the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Jun 2024
The role of OCO-3 XCO2 retrievals in estimating global terrestrial net ecosystem exchanges
Abstract. Satellite-based column-averaged dry air CO2 mole fraction (XCO2) retrievals are frequently used to improve the estimates of terrestrial net carbon exchanges (NEE). The Orbiting Carbon Observatory 3 (OCO-3) satellite, launched in May 2019, was designed to address important questions about the distribution of carbon fluxes on Earth, but its role in estimating global terrestrial NEE remains unclear. Here, using the Global Carbon Assimilation System, version 2, we investigate the impact of OCO-3 XCO2 on the estimation of global NEE by assimilating the OCO-3 XCO2 retrievals alone and in combination with the OCO-2 XCO2 retrievals. The results show that when only the OCO-3 XCO2 is assimilated (Exp_OCO3), the estimated global land sink is significantly lower than that from the OCO-2 experiment (Exp_OCO2). The estimate from the joint assimilation of OCO-3 and OCO-2 (Exp_OCO3&2) is comparable on a global scale to that of Exp_OCO2. However, there are significant regional differences. Compared to the observed global annual CO2 growth rate, Exp_OCO3 has the largest bias, and Exp_OCO3&2 shows the best performance. Furthermore, validation with independent CO2 observations shows that the biases of the Exp_OCO3 are significantly larger than those of Exp_OCO2 and Exp_OCO3&2 at mid and high latitudes, probably due to the fact that OCO-3 only has observations from 52° S to 52° N. Our study indicates that assimilating OCO-3 XCO2 retrievals alone leads to an underestimation of land sinks at high latitudes, and that a joint assimilation of OCO-2 and OCO-3 XCO2 retrievals is required for a better estimation of global terrestrial NEE.
- Preprint
(1628 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-1568', Anonymous Referee #1, 10 Sep 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1568/egusphere-2024-1568-RC1-supplement.pdf
-
AC1: 'Reply on RC1', Xingyu Wang, 16 Nov 2024
Publisher’s note: this comment was edited on 18 November 2024. The following text is not identical to the original comment, but the adjustments were minor without effect on the scientific meaning.
We thank the referee for taking the time to review our manuscript and provide the supportive comments. Please find our responses in the attached file.
-
AC1: 'Reply on RC1', Xingyu Wang, 16 Nov 2024
-
RC2: 'Comment on egusphere-2024-1568', Anonymous Referee #2, 25 Sep 2024
Summary:
In this work, the authors conduct atmospheric CO_2 inversions to estimate global NEEs using OCO-2 and OCO-3 XCO_2 retrievals and the Global Carbon Assimilation System, version 2 (GCASv2). Three sets of experiments have been designed by the authors to evaluate the impact of using different OCO XCO_2 observations to constrain the posterior carbon fluxes: using OCO-3 XCO_2 only; using OCO-2 XCO_2 only; and using OCO-2 & OCO-3 XCO_2 combined. The overall results suggest using combined OCO-2&OCO-3 XCO_2 retrievals can yield better consistency when compared with in-situ observations while using OCO-3 XCO_2 retrievals alone presents largest biases. The results and discussion reveal some interesting patterns in global and regional NEEs across different experimental setups and provided some insights on the choice of satellite observations to constrain global NEEs, but lack in-depth discussion of the resulted behavior of using OCO-3 XCO_2 only, OCO-2 XCO_2 only, and OCO-2&OCO-3 XCO_2 combined. Please see below sections for detailed comments and I would expect the manuscript to be published once comments and questions have been resolved.
Main comments/questions:
- More information needed for the GCASv2. I understand the GCASv2 is an established model and described in at least two other published journal articles, but detailed information on the model setup, inversion methods, and error covariance metrics can be very helpful for readers of this manuscript to better understand the inversion system and results. Also see related comments in the Technical notes section.
- How are the posterior fluxes constrained when there’s no observation data in GCASv2? For example, in the Exp_OCO3 at high latitudes, I would assume the posterior fluxes are less updated and would be similar to prior fluxes since no new information has been presented to the inversion system, but Figure 3 and Figure 6 seem to suggest the posterior fluxes changed substantially when compared to prior. More information on the EnSRF would be helpful for the readers to understand the inversion process.
- I’m curious about the authors' insights on why in general using OCO-2 XCO_2 alone and OCO-2&OCO-3 XCO_2 combined outperforms the experiment using OCO-3 only? Would there be any other reason except the spatial coverage and potential bias in OCO-3 XCO_2 (line 263)?
General comments:
Line 113: How does GCASv2 handle parameters of the aggregated ‘super-observation’? For example, if multiple OCO soundings has been aggregated into one ‘super-observation’, how does GCASv2 incorporate information such as pressure weighting function and averaging kernels from each individual soundings?
