the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 15 Aug 2023
Long-term reduction in CO2 emissions from the Elbe River due to water quality improvement
Abstract. Polluted rivers transport and transform large quantities of anthropogenically-derived organic carbon to coastal regions, and account for an unneglectable share of global CO2 emissions. Effective river water management can enhance water quality and reduce CO2 emissions from the surface water to the atmosphere. However, the effect of water management on future riverine CO2 emissions remains a topic yet to be explored. In this context, the effects of water quality on riverine carbon dynamics were evaluated by examining the temporal variations in carbon fluxes from the Elbe River during a climate base period (around 30 years) from 1984 to 2018. The analysis of long-term data reveals that annual CO2 emissions from the Elbe River have decreased from 3.8±1.7 Tg C yr–1 in 1984 to 1.3±0.6 Tg C yr–1 in 2018 (1 T = 1012), and the largest reduction occurred after the initial decade of reunification of Germany. The changes in wastewater treatment have largely reduced nutrient loads, mitigated eutrophication, impacted the quality of the transported carbon to the ocean, resulting in concurrent decreases in CO2 emissions. The long-term trends in the Elbe River underscore the importance of water quality management for mitigating CO2 emissions from polluted rivers around the globe.
- Preprint
(1751 KB) - Metadata XML
-
Supplement
(1674 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on bg-2023-131', Anonymous Referee #1, 12 Sep 2023
MAJOR COMMENTS
Long-term patterns of CO2 levels and emissions in rivers have been reported by several studies (Jones et al. 2003; Ran et al. 2015; 2021 ; Nydahl et al. 2017 ; Marescaux et al. 2018) (non-exhaustive list). Findings from these studies could be used to contextualize the present study (Introduction) and to discuss differences or convergences by comparison (Discussion).
L 30 : « water pollution » is extremely vague. This should be broken down into several human impacts on riverine systems that do not necessarily lead to the same change in CO2 emissions. Eutrophication (increase of nutrient inputs) can potentially lead to enhanced primary production and a CO2 sink in impounded large rivers such as the Mississippi (Crawford et al. 2016). Conversely, croplands seem to also lead to enhanced organic carbon inputs from soils enhancing CO2 emissions compared to more natural land cover such as forests (Borges et al. 2018; Mwanake et al. 2023) Waste water inputs lead to CO2 production in the river, although this impact seems very local, in the near vicinity of the emissary (Marescaux et al. 2018).
L 30 “this percentage continues to increase because the unprecedented anthropogenic stresses on riverine systems have led to many negative issues such as water pollution”. I’m not sure this statement applies assertively to all climate zones. According to Liu et al. (2022), tropical rivers are responsible for 57% of the riverine CO2 global emission, followed by temperate (30%) and Arctic regions (13%). The most direct anthropogenic impacts expected to affect riverine CO2 emissions should occur at temperate latitudes (North America, Europe and parts of Asia) that account for less than a third of total emissions. Note that this percentage was lower in earlier estimates for which tropical rivers accounted for 80% of riverine CO2 emissions (Raymond et al. 2013; Lauewarld et al. 2015).
L 34: Rivers do not have “ecosystem's natural carbon absorption and storage capabilities”. Rivers do not store carbon in sediments and do not “absorb” carbon on contrary tend to emit CO2 to the atmosphere. High CO2 over-saturation in rivers occurs ubiquitously even in pristine (or near pristine) river basins such as the Amazon and Congo.
L 37 : It has been argued that CO2 emissions from lowland rivers in particular in the tropics are related to inputs from wetlands (Abril et al. 2014; Borges et al. 2015) that are conceptually different (Abril and Borges 2019) from “terrestrial organic carbon (OC)» (as stated).
L38 : Can you please clarify the role of «nutrient availability” in this context ?
L44-46: This argument is awkward. DOM produced by phytoplankton should indeed sustain microbial respiration but phytoplankton also photosynthesized prior to DOM release, so both effects should cancel each other in terms of net carbon fluxes.
L 44: reference to “lakes and reservoirs » seems to be out of context here.
L49-50: statement “trophic status related to nutrient availability significantly impacts the levels of CO2 in rivers” is contradicted by the fact that CO2 emissions in rivers are in majority related to lateral inputs of carbon from soils and ground-waters (Hotchkiss et al. 2015) or from wetlands (Abril and Borges 2019), and are not related to in-stream CO2 production from metabolism (Hotchkiss et al. 2015; Abril et al. 2014; Borges et al. 2019).
L 51: reference to “biodiversity” seems out of context here.
L 55 : The authors should cite the “existing studies” they critique rather than stating this in a vague way.
L 55 : Please clarify what is meant by “short term effects » ? “effects” of what on what ? Do you mean short-term time-series ? Some studies have reported relatively long time series (Jones et al. 2003; Ran et al. 2015; 2021 ; Nydahl et al. 2017 ; Marescaux et al. 2018). It is not necessary to downplay existing literature to put forward your own study.
L 55 : What do you mean by «hydrological conditions» ? CO2 emissions from rivers depend on CO2 concentration between water and air, and on the gas transfer velocity. Both are more or less indirectly linked to “hydrological conditions” but this should be clarified, especially when criticizing “existing studies”.
