the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 27 Aug 2026
Structural divergence between satellite and surface CO observations across China, North America, and Europe (2015–2024)
Abstract. Satellite and surface observations of atmospheric carbon monoxide (CO) yield divergent emission estimates, yet the underlying causes of this discrepancy remain poorly understood. Here we present a multi-region analysis spanning 2015–2024 over China, North America, and Europe, assimilating MOPITT satellite retrievals and ground-based CO measurements from over 3300 monitoring stations. Our results reveal that the systematic underestimation of modeled surface CO is a common feature across all three continents, with median observation-to-model ratios of 1.82 (China), 1.65 (North America), and 2.25 (Europe). This bias progressively diminishes from urban to rural sites, by 44 % in North America and 58 % in Europe, and almost disappears at remote sites, demonstrating that representation error is an important driver of the surface-model mismatch. In contrast, comparisons of column CO between MOPITT and model show no systematic underestimation. Assimilation experiments further illuminate the structural nature of this divergence: while surface-data assimilation greatly improves agreement with in situ observations, it introduces a positive column bias relative to MOPITT. This bias is most pronounced over China and Europe but remains modest over North America, and accordingly the decadal trends from surface CO show a sharp decline in China (-5.92 % yr-1) vs. plateaus after 2018 and 2016 in North America and Europe, while satellite-derived column trends are substantially weaker. Our multi-region analysis demonstrates that the satellite-surface discrepancy is a structurally embedded feature of the observing system, and advocates for joint utilization of both observation platforms in future inversion studies to achieve robust and self-consistent CO emission estimates.
- Preprint
(3296 KB) - Metadata XML
-
Supplement
(856 KB) - BibTeX
- EndNote
Status: open (until 28 Oct 2026)
- RC1: 'Comment on egusphere-2026-4938', Anonymous Referee #1, 09 Oct 2026 reply
-
RC2: 'Comment on egusphere-2026-4938', Anonymous Referee #2, 15 Oct 2026
reply
Major comment
1. Separation of representation error from emission and transport errors:
In Sections 3.1.2–3.2, the authors suggest that site representation error is an important reason for the discrepancy between surface and satellite CO observations. Though the reduction in model bias from urban to remote stations supports this explanation, it does not clearly separate representation error from uncertainties in emissions and boundary-layer mixing. Underestimated urban emissions or excessive vertical mixing could also produce lower surface CO without a comparable underestimate in the total column. Moreover, the increase in column CO after surface assimilation does not independently establish the cause of the original bias. I suggest that the authors include a targeted resolution or sub-grid representativeness experiment for a selected period, together with an emission or mixing sensitivity test. This would help establish how much of the discrepancy can be attributed to representation error. Otherwise, the conclusions should acknowledge that these contributions have not been quantitatively separated.Minor comments
1. MOPITT observation operator — Sections 2.1 and 2.4:
Please clarify whether the assimilation uses MOPITT CO profiles or total columns. The retrieval equation is provided, but its application to the model profiles is not sufficiently described. The authors should explain how the averaging kernels and a priori profiles are applied, including the treatment of the logarithmic retrieval state. Also, please define how the column quantity reported in ppb is calculated and how its uncertainty is obtained from the retrieval covariance.2. Background-error covariance and vertical increments — Section 2.4 and Figure 6:
The construction and updating of the background-error covariance need further explanation, particularly the horizontal and vertical correlations. Please clarify whether surface assimilation directly modifies CO above the surface through these correlations. This information is needed to distinguish the immediate vertical distribution of analysis increments from their subsequent redistribution by model transport. The uncertainty assigned to the surface super-observations should also be specified.3. Consistent sampling for experiment comparison — Table 1 and Figures 3–4:
Table 1 evaluates the a priori simulation against MOPITT observations, whereas KF-Surface is evaluated against MOPITT observations sampled near surface stations. Please compare the experiments using the same satellite observations and sampling times before attributing the change in column bias entirely to assimilation. Also, clarify whether the surface correlations of 0.94–0.95 are calculated using assimilated observations. If so, these values describe analysis agreement rather than independent validation.4. Interannual treatment of model inputs — Section 2.3:
Please explain how anthropogenic emissions were prescribed for each year during 2015–2024, particularly the temporal extension or scaling of MIX v2, NEI2016 and CEDS-CMIP6. It is also unclear whether the archived monthly OH fields vary between years or represent a repeated climatology. These details are needed to interpret the simulated trends and the differences between the a priori and assimilated CO fields.5. Trend estimation and sampling uncertainty — Section 3.3 and Table 3:
Please describe the trend-estimation method and specify whether the reported uncertainties represent standard errors or confidence intervals. The treatment of temporal autocorrelation should also be explained. Since the station-selection criterion permits some unqualified years, please clarify how these years enter the regional averages. A common-sampling sensitivity test and a statistical assessment of the proposed plateaus after 2018 and 2016 would strengthen the trend interpretation.Citation: https://doi.org/10.5194/egusphere-2026-4938-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 227 | 130 | 55 | 412 | 78 | 49 | 59 |
- HTML: 227
- PDF: 130
- XML: 55
- Total: 412
- Supplement: 78
- BibTeX: 49
- EndNote: 59
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Tang et al. present results of GEOS-Chem CO emission inversions over China, the USA and Europe with different sources of assimilated observations: in-situ observations from surface CO monitoring networks, covering a broad spectrum of pollution conditions, and column-averaged satellite CO observations from the MOPITT instrument. They find that the model is in good agreement with satellite CO observations but strongly underestimates surface CO compared to in-situ observations, particularly in (sub-)urban regions.
While the manuscript points out an important issue related to representativeness in the CO observing system (most in-situ observations are in urban environments not well represented in models), the results section feels repetitive and does not provide much in-depth analysis of this issue, nor does it attempt to resolve it. The manuscript does not place the results in the wider context or present a direction for research on this topic. The writing style is clear, but there is some repetition in the text, and most figures are missing labels and units. I recommend that this article may be considered for publication after addressing the following major issues:
Minor issues