Line 131: Can you justify the use of ocean glint? Ocean glint data is in general avoided in inversions due to potential high bias.
Line 134: Please explain the regridding process. Does the regridding process refer to the ‘super-observation’ described in section 2.1? How did the XCO_2 values and parameters for each sounding been processed? Did you take the mean, or median, or other methods? And can you justify the method you used? How did you handle the outliers in the observations with one grid box? Also, for the ‘super-observation’, does it mean that for each model grid box, there’s essentially only one observation being used by the model to constrain the posterior fluxes? If that’s the case, why does the data amount (Figure 1) matter (except for the grids containing 0 OCO soundsing)?
Line 209: Could you list out the annual CO_2 growth rates for 2020-2022 that you used to calculate the average growth rates?
Line 214: Why does the joint assimilation of OCO-2 and OCO-3 XCO_2 give the best performance on a global scale? One potential reason - spatial coverage of OCO-3 XCO_2 - has been mentioned briefly in several places in the manuscript, but an in-depth discussion would be expected.
Line 221 - 224: Is the word ‘sinks’ in line 22 a typo? Otherwise the sentence does not make sense – the listed locations seem to have positive NEE values suggesting being CO_2 sources.
Line 236: I would suggest the authors avoid using ‘peaks’ when describing the negative values to clear confusion, or maybe specify the values when doing comparison. For example, the ‘peaks’ for ExpOCO2 and Exp_OCO3&2 are higher than the prior when rotation 90 degrees for Figure 3 (f) and (i), but the actually corresponding values at the ‘peaks’ are lower because they are CO_2 sinks and the NEE values are negative. Same for ‘the lowest peak’ in line 237.
Line 251: Potential confusion – by the word ‘lower’ do you mean the NEE value is lower (strong sinks) or the NEE value is higher (weaker sinks)?
Line 301: Which experiment are those numbers from?
Table 2 and Figure 4: Is the information presented in Table 2 and Figure 4 largely duplicated? If so, authors may consider removing Figure 4 if additional paragraphs are needed.
Figure 6, Figure 3 and Line 358: For high latitude areas (> 60 degree N), why is the BIAS from Exp_OCO3 not consistent with prior fluxes? Given the fact that no OCO-3 observations available beyond 52 degree north, I would expect the posterior fluxes are very similar to prior fluxes in high latitude areas since no observation can be used to constrain and optimize prior emissions, yet both Figure 3 and Figure 6 showed substantial changes when comparing posterior to prior from Exp_OCO3. It’s possible that fluxes in high latitude can be updated due to spatial covariance assumed in the inversion system, therefore more details on the GCASv2 is needed in Section 2.1.
Line 367: Could the bias exist prior? If there’s no OCO-3 observation available in high latitudes, how can the OCO-3 observation introduce additional bias?
Line 376: period ‘1’?
Citation: https://doi.org/10.5194/egusphere-2024-1568-RC2 - AC2: 'Reply on RC2', Xingyu Wang, 16 Nov 2024
Status: closed
-
RC1: 'Comment on egusphere-2024-1568', Anonymous Referee #1, 10 Sep 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-1568/egusphere-2024-1568-RC1-supplement.pdf
-
AC1: 'Reply on RC1', Xingyu Wang, 16 Nov 2024
Publisher’s note: this comment was edited on 18 November 2024. The following text is not identical to the original comment, but the adjustments were minor without effect on the scientific meaning.
We thank the referee for taking the time to review our manuscript and provide the supportive comments. Please find our responses in the attached file.
-
AC1: 'Reply on RC1', Xingyu Wang, 16 Nov 2024
-
RC2: 'Comment on egusphere-2024-1568', Anonymous Referee #2, 25 Sep 2024
Summary:
In this work, the authors conduct atmospheric CO_2 inversions to estimate global NEEs using OCO-2 and OCO-3 XCO_2 retrievals and the Global Carbon Assimilation System, version 2 (GCASv2). Three sets of experiments have been designed by the authors to evaluate the impact of using different OCO XCO_2 observations to constrain the posterior carbon fluxes: using OCO-3 XCO_2 only; using OCO-2 XCO_2 only; and using OCO-2 & OCO-3 XCO_2 combined. The overall results suggest using combined OCO-2&OCO-3 XCO_2 retrievals can yield better consistency when compared with in-situ observations while using OCO-3 XCO_2 retrievals alone presents largest biases. The results and discussion reveal some interesting patterns in global and regional NEEs across different experimental setups and provided some insights on the choice of satellite observations to constrain global NEEs, but lack in-depth discussion of the resulted behavior of using OCO-3 XCO_2 only, OCO-2 XCO_2 only, and OCO-2&OCO-3 XCO_2 combined. Please see below sections for detailed comments and I would expect the manuscript to be published once comments and questions have been resolved.