L61 : Please provide a reference to back this statement, and clarify compared to which other rivers was it the most polluted? At European level ? Globally ? It could be also useful to take into account size effects. A very small stream can be extremely impacted by wastewater from a small village, while very large rivers are unaffected by large cities because all inputs are diluted by high discharge.
L 163 : the equation relating river width and Q given by Raymond et al. (2012) was derived for small streams. Can you comment on its applicability to large rivers ? Also this relation is probably affected by channelization and probably does not apply to highly engineered rivers such as the Elbe.
L 300 : can you please provide a numerical comparison and a reference for the data for the 1954–1977 period ?
Can you please explain somewhere in text why the analysis was not extended back to 1954 and only started in 1984 ?
L341-344: This statement does not seem relevant. Indeed, it is conceivable that light absorption by CDOM limits photosynthesis from aquatic primary producers, but in rivers CDOM mostly originates from soils. Also, DOM from phytoplankton is usually very labile and is quickly consumed by micro-organisms. CDOM is usually related to highly refractory substances, typically from soils.
L 370-373: Please clarify the text of the two hypothesis and also provide extra arguments and references to back them.
What do you mean by “biomass amount » and why should it not increase in « restored aquatic system” ?
What do you mean by “challenging through water quality treatments.”
MINOR COMMENTS
Text contains numerous awkward phrasing or typos or redundancies. The senior co-authors should spend some time looking through the text and make the necessary improvements; this is not the reviewer’s job. Nevertheless, some are listed hereafter (not an exhaustive list):
L40 + L 337: Labile instead of “liable” ?
L 55 : context instead of « contest” ?
L42: “phytoplankton behaviors » is awkward, please rephrase.
L61: most instead of “highest”
L66: “FCO2 efflux » is redundant sinc "F" of "FCO2" abbreviates the word flux.
L68: "high-resolution" is self-evaluation, please simply state instead the actual time step of the data.
L 368: “CO2 drawdown ratio by water quality management” is awkward, please rephrase.
REFERENCES
Abril G & AV Borges (2019) Carbon leaks from flooded land: do we need to re-plumb the inland water active pipe? Biogeosciences, 16, 769-784, https://doi.org/10.5194/bg-16-769-2019
Abril, G. et al. Amazon River carbon dioxide outgassing fuelled by wetlands. Nature 505, 395-398 (2014). https://doi.org/10.1038/nature12797
Borges A.V. et al. (2018) Effects of agricultural land use on fluvial carbon dioxide, methane and nitrous oxide concentrations in a large European river, the Meuse (Belgium), Science of the Total Environment, 610-611, 342-355, https://doi.org/10.1016/j.scitotenv.2017.08.047
Borges AV et al. (2015) Globally significant greenhouse gas emissions from African inland waters, Nature Geoscience, 8, 637-642, https://doi.org/10.1038/NGEO2486
Borges AV et al. (2019) Variations in dissolved greenhouse gases (CO2, CH4, N2O) in the Congo River network overwhelmingly driven by fluvial-wetland connectivity, Biogeosciences, 16, 3801-3834, https://doi.org/10.5194/bg-16-3801-2019
Crawford J.T. et al. (2016) Basin scale controls on CO2 and CH4 emissions from the Upper Mississippi River, Geophysical Research Letters, 43, 1973-1979, 10.1002/2015gl067599
Hotchkiss E et al. (2015) Sources of and processes controlling CO2 emissions change with the size of streams and rivers. Nature Geosci 8, 696–699. https://doi.org/10.1038/ngeo2507
Jones J.B. et al. (2003), Long-term decline in carbon dioxide supersaturation in rivers across the contiguous United States, Geophys. Res. Lett., 30, 1495, doi:10.1029/2003GL017056, 10.
Lauerwald, R. et al. (2015) Spatial patterns in CO2 evasion from the global river network. Global Biogeochem. Cycles 29, 534-554. doi:10.1002/2014GB004941.
Liu, S. et al. (2022) The importance of hydrology in routing terrestrial carbon to the atmosphere via global streams and rivers. Proc. Natl. Acad. Sci. USA 119, e2106322119.
Marescaux A et al. (2018) Seasonal and spatial variability of the partial pressure of carbon dioxide in the human-impacted Seine River in France, Scientific Reports, Scientific Reports, 8(13961), https://doi.org/10.1038/s41598-018-32332-2
Mwanake RM et al. (2023) Anthropogenic activities significantly increase annual greenhouse gas (GHG) fluxes from temperate headwater streams in Germany, Biogeosciences, 20, 3395–3422, https://doi.org/10.5194/bg-20-3395-2023
Nydahl A. C. et al. (2017), No long-term trends in pCO2 despite increasing organic carbon concentrations in boreal lakes, streams, and rivers, Global Biogeochem. Cycles, 31, 985–995, doi:10.1002/2016GB005539.
Prasad M.B.K. et al. (2013) Long-term pCO2 dynamics in rivers in the Chesapeake Bay watershed, Applied Geochemistry, 31, 209-215, https://doi.org/10.1016/j.apgeochem.2013.01.006.
