Quantifying the economic opportunity for afforestation and reforestation in Maryland (USA) using high resolution carbon monitoring products

Dedicated satellite retrievals of total column CO₂ first became available with the launch of the GOSAT satellite in 2009, followed by NASA's Orbiting Carbon Observatory-2 in 2014. GOSAT also measures CH₄. Additional satellite missions to measure CO₂ and CH₄ are planned for the coming decade, including the NASA GeoCarb mission, which will provide greatly improved spatial and temporal coverage of the continental US and the Amazon compared to current missions. The signatures of regional scale biospheric and anthropogenic fluxes in total column retrievals are relatively small, and flux estimates that rely on remote sensing data can be confounded by systematic errors. Conversely, the usefulness of in situ point measurements is limited by sparseness in space, and, in the case of flask or aircraft sampling, in time. Interpretation of surface observations is further complicated by covariance of surface fluxes with diurnal and seasonal variations in planetary boundary layer height (the so-called rectifier effect). In addition, the footprint (i.e., the upwind surface influences) for a particular measurement location and type strongly depends on meteorological conditions, a problem that is especially important for sparse networks. Consequently, the best flux estimates will come from a rigorous combination of a variety of data types. Although North America is relatively well-observed compared to other regions, opportunities for direct comparisons between in situ measurements and satellite retrievals are extremely limited due to sparsity of both types of data. Here, we investigate the consistency among GOSAT, OCO-2 and TCCON retrievals, and simulated profiles of CO₂ and CH₄ generated using the high-resolution CarbonTracker-Lagrange inverse modeling framework. The CarbonTracker-Lagrange profiles are strongly constrained by in situ measurements over North America. Detailed and rigorous uncertainty estimates corresponding to each simulated profile enable meaningful comparisons with the retrievals. These comparisons represent an important step toward our long-term goal of developing robust data assimilation strategies that optimally leverage both in situ measurements and satellite retrievals and that can detect any systematic biases in remote sensing retrievals or in simulated atmospheric transport.