Assessing regional anthropogenic emissions from observations of atmospheric CO2

Monday, 15 December 2014: 8:45 AM
Anna M Michalak1, Jaideep Ray2, Yoichi Paolo Shiga3,4 and Vineet Yadav4, (1)Carnegie Institution for Science, Washington, DC, United States, (2)Sandia National Laboratories, Albuquerque, NM, United States, (3)Stanford University, Stanford, CA, United States, (4)Carnegie Institution for Science, Stanford, CA, United States
Atmospheric inverse modeling has been used to constrain biospheric and oceanic CO2 flux budgets for two decades. Recently, the idea of using such tools for quantifying anthropogenic CO2 emissions has also emerged. To date, most such applications have been at local to urban scales, which represent an important component of anthropogenic emissions but cannot be used directly to understand or verify regional to national scale budgets.

This presentation will bring together a number of efforts to address this gap through the development and application of inversion tools tailored to anthropogenic emissions. Rather than relying on isotopic signatures or co-emitted tracers, these studies are based on characterizing and leveraging the distinct spatiotemporal patterns of fossil fuel emissions.

The first set of studies assesses the extent to which fossil fuel emissions are discernable given current in situ and realistic satellite-based observations of atmospheric CO2. The second develops a wavelet-based parameterization that enables inversions that can resolve the fine-scale features characteristic of fossil fuel emissions from the diffuse information provided by atmospheric observations. The final set focuses on statistical strategies for disaggregating the fossil fuel and biospheric components of terrestrial CO2 fluxes.

Taken together, these studies indicate that a number of confounding factors make the detection and attribution of regional-scale fossil fuel emissions very challenging using the current in situ monitoring network as well as satellite-based observations, especially during the growing season. They further indicate, however, that inversion approaches that leverage the distinct spatiotemporal statistical features of biospheric and anthropogenic emissions can be used to disaggregate these two types of fluxes directly from CO2 observations, given sufficient network sensitivity.