Using Space-Based CO2 and NO2 Observations to Estimate Urban CO2 Emissions

As the majority of fossil fuel carbon dioxide (CO2) emissions originate from cities, the use of novel techniques to leverage available satellite observations of CO2 and proxy species to constrain urban CO2 is of great importance. In this study, we seek to empirically determine relationships between satellite observations of CO2 and the proxy species nitrogen dioxide (NO2), applying these relationships to NO2 fields to generate NO2-derived CO2 fields (NDCFs) from which CO2 emissions can be estimated. We first establish this method using simulations of CO2 and NO2 for the cities of Buenos Aires, Melbourne, and Mexico City, finding that the method is viable throughout the year. For the same three cities, we next calculate empirical relationships (slopes) between co-located observations of NO2 from the Tropospheric Monitoring Instrument and Snapshot Area Mode observations of CO2 from Orbiting Carbon Observatory-3. Applying varying combinations of slopes to generate NDCFs, we evaluate methodological uncertainties for each slope application method and use a simple mass balance method to estimate CO2 emissions from NDCFs. We demonstrate monthly urban CO2 emissions estimates that are comparable to emissions inventory estimates. We additionally prove the utility of our method by demonstrating how large uncertainties at a grid cell level (equivalent to similar to 1-3 ppm) can be reduced substantially when aggregating emissions estimates from NDCFs generated from all NO2 swaths (about 1%-6%). Rather than rely on prior knowledge of emission ratios, our method circumvents such assumptions and provides a valuable observational constraint on urban CO2 emissions.