3. Correct: We compared two methods to correct BC biases. We found that directly quantifying BC concentrations is more effective and computationally tractable than correcting emissions outside the region of interest.

2. Quantify: We derived two metrics to estimate errors induced by BCs: the preview estimates errors before completing the analysis to support domain specification, while the diagnostic estimates errors after the analysis to improve error quantification.

This fraction is determined by the boundary conditions (BCs). The paper works to understand, quantify, and correct errors induced by BCs:

1. Understand: We explored how BCs impact emission estimates as a function of emission and observation uncertainty and of wind speed.

Quantifying regional GHG emissions (e.g., from cities or oil/gas basins) is important to mitigation. We studied a significant source of uncertainty in such analyses: the fraction of atmospheric concentrations from sources inside vs. outside of the region.

It's methane emissions! (That is, we report emissions as mass methane.)

There are a lot of numbers here, but I hope that this work gives a bit of hope and direction to efforts to decrease methane emissions. We can do a lot by focusing on the largest emitters! (If you’re interested in any of this in more detail, please read the paper or contact me.)

4. We calculate methane emissions for 95 urban areas. We find that these urban areas emit 20% of human-caused methane in the contiguous US. Their emissions are 38% larger than a gridded version of the EPA Inventory. We think the change may be from landfill and gas emissions.

Texas alone produces 21% of contiguous US emissions in our estimate! The Permian Basin explains almost 40% of Texas’ emissions. California, the second largest methane-producing state, generates 7% of contiguous US methane emissions.

3. We compare our emissions to state estimates from the EPA Inventory. We find that the top 10 methane-producing states are responsible for 55% of human-caused emissions in the contiguous US. Our emissions are on average 27% larger than the EPA state inventories in these states.

We attribute the change in landfill emissions to overestimated recovery efficiencies at landfill gas facilities and to under-accounting of site-specific operational changes and leaks.

2. Our landfill emissions are 51% larger than EPA’s Inventory. At 70 landfills that report to EPA’s Greenhouse Gas Reporting Program—a key input to the Inventory—we find a 77% median increase from reported emissions (and a 204% median increase at gas recovery facilities)!

1. Our emission estimate is 13% higher than the EPA’s 2023 Greenhouse Gas Inventory for 2019. We find that oil and gas, livestock, and landfills explain 89% of human-caused emissions in the contiguous US.