Ran L et al. (2015) Long-term spatial and temporal variation of CO2 partial pressure in the Yellow River, China, Biogeosciences, 12, 921–932, https://doi.org/10.5194/bg-12-921-2015
Ran L. et al. (2021) Substantial decrease in CO2 emissions from Chinese inland waters due to global change. Nat Commun 12, 1730. https://doi.org/10.1038/s41467-021-21926-6
Raymond, P.A. et al. (2013) Global carbon dioxide emissions from inland waters. Nature 503, 355-359. https://doi.org/10.1038/nature12760
Citation: https://doi.org/10.5194/bg-2023-131-RC1 -
AC1: 'Reply on RC1', Mingyang Tian, 02 Jan 2024
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2023-131/bg-2023-131-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Mingyang Tian, 02 Jan 2024
-
RC2: 'Comment on bg-2023-131', Anonymous Referee #2, 19 Sep 2023
This manuscript presents long-term water quality time series data from the Elbe River in Europe. The authors use alkalinity and pH measurements to estimate dissolved CO2 concentrations, which they use to estimate CO2 emissions from the river and tributaries from 1984 to 2018. They then compare the temporal changes in CO2 emissions with the temporal changes in DIC, DOC and POC loads at the watershed’s outlet, along with other water quality parameters. The authors show a decrease in CO2 emissions, which they relate to an improvement in water quality, particularly a decrease in DOC.
The paper suffers from several shortcomings in methodology, a poor presentation of results, and considerable issues with the English language. I must admit this comes as a surprise considering the list of authors, some of whom are widely recognized and respected in the scientific community. I think there is potential to improve this paper substantially, because the dataset holds significant value—but much more guidance will need to be provided by the co-authors. In the following I will elaborate on the three main concerns I have.
Methodological limitations
One limitation is that the entire paper is based on the use of two indirect methods to estimate CO2 emissions. First, pCO2 estimates are indirectly calculated from pH and alkalinity measurements. While this is a common undertaking, the authors must at least provide a quantification of uncertainties. Their plot comparing pCO2 estimates based on two different packages (PHREEQC and CO2SYS) raises concerns as it shows large differences between the two sets of estimates. Second, the CO2 emission estimates lack actual measurements. The authors use an empirical model which was primarily developed for smaller streams and might not be suitable to large rivers. The model results are not evaluated against actual measurements. Again, this needs to be justified (i.e. why was this particular model chosen and not another one?), and an assessment of uncertainties should be presented.
Another critical issue is with the use of discharge values for k600 estimates. From what I gather, the authors have used only one discharge value for each river location. This approach is problematic because k600 is highly influenced by discharge fluctuations, and failing to account for discharge fluctuations will result in erroneous CO2 emission flux estimates. This issue becomes evident in Figure 3f, where the relationship between FCO2 and pCO2 is almost perfectly linear—either suggesting that k600 has no influence on FCO2, or that k600 remains constant across space and time, both of which are improbable.
A third issue is with the DOC data. It appears that two methods are used for the DOC flux estimation, yet only one is presented in the Results section. Furthermore, the first method does not present a way to calculate loads, but simply provides a framework for classifying C-Q patterns, which is rather confusing.
Presentation of results
The results of statistical tests are not consistently reported throughout the paper. For example, Mann-Kendall trend test results are not presented for pCO2 and FCO2 (L231-261) as well as for DIC, DOC and POC (L276-291), making it challenging to assess the significance of the observed trends. Furthermore, there are no reported step change test results, despite the mention of these tests in the Methods section.
I also noted some inconsistent statements between the results and discussion: while on L281 the authors state that “POC, DOC and DIC loads did not show significant trends”, this contradicts the following statement that the DOC load “showed relatively robust decreasing trend” (L310-311).
Lastly, several figures are missing. For instance, the pCO2 time-series data are not shown despite these data being arguably one of the most critical data of the paper.
English language
The paper is very challenging to understand, and clearly the more senior authors (some of whom are well-published) have not provided the necessary feedback. It seems like only the abstract and the first few paragraphs of the introduction have been edited. The language used is awkward at best, and completely incoherent at worst. As a reviewer, I am not willing to invest one or two days correcting grammar and editing the entire paper. I strongly recommend that the senior authors fulfil their responsibilities of reviewing and editing this paper.
Citation: https://doi.org/10.5194/bg-2023-131-RC2 -
AC2: 'Reply on RC2', Mingyang Tian, 02 Jan 2024
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2023-131/bg-2023-131-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Mingyang Tian, 02 Jan 2024
-
CC1: 'Comment on bg-2023-131', Matthias Koschorreck, 20 Sep 2023
Comment on Tian et al. „Long-term reduction in CO2 emissions from the Elbe River due to water quality improvement
Matthias Koschorreck
It is a very good idea to use long term monitoring data to investigate the effect of the socioeconomic changes in Germany after re-unification on GHG emissions from a large river. The paper contains a very nice dataset including both main river and tributary data which allows the investigation of both spatial and inter-annual pattern. However, in my eyes the manuscript does not fully exploit the potential of the dataset and has some serious issues which I would like to address in the following:
- I cannot follow the argumentation that nutrient driven eutrophication should increase CO2 Any CO2 produced from decaying algae was fixed by those algae before. Thus, the cycle of primary production and algae mineralization cannot increase CO2 emissions. In contrast it has the potential to reduce CO2 emissions if algae are buried in the sediments – a scenario relevant for lakes but probably not for rivers.
