Comparison of global high-resolution fossil fuel CO2 emissions data products to Vulcan v4.0: sector differences, urban geographies, and methodological guidance

New global greenhouse gas emission products have emerged in recent years providing emissions estimation at increasing fine space/time scales. Furthermore, these efforts are moving away from traditional forms of proxy linear downscaling and toward the use of machine learning and integration of some forms of “bottom-up” data. In terms of application, there is interest in applying these data products to the city-scale, assisting and supporting mitigation activities in the global urban governance level.

 

However, it is also acknowledged that developing these very high-resolution efforts at the global scale come with particular challenges associated with data availability, method limitations, and data quality variations. Here we use a very high-resolution data product developed in the United States, the Vulcan version 4.0 emissions, as a point of comparison with two of the new global very high-resolution efforts: Climate Trace, and GRACED. The ‘Vulcan Project’ is an effort to compute bottom-up CO2 emissions from fossil fuel combustion (FFCO2) and cement production for the entire USA. Vulcan v4.0 quantifies emissions from 2010 to 2021 for multiple sectors to the point, line, and polygon spatial scale.

 

We use detailed comparison with Vulcan to illuminate and inform aspects of the global efforts that many warrant further investigation or methodological development. We upscale Vulcan to match the resolution of the global data products and aggregate as necessary to isolate sectoral matches. Using statistical analysis techniques we isolate differences that may be systematic and explainable via alternative methodologies and/or data sources. Our aim is to strengthen and improve all high-resolution efforts at multiple scales and recommend where scale limitation may exist.