the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Nov 2023
A global surface CO2 flux dataset (2015–2022) inferred from OCO-2 retrievals using the GONGGA inversion system
Abstract. Accurate assessment of the size and distribution of carbon dioxide (CO2) sources and sinks is important for efforts to understand the carbon cycle and support policy decisions regarding climate mitigation actions. Satellite retrievals of the column-averaged dry-air mole fractions of CO2 (XCO2) have been widely used to infer spatial and temporal variations of carbon fluxes through atmospheric inversion techniques. In this study, we present a global spatially resolved terrestrial and ocean carbon flux dataset for 2015–2022. The dataset was generated by the Global ObservatioN-based system for monitoring Greenhouse GAses (GONGGA) atmospheric inversion system through the assimilation of Orbiting Carbon Observatory 2 (OCO-2) XCO2 retrievals. We describe the carbon budget, interannual variability, and seasonal cycle for the global scale and a set of TransCom regions. The 8-year mean net biosphere exchange and ocean carbon fluxes were −2.22 ± 0.75 PgC yr−1 and –2.32 ± 0.18 PgC yr−1, absorbing approximately 23 % and 24 % of contemporary fossil fuel CO2 emissions, respectively. The annual mean global atmospheric CO2 growth rate was 5.17 ± 0.68 PgC yr−1, which is consistent with the National Oceanic and Atmospheric Administration (NOAA) measurement (5.24 ± 0.59 PgC yr−1). Europe has the largest terrestrial sink among the 11 TransCom land regions, followed by Boreal Asia and Temperate Asia. The dataset was evaluated by comparing posterior CO2 simulations with the observations from Total Carbon Column Observing Network (TCCON) and Observation Package (ObsPack). Compared with CO2 simulations using the unoptimized fluxes, the bias and root mean square error of posterior CO2 simulations were largely reduced across the full range of locations, confirming that the GONGGA system improves the estimates of spatial and temporal variations in carbon fluxes by assimilating OCO-2 XCO2 data. This dataset will improve the broader understanding of global carbon cycle dynamics and their response to climate change. The dataset can be accessed at https://doi.org/10.5281/zenodo.8368846 (Jin et al., 2023a).
- Preprint
(5886 KB) - Metadata XML
-
Supplement
(4660 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on essd-2023-449', Anonymous Referee #1, 24 Jan 2024
This manuscript presents an 8-year dataset of surface-atmosphere CO2 fluxes estimated by the GONGGA inversion system constrained by OCO-2 XCO2 retrievals. This provides a useful dataset to the community and the paper is well written and structured to present the dataset and its evaluation. However, I feel that some important details are missing as described below. I recommend publication after addressing the following minor comments.
Main comments:
- Sec. 2.1: Some details are missing here. What is the spatial and temporal resolution of the optimization. Is it at 2x2.5 and monthly (weekly?)? How is the is the covariance between surface flux and atmospheric CO2 constructed. If the set-up follows the set-up of a previous study then explicitly state this.
- Sec. 2.2: How was prior error covariance matrix created? Is it diagonal? Is it an output of ORCHIDEE-MCT? Same question for ocean flux uncertainties. Based of Fig. 2 is seems that the global land and ocean uncertainties are very different in magnitude, despite the fact that the GCP gives similar order of magnitude uncertainties, why is this?
- Sec 2.3. The OCO-2 XCO2 dataset is not properly cited. There is the v11r standard XCO2 product (no bias correction, JPL DEM, still running routinely), the v11r Lite XCO2 product (bias corrected, JPL NASADEM+, available up to April 2023) and the v11.1r Lite XCO2 product (bias corrected, Copernicus DEM, still running routinely). Please clearly state and cite which dataset was used. An important point is that the DEM used in v11r cause a systematic error over the northern high latitudes that may have impacted the inversion results, if used. The impact of the DEM change is described in Jacobs et al. (2023): https://doi.org/10.5194/amt-2023-151.