I would hypothesize that correlation between N or P with CO2 might be a pseudo correlation and not a direct mechanistic link. As written in the manuscript, wastewater contains both DOC and inorganic nutrients. - In the manuscript a rather crude method is used to estimate river surface area. The resulting surface area of 735 km2 (supplement) looks rather high. Divided by river length this means a river width of about 1 km – an unrealistic high value. In Mallast et al. (2020) we determined a surface area of 106 km2 from satellite images.
- The gas transfer velocity was estimated from slope and flow velocity. However, there are also k600 data from River Elbe published (Matoušů et al., 2019). It should at least be checked how estimated k600 data compare to measured ones.
- In Kamjunke et al. (2022) and Kamjunke et al. (2023) it was shown that there is a longitudinal gradient with plankton concentrations increasing downstream the river. It would be interesting to analyze the dataset in this paper with respect to this gradient. Was the transition zone between plankton poor and plankton rich water moving downstream after 1990?
- The dataset also should allow the comparison of different tributaries. Statistical relations between CO2 and other parameters could be checked for each tributary separately. This can be used to investigate the drivers of CO2 in the different sub-catchments. The effect of the tributaries on the main stream, however, is probably difficult to detect. In Bussmann et al. (2022) for example we showed that the high dilution effect at the confluence did not allow the detection of CH4 import from the tributaries into the main river.
- Recent literature shows that CO2 concentrations in rivers fluctuate diurnally (Gómez-Gener et al., 2021). Thus, scaling up CO2 emissions from single datapoints means accepting a systematic uncertainty. Our own measurements show that diurnal fluctuation of CO2 is an issue in River Elbe (manuscript in preparation). This could be relevant in long term time series, if the time of day when samples were taken changed during the time series.
- If CO2 emissions are primarily driven by DOC mineralization the dataset should allow a quantitative comparison between the two. Was DOC decreasing downstream and how does that downstream decrease of DOC compare quantitatively to CO2 emissions? Such a question could be investigated by looking at monitoring data from longer reaches without major tributaries.
- An analysis of long term changes of water quality in river Elbe was recently published by (Wachholz et al., 2022)
references
Bussmann, I., Koedel, U., Schütze, C., Kamjunke, N., and Koschorreck, M.: Spatial Variability and Hotspots of Methane Concentrations in a Large Temperate River, Frontiers in Environmental Science, 10, 10.3389/fenvs.2022.833936, 2022.
Gómez-Gener, L., Rocher-Ros, G., Battin, T., Cohen, M. J., Dalmagro, H. J., Dinsmore, K. J., Drake, T. W., Duvert, C., Enrich-Prast, A., Horgby, Å., Johnson, M. S., Kirk, L., Machado-Silva, F., Marzolf, N. S., McDowell, M. J., McDowell, W. H., Miettinen, H., Ojala, A. K., Peter, H., Pumpanen, J., Ran, L., Riveros-Iregui, D. A., Santos, I. R., Six, J., Stanley, E. H., Wallin, M. B., White, S. A., and Sponseller, R. A.: Global carbon dioxide efflux from rivers enhanced by high nocturnal emissions, Nature Geoscience, 14, 289-294, 10.1038/s41561-021-00722-3, 2021.
Kamjunke, N., Beckers, L. M., Herzsprung, P., von Tumpling, W., Lechtenfeld, O., Tittel, J., Risse-Buhl, U., Rode, M., Wachholz, A., Kallies, R., Schulze, T., Krauss, M., Brack, W., Comero, S., Gawlik, B. M., Skejo, H., Tavazzi, S., Mariani, G., Borchardt, D., and Weitere, M.: Lagrangian profiles of riverine autotrophy, organic matter transformation, and micropollutants at extreme drought, Science of the Total Environment, 828, ARTN 154243
10.1016/j.scitotenv.2022.154243, 2022.
Kamjunke, N., Brix, H., Floeser, G., Bussmann, I., Schuetze, C., Achterberg, E. P., Koedel, U., Fischer, P., Rewrie, L., Sanders, T., Borchardt, D., and Weitere, M.: Large-scale nutrient and carbon dynamics along the river-estuary-ocean continuum, Science of the Total Environment, 890, ARTN 164421
10.1016/j.scitotenv.2023.164421, 2023.
Mallast, U., Staniek, M., and Koschorreck, M.: Spatial upscaling of CO2 emissions from exposed river sediments of the Elbe River during an extreme drought, Ecohydrology, 13, ARTN e221610.1002/eco.2216, 2020.
Matoušů, A., Rulík, M., Tušer, M., Bednařík, A., Šimek, K., and Bussmann, I.: Methane dynamics in a large river: a case study of the Elbe River, Aquat Sci, 81, 12, 2019.
Wachholz, A., Jawitz, J. W., Büttner, O., Jomaa, S., Merz, R., Yang, S., and Borchardt, D.: Drivers of multi-decadal nitrate regime shifts in a large European catchment, Environmental Research Letters, 17, 064039, 10.1088/1748-9326/ac6f6a, 2022.
