A comparison of five high-resolution spatially-explicit, fossil-fuel, carbon dioxide emission inventories for the United States

The quantification of fossil-fuel-related emissions of carbon dioxide to the atmosphere is necessary in order to accurately represent carbon cycle fluxes and to understand and project the details of the global carbon cycle. In addition, the monitoring, reporting, and verification (MRV) of carbon dioxide emissions is necessary for the success of international agreements to reduce emissions. However, existing fossil-fuel carbon dioxide (FFCO2) emissions inventories vary in terms of the data and methods used to estimate and distribute FFCO2. This paper compares how the approaches used to create spatially explicit FFCO2 emissions inventories affect the spatial distribution of emissions estimates and the magnitude of emissions estimates in specific locales. Five spatially explicit FFCO2 emission inventories were compared: Carbon Dioxide Information and Analysis Center (CDIAC), Emission Database for Global Atmospheric Research (EDGAR), Fossil Fuel Data Assimilation System (FFDAS), Open-source Data Inventory for Anthropogenic CO2 (ODIAC), and Vulcan. The effects of using specific data and approaches in the creation of spatially explicit FFCO2 emissions inventories, and the effect of resolution on data representation are analyzed using graphical, numerical, and cartographic approaches. We examined the effect of using top-down versus bottom-up approaches, nightlights versus population proxies, and the inclusion of large point sources. The results indicate that the approach used to distribute emissions in space creates distinct patterns in the distribution of emissions estimates and hence in the estimates of emissions in specific locations. The different datasets serve different purposes but collectively show the key role of large point sources and urban centers and the strong relationship between scale and uncertainty.