Instructions for citing the OCO-2 retrievals are given on the GES DISC website. For example, if this was V11.1r downloaded from GES DISC then citation should be: OCO-2/OCO-3 Science Team, Vivienne Payne, Abhishek Chatterjee (2022), OCO-2 Level 2 bias-corrected XCO2 and other select fields from the full-physics retrieval aggregated as daily files, Retrospective processing V11.1r, Greenbelt, MD, USA, Goddard Earth Sciences Data and Information Services Center (GES DISC), Accessed: [Data Access Date], 10.5067/8E4VLCK16O6Q” - L190-191: I think that the definition “SLAND” is confusing here. In the Global Carbon Budget papers, the term SLAND is the net land sink after accounting for net land-use change emissions. However, in this paper, SLAND is defined as NEE (e.g., SLAND = NBE – Fire). But fire does not equal ELUC, so the definitions are different. I recommend not using SLAND to define this quantity. It may be best to compare the NBE terms between the two studies after accounting for lateral fluxes. I recommend reviewing Sec. 7 of Byrne et al. (2023; https://essd.copernicus.org/articles/15/963/2023/) to see a comparison between the OCO-2 v10 MIP and Global Carbon Budget numbers.
Specific comments
- L25: Specify that these are in situ and flask CO2 ObsPack data.
- L102-103: I think Liu et al. (2021) optimized NBE, so may not be an applicable reference.
- L115: “to December 21, 2022”. Typically, inversions have a spin down period to increase data constraints at the end of the period, why was the inversion not extended into 2023?
- Table 1 should be referenced in Sec. 2.4.1
- L183: would be clearer to say “ocean-atmosphere” than “ocean”
- L193-194: "NEE had substantial interannual variability (-4.08 +/- 0.53 PgC yr-1)". This phrasing makes it seem like -4.08 is the interannual variability. I would suggest re-phasing "NEE had substantial mean sink with considerable interannual variability, estimated as the standard deviation across years (-4.08 _/- 0.53 PgC yr-1) “.
- L232: These are the incorrect citations for the v10 OCO-2 MIP. The documentation of the OCO-2 v10 MIP should be cited as:
Byrne, B., Baker, D. F., Basu, S., Bertolacci, M., Bowman, K. W., Carroll, D., Chatterjee, A., Chevallier, F., Ciais, P., Cressie, N., Crisp, D., Crowell, S., Deng, F., Deng, Z., Deutscher, N. M., Dubey, M. K., Feng, S., García, O. E., Griffith, D. W. T., Herkommer, B., Hu, L., Jacobson, A. R., Janardanan, R., Jeong, S., Johnson, M. S., Jones, D. B. A., Kivi, R., Liu, J., Liu, Z., Maksyutov, S., Miller, J. B., Miller, S. M., Morino, I., Notholt, J., Oda, T., O'Dell, C. W., Oh, Y.-S., Ohyama, H., Patra, P. K., Peiro, H., Petri, C., Philip, S., Pollard, D. F., Poulter, B., Remaud, M., Schuh, A., Sha, M. K., Shiomi, K., Strong, K., Sweeney, C., Té, Y., Tian, H., Velazco, V. A., Vrekoussis, M., Warneke, T., Worden, J. R., Wunch, D., Yao, Y., Yun, J., Zammit-Mangion, A., and Zeng, N.: National CO2 budgets (2015–2020) inferred from atmospheric CO2 observations in support of the global stocktake, Earth Syst. Sci. Data, 15, 963–1004, https://doi.org/10.5194/essd-15-963-2023, 2023.
While the dataset should be cited as:
Baker, D. F., Basu, S., Bertolacci, M., Chevallier, F., Cressie, N., Crowell, S., Deng, F., He, W., Jacobson, A. R., Janardanan, R., Jiang, F., Johnson, M. S., Jones, D. B. A., Liu, J., Liu, Z., Maksyutov, S., Miller, S. M., Philip, S., Schuh, A., Weir, B., Zammit-Mangion, A., and Zeng, N.: v10 Orbiting Carbon Observatory-2 model intercomparison project, NOAA Global Monitoring Laboratory [data set], https://gml.noaa.gov/ccgg/OCO2_v10mip/, last access:. XXX - L245-248: It could be interesting to plot the GONGGA prior and OCO v10 MIP priors as a supplementary figure. Would be interesting if these differences are also present there.
- L258-259: I don’t understand the logic in this sentence: “In the Amazon, the mean gross emissions from forest fires from 2003 to 2015 was 454 ± 496 Tg CO2 yr−1, which may counteract the decline of Amazon deforestation carbon emissions (Aragão et al., 2018).”