Citation: https://doi.org/10.5194/bg-2023-131-CC1 -
AC3: 'Reply on CC1', Mingyang Tian, 02 Jan 2024
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2023-131/bg-2023-131-AC3-supplement.pdf
- I cannot follow the argumentation that nutrient driven eutrophication should increase CO2 Any CO2 produced from decaying algae was fixed by those algae before. Thus, the cycle of primary production and algae mineralization cannot increase CO2 emissions. In contrast it has the potential to reduce CO2 emissions if algae are buried in the sediments – a scenario relevant for lakes but probably not for rivers.
Status: closed
-
RC1: 'Comment on bg-2023-131', Anonymous Referee #1, 12 Sep 2023
MAJOR COMMENTS
Long-term patterns of CO2 levels and emissions in rivers have been reported by several studies (Jones et al. 2003; Ran et al. 2015; 2021 ; Nydahl et al. 2017 ; Marescaux et al. 2018) (non-exhaustive list). Findings from these studies could be used to contextualize the present study (Introduction) and to discuss differences or convergences by comparison (Discussion).
L 30 : « water pollution » is extremely vague. This should be broken down into several human impacts on riverine systems that do not necessarily lead to the same change in CO2 emissions. Eutrophication (increase of nutrient inputs) can potentially lead to enhanced primary production and a CO2 sink in impounded large rivers such as the Mississippi (Crawford et al. 2016). Conversely, croplands seem to also lead to enhanced organic carbon inputs from soils enhancing CO2 emissions compared to more natural land cover such as forests (Borges et al. 2018; Mwanake et al. 2023) Waste water inputs lead to CO2 production in the river, although this impact seems very local, in the near vicinity of the emissary (Marescaux et al. 2018).
L 30 “this percentage continues to increase because the unprecedented anthropogenic stresses on riverine systems have led to many negative issues such as water pollution”. I’m not sure this statement applies assertively to all climate zones. According to Liu et al. (2022), tropical rivers are responsible for 57% of the riverine CO2 global emission, followed by temperate (30%) and Arctic regions (13%). The most direct anthropogenic impacts expected to affect riverine CO2 emissions should occur at temperate latitudes (North America, Europe and parts of Asia) that account for less than a third of total emissions. Note that this percentage was lower in earlier estimates for which tropical rivers accounted for 80% of riverine CO2 emissions (Raymond et al. 2013; Lauewarld et al. 2015).
L 34: Rivers do not have “ecosystem's natural carbon absorption and storage capabilities”. Rivers do not store carbon in sediments and do not “absorb” carbon on contrary tend to emit CO2 to the atmosphere. High CO2 over-saturation in rivers occurs ubiquitously even in pristine (or near pristine) river basins such as the Amazon and Congo.
L 37 : It has been argued that CO2 emissions from lowland rivers in particular in the tropics are related to inputs from wetlands (Abril et al. 2014; Borges et al. 2015) that are conceptually different (Abril and Borges 2019) from “terrestrial organic carbon (OC)» (as stated).
L38 : Can you please clarify the role of «nutrient availability” in this context ?
L44-46: This argument is awkward. DOM produced by phytoplankton should indeed sustain microbial respiration but phytoplankton also photosynthesized prior to DOM release, so both effects should cancel each other in terms of net carbon fluxes.
L 44: reference to “lakes and reservoirs » seems to be out of context here.
L49-50: statement “trophic status related to nutrient availability significantly impacts the levels of CO2 in rivers” is contradicted by the fact that CO2 emissions in rivers are in majority related to lateral inputs of carbon from soils and ground-waters (Hotchkiss et al. 2015) or from wetlands (Abril and Borges 2019), and are not related to in-stream CO2 production from metabolism (Hotchkiss et al. 2015; Abril et al. 2014; Borges et al. 2019).
L 51: reference to “biodiversity” seems out of context here.
L 55 : The authors should cite the “existing studies” they critique rather than stating this in a vague way.
L 55 : Please clarify what is meant by “short term effects » ? “effects” of what on what ? Do you mean short-term time-series ? Some studies have reported relatively long time series (Jones et al. 2003; Ran et al. 2015; 2021 ; Nydahl et al. 2017 ; Marescaux et al. 2018). It is not necessary to downplay existing literature to put forward your own study.
L 55 : What do you mean by «hydrological conditions» ? CO2 emissions from rivers depend on CO2 concentration between water and air, and on the gas transfer velocity. Both are more or less indirectly linked to “hydrological conditions” but this should be clarified, especially when criticizing “existing studies”.
L61 : Please provide a reference to back this statement, and clarify compared to which other rivers was it the most polluted? At European level ? Globally ? It could be also useful to take into account size effects. A very small stream can be extremely impacted by wastewater from a small village, while very large rivers are unaffected by large cities because all inputs are diluted by high discharge.
L 163 : the equation relating river width and Q given by Raymond et al. (2012) was derived for small streams. Can you comment on its applicability to large rivers ? Also this relation is probably affected by channelization and probably does not apply to highly engineered rivers such as the Elbe.
L 300 : can you please provide a numerical comparison and a reference for the data for the 1954–1977 period ?
Can you please explain somewhere in text why the analysis was not extended back to 1954 and only started in 1984 ?
