Exploring the Synergistic Use of OMI NO2 Data with CO2 Data Collected from GOSAT and OCO-2: Current Status, Challenges and Implications for Upcoming Carbon Missions

Carbon dioxide (CO₂) emissions from fossil fuel combustion (Fossil fuel CO₂ emissions, FFCO2) need to be curbed to achieve the global climate mitigation goal defined under the Paris Agreement. FFCO2 is primarily derived from “bottom-up” emission inventories (EIs), which are thought to be accurately calculated at national level. However, EIs are fundamentally prone to systematic biases due to the calculation method (emission factors x activity data) and they need independent evaluation (e.g., “top-down” estimates from satellite data). Assuring the accuracy of emission estimates is crucial for successful implementation of United Nations Framework Convention on Climate Change (UNFCCC) under the Paris Agreement. Current missions such as Japan’s Greenhouse gas Observing SATellite (GOSAT, 2009-current) and NASA’s Orbiting Carbon Observatory-2 (OCO-2, 2012-current) have routinely collected column CO₂ data globally and demonstrated the observing capability of CO₂ enhancement signatures from large localized sources such as cities and power plants.

This study explores the synergistic use of NO₂ tropospheric columns from the Ozone Monitoring Instrument (OMI) on NASA’s Aura and CO₂ data collected from GOSAT and OCO-2 to better quantify FFCO2. Estimating FFCO2 from the CO₂ data alone is challenging because CO₂ is a long-lived gas and has a high background. NO₂ is co-emitted with CO₂, so it serves as an excellent tracer of fossil fuel combustion sources of CO₂. Furthermore, the ratio of NO₂ to CO₂ can help us studying the characteristics of the combustion technologies and control devices (e.g. combustion efficiency) that are often poorly represented in EIs.

One of the major challenges in this study is the very limited amount of CO₂ data collected over localized sources. While OMI currently has a 2-day revisit time, OCO-2, which is also in the A-train constellation of satellites, collects data with a narrow swath (1.29x2.25km footprint x 8) with a 16 day revisit time. With the lack of a pointing capability, OCO-2 does not routinely collect data for specific sources. GOSAT on the other hand collects data over and near major localized sources globally (3 day revisit) with a target mode observation, but its relatively large footprint (10 km in diameter) limits the data yield due to the clouds (GOSAT and OCO-2 retrievals require a cloud-free condition). Thus, a simple colocation approach does not yet provide robust statistics of soundings over localized source.

To maximize the number of matched NO₂/CO₂ soundings available for this study, we used a global Eulerian-Lagrangian coupled atmospheric transport model (GELCA) driven by a global high-resolution FFCO2 emissions (ODIAC). Based on the model simulation, we identified CO₂ soundings that record CO₂ enhancement due to FFCO2 and exclude biogenic contributions. We then calculated the ratios of the collocated NO₂ and GOSAT-CO₂ data during 2009-2014 over aggregated large continental regions (e.g., North America and East Asia) and examined them by comparing to ratios derived from available gridded EIs. In our presentation, we will also show the results with OCO-2 data and discuss the current challenges and difficulties. We will also discuss targeting strategies for upcoming satellite carbon missions (e.g. OCO-3) to enhance their contribution to global carbon monitoring in combination with OMI data.