- L260: In addition to van der Velde et al. (2021), there were two studies that examined the CO2 emissions from the 2019-20 Australian fires using OCO-2 data:
1. Byrne, B., Liu, J., Lee, M., Yin, Y., Bowman, K. W., Miyazaki, K., et al. (2021). The carbon cycle of southeast Australia during 2019–2020: Drought, fires, and subsequent recovery. AGU Advances, 2, e2021AV000469. https://doi.org/10.1029/2021AV000469
2. Wang, J., Liu, Z., Zeng, N., Jiang, F., Wang, H., & Ju, W. (2020). Spaceborne detection of XCO2 enhancement induced by Australian mega-bush-fires. Environmental Research Letters, 15(12), 124069. https://doi.org/10.1088/1748-9326/abc846 - L272: Please be more specific. I suggest re-writting “the magnitude of global NBE IAV” as “the standard deviation of global NBE IAV”
- L276-277: “Considering the short time series of the carbon cycle, the latitudinal contributions in this study are qualitative, rather than quantitative.” I think this would be better written as “Considering the short time series of the carbon cycle, the latitudinal contributions in this study are suggestive but not statistically robust.”
- L298: “more flatten” should be “smaller amplitude”
- Figure 9-12 captions. Specify “posterior simulations”
- L363: Just for your information, there is a known difference in the mean atmospheric CO2 abundance between TCCON and posterior CO2 fields from in situ inversions, which is not well understood. I'm not sure if this has been documented in a paper, but it is known to some researchers. This could cause the differences seen here
- L366: BIAS shouldn’t be all capitalized.
Citation: https://doi.org/10.5194/essd-2023-449-RC1 -
RC2: 'Comment on essd-2023-449', Anonymous Referee #2, 13 Feb 2024
The authors introduce a new inversion system (GONGGA) that assimilates total column CO2 from NASA’s OCO2 satellite to optimize terrestrial and oceanic carbon fluxes (NEE). The results are compared against a recent model inter-comparison project that assimilated an older version of this dataset. Results are also evaluated against a network of upward looking forward scattering radiometers as well as in-situ surface and aircraft observations.
The manuscript adds a novel inversion system to a growing list of similar simulations (global models that estimate CO2 fluxes by assimilating total column CO2 retrievals). The manuscript is generally well written. Therefore, I think this is suitable for publication in ESSD.
I only have a few concerns at this point:
1. It isn’t clear how XCO2 uncertainties are treated in the inversion system. It is generally assumed that the reported XCO2 uncertainty in the lite files is likely too low. Moreover, unlike in-situ observations, XCO2 data exhibit high correlation (given that individual soundings are only 300 ms apart). Therefore the information content as well as errors are highly correlated for adjacent soundings. Generally, studies have relied on averaging. See Piero et al. 2022, Byrne et al. 2023, and Baker et al., 2022. I would recommend expanding the methods section to describe exactly how retrieval uncertainties are treated (given the context of the afore-mentioned studies) and perform some sensitivity analyses ( e.g., tests where uncertainties are inflated) to estimate the impact of data error on retrieved fluxes.
2. The authors note that the main differences from the OCO v10 MIP arise in the high northern latitudes. At one point this was due an issue with the OCO-2 retrievals in the V11 dataset. I wonder if the retrievals used in the inversion system are impacted by this. I would check with the dataset providers to see if the authors are using a version that is known to have issues. Also see the data quality statement:
https://docserver.gesdisc.eosdis.nasa.gov/public/project/OCO/OCO2_L2_Data_Release_Statement_v11.1_Lite_Files.pdf
Minor comments
Lines 99-100: Biomass burning carbon emissions are also terrestrial ecosystem fluxes, so I would just define NEE instead (i.e., balance of photosynthesis and respiration).
Line 153- Cite ObsPack and also specify which version was used.
Line 176-178: CARVE aircraft observations may not be appropriate for evaluation, given that CARVE flight tracks did not intend to sample regions that were representative of large areas. I would recommend removing CARVE, or discussing this when you discuss results for Fig. 12.
Line 190-92: Earlier it was stated that fossil fuel and biomass burning CO2 emissions were not optimized. So it would be incorrect to say that E_FOS and E_FIRE were quantified. Instead they were specified.
Line 192: S_LAND generally refers to NEE + BMB fluxes. I would suggest changing this to NEE throughout. You could then call the sum of NEE and BMB S_Land or NBE.