L341-344: This statement does not seem relevant. Indeed, it is conceivable that light absorption by CDOM limits photosynthesis from aquatic primary producers, but in rivers CDOM mostly originates from soils. Also, DOM from phytoplankton is usually very labile and is quickly consumed by micro-organisms. CDOM is usually related to highly refractory substances, typically from soils.
L 370-373: Please clarify the text of the two hypothesis and also provide extra arguments and references to back them.
What do you mean by “biomass amount » and why should it not increase in « restored aquatic system” ?
What do you mean by “challenging through water quality treatments.”
MINOR COMMENTS
Text contains numerous awkward phrasing or typos or redundancies. The senior co-authors should spend some time looking through the text and make the necessary improvements; this is not the reviewer’s job. Nevertheless, some are listed hereafter (not an exhaustive list):
L40 + L 337: Labile instead of “liable” ?
L 55 : context instead of « contest” ?
L42: “phytoplankton behaviors » is awkward, please rephrase.
L61: most instead of “highest”
L66: “FCO2 efflux » is redundant sinc "F" of "FCO2" abbreviates the word flux.
L68: "high-resolution" is self-evaluation, please simply state instead the actual time step of the data.
L 368: “CO2 drawdown ratio by water quality management” is awkward, please rephrase.
REFERENCES
Abril G & AV Borges (2019) Carbon leaks from flooded land: do we need to re-plumb the inland water active pipe? Biogeosciences, 16, 769-784, https://doi.org/10.5194/bg-16-769-2019
Abril, G. et al. Amazon River carbon dioxide outgassing fuelled by wetlands. Nature 505, 395-398 (2014). https://doi.org/10.1038/nature12797
Borges A.V. et al. (2018) Effects of agricultural land use on fluvial carbon dioxide, methane and nitrous oxide concentrations in a large European river, the Meuse (Belgium), Science of the Total Environment, 610-611, 342-355, https://doi.org/10.1016/j.scitotenv.2017.08.047
Borges AV et al. (2015) Globally significant greenhouse gas emissions from African inland waters, Nature Geoscience, 8, 637-642, https://doi.org/10.1038/NGEO2486
Borges AV et al. (2019) Variations in dissolved greenhouse gases (CO2, CH4, N2O) in the Congo River network overwhelmingly driven by fluvial-wetland connectivity, Biogeosciences, 16, 3801-3834, https://doi.org/10.5194/bg-16-3801-2019
Crawford J.T. et al. (2016) Basin scale controls on CO2 and CH4 emissions from the Upper Mississippi River, Geophysical Research Letters, 43, 1973-1979, 10.1002/2015gl067599
Hotchkiss E et al. (2015) Sources of and processes controlling CO2 emissions change with the size of streams and rivers. Nature Geosci 8, 696–699. https://doi.org/10.1038/ngeo2507
Jones J.B. et al. (2003), Long-term decline in carbon dioxide supersaturation in rivers across the contiguous United States, Geophys. Res. Lett., 30, 1495, doi:10.1029/2003GL017056, 10.
Lauerwald, R. et al. (2015) Spatial patterns in CO2 evasion from the global river network. Global Biogeochem. Cycles 29, 534-554. doi:10.1002/2014GB004941.
Liu, S. et al. (2022) The importance of hydrology in routing terrestrial carbon to the atmosphere via global streams and rivers. Proc. Natl. Acad. Sci. USA 119, e2106322119.
Marescaux A et al. (2018) Seasonal and spatial variability of the partial pressure of carbon dioxide in the human-impacted Seine River in France, Scientific Reports, Scientific Reports, 8(13961), https://doi.org/10.1038/s41598-018-32332-2
Mwanake RM et al. (2023) Anthropogenic activities significantly increase annual greenhouse gas (GHG) fluxes from temperate headwater streams in Germany, Biogeosciences, 20, 3395–3422, https://doi.org/10.5194/bg-20-3395-2023
Nydahl A. C. et al. (2017), No long-term trends in pCO2 despite increasing organic carbon concentrations in boreal lakes, streams, and rivers, Global Biogeochem. Cycles, 31, 985–995, doi:10.1002/2016GB005539.
Prasad M.B.K. et al. (2013) Long-term pCO2 dynamics in rivers in the Chesapeake Bay watershed, Applied Geochemistry, 31, 209-215, https://doi.org/10.1016/j.apgeochem.2013.01.006.
Ran L et al. (2015) Long-term spatial and temporal variation of CO2 partial pressure in the Yellow River, China, Biogeosciences, 12, 921–932, https://doi.org/10.5194/bg-12-921-2015
Ran L. et al. (2021) Substantial decrease in CO2 emissions from Chinese inland waters due to global change. Nat Commun 12, 1730. https://doi.org/10.1038/s41467-021-21926-6
Raymond, P.A. et al. (2013) Global carbon dioxide emissions from inland waters. Nature 503, 355-359. https://doi.org/10.1038/nature12760
Citation: https://doi.org/10.5194/bg-2023-131-RC1 -
AC1: 'Reply on RC1', Mingyang Tian, 02 Jan 2024
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2023-131/bg-2023-131-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Mingyang Tian, 02 Jan 2024
-
RC2: 'Comment on bg-2023-131', Anonymous Referee #2, 19 Sep 2023
This manuscript presents long-term water quality time series data from the Elbe River in Europe. The authors use alkalinity and pH measurements to estimate dissolved CO2 concentrations, which they use to estimate CO2 emissions from the river and tributaries from 1984 to 2018. They then compare the temporal changes in CO2 emissions with the temporal changes in DIC, DOC and POC loads at the watershed’s outlet, along with other water quality parameters. The authors show a decrease in CO2 emissions, which they relate to an improvement in water quality, particularly a decrease in DOC.
