Nasa's GEOS-Carb Modeling and Assimilation System in Support of the OCO-2 Mission

The GEOS-Carb system at NASA’s Global Modeling and Assimilation Office (GMAO) leverages the GEOS family of models to deliver targeted modeling and assimilation products in support of the Orbiting Carbon Observatory-2 (OCO-2) mission. Recently as part of this system, we have developed 2 new capabilities: (a) assimilation of OCO-2 observations into the GEOS-5 constituent data assimilation system to produce full-coverage maps (Level 3 product) of CO₂, and (b) inversion of OCO-2 observations to estimate fluxes (Level 4 product) at high spatiotemporal resolutions (spatial: ~1° and temporal: daily). With the first 2 years of data available from the OCO-2 mission, the GEOS-Carb modeling and assimilation system is being used to examine the degree to which the OCO-2 total column observations (X_CO2): (a) can be used to estimate and predict in near real-time the global response to interannual changes in atmospheric transport, precipitation, and wildfires, (b) are constraining surface fluxes with reasonable precision and accuracy, and (c) providing additional information about carbon sources and sinks relative to the high precision but sparse in situ observational network.

The GEOS-Carb modeling and assimilation system is an extension of the GEOS-5 general circulation model and data assimilation system developed at NASA's Global Modeling and Assimilation Office (GMAO). In particular, the Level 3 (L3) system processes retrievals of column-average carbon dioxide (XCO2) based on near-infrared radiance measurements from the Orbiting Carbon Observatory 2 (OCO-2) - the resulting L3 maps are the most complete and highest resolution assimilated picture of global CO₂ yet. The extent of its coverage makes the L3 product a powerful tool for the evaluation of the satellite data against independent observations. The inverse modeling system, on the other hand, uses a separate ensemble square root filter to estimate global sources and sinks of CO₂. Results show that OCO-2 observations are able to capture the broad spatial and temporal patterns of flux estimates, and are especially beneficial over regions poorly constrained by the existing ground-based monitoring network (for e.g., South America). As the OCO-2 data matures, we will refine both the L3 and the L4 system and explore specific cases where we observe anomalies in the X_CO2 distribution, for example, the Tropical and the Southern Ocean sectors. Such anomalies indicate the possibility of previously unobserved fluxes, transport variability, and/or impact of the current El Nino conditions on land- /ocean-atmosphere carbon exchanges.