XCO2 observations using satellite measurements with moderate spectral resolution: investigation using GOSAT and OCO-2 measurements

In light of the proposed space segment of Europe’s future CO₂ monitoring system, we investigate the spectral resolution of the CO₂ spectrometer, which measures earthshine radiance in the three relevant spectral bands at 0.76, 1.61, and 2.06 µm. The Orbiting Carbon Observatory-2 (OCO-2) mission covers these bands with fine spectral resolution but limited spatial coverage, which hampers the monitoring of localized anthropogenic CO₂ emission. The future European CO₂ monitoring constellation, currently undergoing feasibility studies at the European Space Agency (ESA), is targeting a moderate spectral resolution of 0.1, 0.3, and 0.3-0.55 nm in the three spectral bands with a high signal-to-noise ratio (SNR) as well as a spatial resolution of 4 km² and an across-track swath width >250 km. This spectral and radiometric sizing is deemed to be favorable for large-swath imaging of point sources of CO₂ emission. To assess this choice, we use real and synthetic OCO-2 satellite observations, which we spectrally degrade to the envisaged lower spectral resolution. We evaluate the corresponding CO₂ retrieval accuracy by taking the Total Carbon Column Observing Network (TCCON) observations as reference. Here, a lower spectral resolution enhances the scatter error of the retrieved CO₂ column mixing ratio (XCO₂) but has little effect on the station-to-station variation in the biases. We show that the scatter error gradually increases with decreasing spectral resolution. Part of the scatter error increase can be attributed to the retrieval noise error which can be compensated for by a future instrument with improved SNR. Moreover, we consider the effect of the reduced spectral resolution on the capability to capture regional XCO₂ variations and XCO₂ plumes from selected OCO-2 orbits. The investigation using measurements from the Greenhouse gases Observing SATellite (GOSAT) and synthetic measurements confirms our finding and indicates that one major source of uncertainties regarding CO₂ retrieval is the insufficient information on aerosol properties that can be inferred from the observations. We hence recommend the implementation of simultaneous, co-located measurements that have a larger information content on aerosols with an auxiliary instrument in the future European observing system.