Line 231: The OCO-MIP V10 citations should be Byrne et al., (2023). I think here it should also be noted that the v10 MIP assimilated v10 OCO-2 retrievals while in this study OCO-2 v11r retrievals are assimilated.
Lines 258-263: This is citing previous work and should go in a discussion section rather than the results, since it reads like it is a result of this study, which it is not. Finally, given that biomass burning fluxes were not optimized, I think these should be discussed in terms of how their magnitude is relative to the NEE fluxes (that were optimized),
Fig. 6: The caption for this figure should say NBE, not IAV of NBE, since each point on the figure refers to a value not an IAV. Also it seems from this figure that most of the IAV comes from North Extra tropics, but in lines 273-76 the authors say that tropics contribute 100% to global IAV. I think this should be clarified.
In all figures with labels “PgC” should be changed to “Pg C”
Line 317: Observed XCO2 should be changed to retrieved XCO2 given that XCO2 cannot be observed directly.
Figures 11 and 12. RMSE folds in both random and systematic error (bias). See Rastogi et al., (2021) for discussion of bias and random error evaluation of OCO-2 relative to in-situ aircraft observations. For vertical profile data, I think it would be useful to look at errors in the column. For instance, the model may have errors that cancel in the column (e.g., high bias near the surface and a low-bias aloft).
Instead I would recommend the authors to report random error and bias separately. Also, why is bias capitalized?
In this section there are other datasets such as the atmospheric tomography mission (ATom) aircraft campaigns would have been valuable for evaluation. See Gaubert et al., (2023) for details. I would also advise the authors to look at the OCO MIP website for evaluation against observations (surface and partial columns).
References
Baker, D. F., Bell, E., Davis, K. J., Campbell, J. F., Lin, B., and Dobler, J.: A new exponentially decaying error correlation model for assimilating OCO-2 column-average CO2 data using a length scale computed from airborne lidar measurements, Geosci. Model Dev., 15, 649–668, https://doi.org/10.5194/gmd-15-649-2022, 2022.
Byrne, B., Baker, D.F., Basu, S., Bertolacci, M., Bowman, K.W., Carroll, D., Chatterjee, A., Chevallier, F., Ciais, P., Cressie, N. and Crisp, D., 2023. National CO 2 budgets (2015–2020) inferred from atmospheric CO 2 observations in support of the global stocktake. Earth System Science Data, 15(2), pp.963-1004.
Gaubert, B., Stephens, B.B., Baker, D.F., Basu, S., Bertolacci, M., Bowman, K.W., Buchholz, R., Chatterjee, A., Chevallier, F., Commane, R. and Cressie, N., 2023. Neutral tropical African CO2 exchange estimated from aircraft and satellite observations. Global Biogeochemical Cycles, 37(12), p.e2023GB007804.
Peiro, H., Crowell, S., Schuh, A., Baker, D. F., O'Dell, C., Jacobson, A. R., Chevallier, F., Liu, J., Eldering, A., Crisp, D., Deng, F., Weir, B., Basu, S., Johnson, M. S., Philip, S., and Baker, I.: Four years of global carbon cycle observed from the Orbiting Carbon Observatory 2 (OCO-2) version 9 and in situ data and comparison to OCO-2 version 7, Atmos. Chem. Phys., 22, 1097–1130, https://doi.org/10.5194/acp-22-1097-2022, 2022.
Rastogi, B., Miller, J.B., Trudeau, M., Andrews, A.E., Hu, L., Mountain, M., Nehrkorn, T., Baier, B., McKain, K., Mund, J. and Guan, K., 2021. Evaluating consistency between total column CO 2 retrievals from OCO-2 and the in situ network over North America: implications for carbon flux estimation. Atmospheric Chemistry and Physics, 21(18), pp.14385-14401.
Citation: https://doi.org/10.5194/essd-2023-449-RC2
Data sets
A global gridded CO2 flux dataset inferred from OCO-2 retrievals using the GONGGA inversion system (v2023.ori) Zhe Jin, Xiangjun Tian, Yilong Wang, Tao Wang, Shilong Piao https://doi.org/10.5281/zenodo.8368846
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 311 | 74 | 30 | 415 | 34 | 20 | 25 |
- HTML: 311
- PDF: 74
- XML: 30
- Total: 415
- Supplement: 34
- BibTeX: 20
- EndNote: 25
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1