Novel Application of NASA’s GEOS-CF CO2 Forecasting System to ACT-America Airborne Campaign

In order to stabilize the ever increasing amounts of CO₂, which significantly contribute to climate change, it is imperative to monitor and quantify these emissions globally. Unfortunately, separating anthropogenic emissions from the spatially and temporally variable uptake and release of CO₂ by the biosphere still remains challenging. Inverse models, which use atmospheric CO₂ observations to infer net flux, are a promising tool that may improve understanding of CO₂ sources and sinks. However, errors in atmospheric transport in such models currently limit the accuracy of the estimated fluxes they produce. NASA’s Atmospheric Carbon and Transport - America (ACT-America), comprised of five airborne campaigns spread out across three specific regions and four different seasons in the eastern U.S., was designed to better understand the transport and fluxes of atmospheric CO₂. During the summer 2019 campaign, we used new experimental CO₂ forecasts from NASA's Goddard Earth Observing System Composition Forecast (GEOS-CF) as a guide for flight planning. The GEOS-CF forecasts are initialized using fields produced by the GEOS meteorological data assimilation system, coupled to the GEOS-Chem chemistry module and are run at 25-km resolution, globally. Fluxes of CO₂ are estimated in near real time (NRT) based on a suite of remote sensing data including greenness, night lights, and fire radiative power. In this preliminary version, the NRT biospheric fluxes are determined by extrapolating the retrospective CASA-GFED fluxes forward in time. The ACT-America team used these GEOS-CF CO₂ forecasts to identify interesting transport patterns across different types of weather systems and then investigated several of these cases using research aircraft. The results provide several examples that illustrate the ability of high-resolution global models to predict CO₂ gradients at 2-3 days lead times and identify the distribution of CO₂ across different layers of the atmosphere. Preliminary results suggest that the biological CO₂ fluxes in the summer of 2019 diverged from our forecast due to extreme wetness in the Midwest. This comparison illustrates the promise of near-real time adjustment of flux estimates along with a potential insight regarding CO₂ transport uncertainties using atmospheric CO₂ observations.