The paper suffers from several shortcomings in methodology, a poor presentation of results, and considerable issues with the English language. I must admit this comes as a surprise considering the list of authors, some of whom are widely recognized and respected in the scientific community. I think there is potential to improve this paper substantially, because the dataset holds significant value—but much more guidance will need to be provided by the co-authors. In the following I will elaborate on the three main concerns I have.
Methodological limitations
One limitation is that the entire paper is based on the use of two indirect methods to estimate CO2 emissions. First, pCO2 estimates are indirectly calculated from pH and alkalinity measurements. While this is a common undertaking, the authors must at least provide a quantification of uncertainties. Their plot comparing pCO2 estimates based on two different packages (PHREEQC and CO2SYS) raises concerns as it shows large differences between the two sets of estimates. Second, the CO2 emission estimates lack actual measurements. The authors use an empirical model which was primarily developed for smaller streams and might not be suitable to large rivers. The model results are not evaluated against actual measurements. Again, this needs to be justified (i.e. why was this particular model chosen and not another one?), and an assessment of uncertainties should be presented.
Another critical issue is with the use of discharge values for k600 estimates. From what I gather, the authors have used only one discharge value for each river location. This approach is problematic because k600 is highly influenced by discharge fluctuations, and failing to account for discharge fluctuations will result in erroneous CO2 emission flux estimates. This issue becomes evident in Figure 3f, where the relationship between FCO2 and pCO2 is almost perfectly linear—either suggesting that k600 has no influence on FCO2, or that k600 remains constant across space and time, both of which are improbable.
A third issue is with the DOC data. It appears that two methods are used for the DOC flux estimation, yet only one is presented in the Results section. Furthermore, the first method does not present a way to calculate loads, but simply provides a framework for classifying C-Q patterns, which is rather confusing.
Presentation of results
The results of statistical tests are not consistently reported throughout the paper. For example, Mann-Kendall trend test results are not presented for pCO2 and FCO2 (L231-261) as well as for DIC, DOC and POC (L276-291), making it challenging to assess the significance of the observed trends. Furthermore, there are no reported step change test results, despite the mention of these tests in the Methods section.
I also noted some inconsistent statements between the results and discussion: while on L281 the authors state that “POC, DOC and DIC loads did not show significant trends”, this contradicts the following statement that the DOC load “showed relatively robust decreasing trend” (L310-311).
Lastly, several figures are missing. For instance, the pCO2 time-series data are not shown despite these data being arguably one of the most critical data of the paper.
English language
The paper is very challenging to understand, and clearly the more senior authors (some of whom are well-published) have not provided the necessary feedback. It seems like only the abstract and the first few paragraphs of the introduction have been edited. The language used is awkward at best, and completely incoherent at worst. As a reviewer, I am not willing to invest one or two days correcting grammar and editing the entire paper. I strongly recommend that the senior authors fulfil their responsibilities of reviewing and editing this paper.
Citation: https://doi.org/10.5194/bg-2023-131-RC2 -
AC2: 'Reply on RC2', Mingyang Tian, 02 Jan 2024
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2023-131/bg-2023-131-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Mingyang Tian, 02 Jan 2024
-
CC1: 'Comment on bg-2023-131', Matthias Koschorreck, 20 Sep 2023
Comment on Tian et al. „Long-term reduction in CO2 emissions from the Elbe River due to water quality improvement
Matthias Koschorreck
It is a very good idea to use long term monitoring data to investigate the effect of the socioeconomic changes in Germany after re-unification on GHG emissions from a large river. The paper contains a very nice dataset including both main river and tributary data which allows the investigation of both spatial and inter-annual pattern. However, in my eyes the manuscript does not fully exploit the potential of the dataset and has some serious issues which I would like to address in the following:
- I cannot follow the argumentation that nutrient driven eutrophication should increase CO2 Any CO2 produced from decaying algae was fixed by those algae before. Thus, the cycle of primary production and algae mineralization cannot increase CO2 emissions. In contrast it has the potential to reduce CO2 emissions if algae are buried in the sediments – a scenario relevant for lakes but probably not for rivers.
I would hypothesize that correlation between N or P with CO2 might be a pseudo correlation and not a direct mechanistic link. As written in the manuscript, wastewater contains both DOC and inorganic nutrients. - In the manuscript a rather crude method is used to estimate river surface area. The resulting surface area of 735 km2 (supplement) looks rather high. Divided by river length this means a river width of about 1 km – an unrealistic high value. In Mallast et al. (2020) we determined a surface area of 106 km2 from satellite images.
- The gas transfer velocity was estimated from slope and flow velocity. However, there are also k600 data from River Elbe published (Matoušů et al., 2019). It should at least be checked how estimated k600 data compare to measured ones.