Main comments/questions:
- More information needed for the GCASv2. I understand the GCASv2 is an established model and described in at least two other published journal articles, but detailed information on the model setup, inversion methods, and error covariance metrics can be very helpful for readers of this manuscript to better understand the inversion system and results. Also see related comments in the Technical notes section.
- How are the posterior fluxes constrained when there’s no observation data in GCASv2? For example, in the Exp_OCO3 at high latitudes, I would assume the posterior fluxes are less updated and would be similar to prior fluxes since no new information has been presented to the inversion system, but Figure 3 and Figure 6 seem to suggest the posterior fluxes changed substantially when compared to prior. More information on the EnSRF would be helpful for the readers to understand the inversion process.
- I’m curious about the authors' insights on why in general using OCO-2 XCO_2 alone and OCO-2&OCO-3 XCO_2 combined outperforms the experiment using OCO-3 only? Would there be any other reason except the spatial coverage and potential bias in OCO-3 XCO_2 (line 263)?
General comments:
Line 113: How does GCASv2 handle parameters of the aggregated ‘super-observation’? For example, if multiple OCO soundings has been aggregated into one ‘super-observation’, how does GCASv2 incorporate information such as pressure weighting function and averaging kernels from each individual soundings?
Line 131: Can you justify the use of ocean glint? Ocean glint data is in general avoided in inversions due to potential high bias.
Line 134: Please explain the regridding process. Does the regridding process refer to the ‘super-observation’ described in section 2.1? How did the XCO_2 values and parameters for each sounding been processed? Did you take the mean, or median, or other methods? And can you justify the method you used? How did you handle the outliers in the observations with one grid box? Also, for the ‘super-observation’, does it mean that for each model grid box, there’s essentially only one observation being used by the model to constrain the posterior fluxes? If that’s the case, why does the data amount (Figure 1) matter (except for the grids containing 0 OCO soundsing)?
Line 209: Could you list out the annual CO_2 growth rates for 2020-2022 that you used to calculate the average growth rates?
Line 214: Why does the joint assimilation of OCO-2 and OCO-3 XCO_2 give the best performance on a global scale? One potential reason - spatial coverage of OCO-3 XCO_2 - has been mentioned briefly in several places in the manuscript, but an in-depth discussion would be expected.
Line 221 - 224: Is the word ‘sinks’ in line 22 a typo? Otherwise the sentence does not make sense – the listed locations seem to have positive NEE values suggesting being CO_2 sources.
Line 236: I would suggest the authors avoid using ‘peaks’ when describing the negative values to clear confusion, or maybe specify the values when doing comparison. For example, the ‘peaks’ for ExpOCO2 and Exp_OCO3&2 are higher than the prior when rotation 90 degrees for Figure 3 (f) and (i), but the actually corresponding values at the ‘peaks’ are lower because they are CO_2 sinks and the NEE values are negative. Same for ‘the lowest peak’ in line 237.
Line 251: Potential confusion – by the word ‘lower’ do you mean the NEE value is lower (strong sinks) or the NEE value is higher (weaker sinks)?
Line 301: Which experiment are those numbers from?
Table 2 and Figure 4: Is the information presented in Table 2 and Figure 4 largely duplicated? If so, authors may consider removing Figure 4 if additional paragraphs are needed.
Figure 6, Figure 3 and Line 358: For high latitude areas (> 60 degree N), why is the BIAS from Exp_OCO3 not consistent with prior fluxes? Given the fact that no OCO-3 observations available beyond 52 degree north, I would expect the posterior fluxes are very similar to prior fluxes in high latitude areas since no observation can be used to constrain and optimize prior emissions, yet both Figure 3 and Figure 6 showed substantial changes when comparing posterior to prior from Exp_OCO3. It’s possible that fluxes in high latitude can be updated due to spatial covariance assumed in the inversion system, therefore more details on the GCASv2 is needed in Section 2.1.
Line 367: Could the bias exist prior? If there’s no OCO-3 observation available in high latitudes, how can the OCO-3 observation introduce additional bias?
Line 376: period ‘1’?
Citation: https://doi.org/10.5194/egusphere-2024-1568-RC2 - AC2: 'Reply on RC2', Xingyu Wang, 16 Nov 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 432 | 95 | 137 | 664 | 20 | 25 |
- HTML: 432
- PDF: 95
- XML: 137
- Total: 664
- BibTeX: 20
- EndNote: 25
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1