Analyzing the Inversion Performance of a Permanent Urban Column GHG Network: An OSSE Perspective

Abstract

Observations of atmospheric columns offer an effective approach to monitoring greenhouse gas (GHG) emissions, as they are less sensitive to the dynamics of atmospheric transport in comparison to in situ measurements. MUCCnet, the world's first permanent urban ground-based column network, has been utilized as an innovative method for measuring column GHGs. We present here an observing system simulation experiment framework to characterize the behavior of this unique network in estimating urban CO₂ emissions. An assumed in situ tower-based network (AISTnet) is performed to improve our understanding of MUCCnet's observing performance. We conduct a set of Bayesian atmospheric inversions to validate the current network deployment strategy and analyze its sensitivity to large point sources (LPSs). From our base inversions, we found overall good consistency between MUCCnet and AISTnet inversions, with nearly all grid cells showing corrections in the same direction during the inversions. While the sensitivities of in situ CO₂ synthetic observations are approximately an order of magnitude higher than those of column measurements, the column measurements have the advantage of broader coverage. This leads to larger uncertainty reduction around the site locations in the AISTnet inversions, while the MUCCnet inversions present larger values over the area away from the network. The inaccurate information of the LPSs provided in the prior estimate can adversely impact the estimated emissions. Our results suggest that MUCCnet is less sensitive to LPSs errors compared to AISTnet. The findings of this work can contribute valuable insights for advancing future observing strategies in an urban environment.