previous "top-down" studies assimilating regional aircraft observations. Nevertheless, we recommend a closer examination of the novel features highlighted by the OCO-2 column data over South Asia in future studies by conducting regional inverse modelling with denser in situ and vertically-resolved regional observations along with satellite data.