- In Kamjunke et al. (2022) and Kamjunke et al. (2023) it was shown that there is a longitudinal gradient with plankton concentrations increasing downstream the river. It would be interesting to analyze the dataset in this paper with respect to this gradient. Was the transition zone between plankton poor and plankton rich water moving downstream after 1990?
- The dataset also should allow the comparison of different tributaries. Statistical relations between CO2 and other parameters could be checked for each tributary separately. This can be used to investigate the drivers of CO2 in the different sub-catchments. The effect of the tributaries on the main stream, however, is probably difficult to detect. In Bussmann et al. (2022) for example we showed that the high dilution effect at the confluence did not allow the detection of CH4 import from the tributaries into the main river.
- Recent literature shows that CO2 concentrations in rivers fluctuate diurnally (Gómez-Gener et al., 2021). Thus, scaling up CO2 emissions from single datapoints means accepting a systematic uncertainty. Our own measurements show that diurnal fluctuation of CO2 is an issue in River Elbe (manuscript in preparation). This could be relevant in long term time series, if the time of day when samples were taken changed during the time series.
- If CO2 emissions are primarily driven by DOC mineralization the dataset should allow a quantitative comparison between the two. Was DOC decreasing downstream and how does that downstream decrease of DOC compare quantitatively to CO2 emissions? Such a question could be investigated by looking at monitoring data from longer reaches without major tributaries.
- An analysis of long term changes of water quality in river Elbe was recently published by (Wachholz et al., 2022)
references
Bussmann, I., Koedel, U., Schütze, C., Kamjunke, N., and Koschorreck, M.: Spatial Variability and Hotspots of Methane Concentrations in a Large Temperate River, Frontiers in Environmental Science, 10, 10.3389/fenvs.2022.833936, 2022.
Gómez-Gener, L., Rocher-Ros, G., Battin, T., Cohen, M. J., Dalmagro, H. J., Dinsmore, K. J., Drake, T. W., Duvert, C., Enrich-Prast, A., Horgby, Å., Johnson, M. S., Kirk, L., Machado-Silva, F., Marzolf, N. S., McDowell, M. J., McDowell, W. H., Miettinen, H., Ojala, A. K., Peter, H., Pumpanen, J., Ran, L., Riveros-Iregui, D. A., Santos, I. R., Six, J., Stanley, E. H., Wallin, M. B., White, S. A., and Sponseller, R. A.: Global carbon dioxide efflux from rivers enhanced by high nocturnal emissions, Nature Geoscience, 14, 289-294, 10.1038/s41561-021-00722-3, 2021.
Kamjunke, N., Beckers, L. M., Herzsprung, P., von Tumpling, W., Lechtenfeld, O., Tittel, J., Risse-Buhl, U., Rode, M., Wachholz, A., Kallies, R., Schulze, T., Krauss, M., Brack, W., Comero, S., Gawlik, B. M., Skejo, H., Tavazzi, S., Mariani, G., Borchardt, D., and Weitere, M.: Lagrangian profiles of riverine autotrophy, organic matter transformation, and micropollutants at extreme drought, Science of the Total Environment, 828, ARTN 154243
10.1016/j.scitotenv.2022.154243, 2022.
Kamjunke, N., Brix, H., Floeser, G., Bussmann, I., Schuetze, C., Achterberg, E. P., Koedel, U., Fischer, P., Rewrie, L., Sanders, T., Borchardt, D., and Weitere, M.: Large-scale nutrient and carbon dynamics along the river-estuary-ocean continuum, Science of the Total Environment, 890, ARTN 164421
10.1016/j.scitotenv.2023.164421, 2023.
Mallast, U., Staniek, M., and Koschorreck, M.: Spatial upscaling of CO2 emissions from exposed river sediments of the Elbe River during an extreme drought, Ecohydrology, 13, ARTN e221610.1002/eco.2216, 2020.
Matoušů, A., Rulík, M., Tušer, M., Bednařík, A., Šimek, K., and Bussmann, I.: Methane dynamics in a large river: a case study of the Elbe River, Aquat Sci, 81, 12, 2019.
Wachholz, A., Jawitz, J. W., Büttner, O., Jomaa, S., Merz, R., Yang, S., and Borchardt, D.: Drivers of multi-decadal nitrate regime shifts in a large European catchment, Environmental Research Letters, 17, 064039, 10.1088/1748-9326/ac6f6a, 2022.
Citation: https://doi.org/10.5194/bg-2023-131-CC1 -
AC3: 'Reply on CC1', Mingyang Tian, 02 Jan 2024
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2023-131/bg-2023-131-AC3-supplement.pdf
- I cannot follow the argumentation that nutrient driven eutrophication should increase CO2 Any CO2 produced from decaying algae was fixed by those algae before. Thus, the cycle of primary production and algae mineralization cannot increase CO2 emissions. In contrast it has the potential to reduce CO2 emissions if algae are buried in the sediments – a scenario relevant for lakes but probably not for rivers.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 566 | 211 | 45 | 822 | 43 | 28 | 34 |
- HTML: 566
- PDF: 211
- XML: 45
- Total: 822
- Supplement: 43
- BibTeX: 28
- EndNote: 34
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1