Characterizing OCO-2 XCO2 Variability Across Land-Ocean Boundaries in the Continental United States

Abstract:

Carbon monitoring and mitigation as well as Earth system modeling require surface CO₂ fluxes to be accurately and robustly quantified. Atmospheric inversion studies that aim to quantify surface CO₂ fluxes call for descriptions of atmosphere CO₂ variations, which are becoming increasingly documented by satellite CO₂ observations. NASA’s Orbiting Carbon Observatory 2 (OCO-2) provides global coverage of total column-averaged CO₂ measurements (XCO₂), which are less sensitive to vertical mixing and exchange with the free troposphere than surface CO₂ data. Differences between CO₂ variability in continental and marine air have been observed and quantified by surface measurement networks, but less is known about this transition throughout the total air column. Using OCO-2 XCO₂, we investigate land-ocean boundaries to better understand the transition between continental air mass XCO₂ and background maritime air mass XCO₂. Our research characterizes spatial and temporal variability in XCO₂ across these land-ocean atmospheric boundaries along the east and west coastlines of the continental United States. The analysis is split into one-degree latitudinal bands to compare onshore to offshore XCO₂ populations. A series of anomalies are computed to investigate XCO₂ variability over a range of spatial gradients and timescales from seasonal to synoptic. Preliminary findings indicate differences in spatial means, synoptic variability, and seasonal cycle amplitude and phasing between land and ocean XCO₂ populations that show a latitudinal dependence. Differences in XCO₂ variance imparted by pointing mode (glint & nadir) are significant. The impact of instrument and algorithm noise as well as small spatial-scale geophysical signals are evaluated by averaging adjacent retrievals along-orbit and found to impart a substantial fraction of the variability in XCO₂ data analyzed. Further work can focus on attributing the observed variability to real signals either from surface fluxes and atmospheric transport or residual bias in OCO-2 soundings. This analysis could help identify real carbon cycle and atmosphere-driven geophysical variation as well as measurement biases over land and ocean, which have different surface reflectance properties, and improve satellite retrieval techniques.