the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Interannual variabilities, long-term trends, and regulating factors of low-oxygen conditions in the eastern Pearl River Estuary
- 1School of Environmental Science and Engineering, Sun Yat-sen University, Guangzhou, 510275, China
- 2Guangdong Provincial Key Laboratory of Environmental Pollution Control and Remediation Technology, Guangzhou, 510275, China
- 3Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, 519000, China
- 4Department of Oceanography, Dalhousie University, Halifax, Nova Scotia, B3H 4R2, Canada
- 5Center for Water Resources and Environment, Sun Yat-sen University, Guangzhou, 510275, China
- 6Guangdong Zhihuan Innovative Environmental Technology Co., Ltd, Guangzhou, 510030, China
- 1School of Environmental Science and Engineering, Sun Yat-sen University, Guangzhou, 510275, China
- 2Guangdong Provincial Key Laboratory of Environmental Pollution Control and Remediation Technology, Guangzhou, 510275, China
- 3Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, 519000, China
- 4Department of Oceanography, Dalhousie University, Halifax, Nova Scotia, B3H 4R2, Canada
- 5Center for Water Resources and Environment, Sun Yat-sen University, Guangzhou, 510275, China
- 6Guangdong Zhihuan Innovative Environmental Technology Co., Ltd, Guangzhou, 510030, China
Abstract. The summertime low-oxygen conditions in the Pearl River Estuary (PRE) have experienced a significant expansion in spatial extent associated with notable deoxygenation in recent decades. Nevertheless, there is still a lack of quantitative understanding of the long-term trends and interannual variabilities in oxygen conditions in the PRE as well as the driving factors, which was comprehensively investigated in this study using monthly observations in the eastern PRE during 1994–2018. To evaluate the changes in scope and intensity of oxygen conditions, an indicator (defined as the Low-oxygen Index, LOI) that integrates several metrics related to low-oxygen conditions was introduced through the principal component analysis (PCA). Moreover, primary physical and biogeochemical factors controlling the interannual variabilities and long-term trends in oxygen conditions were discerned, and their relative contributions were quantified by the multiple regression analysis. Results showed that the regression models explained over 60 % of the interannual variations in LOI. Both the wind speeds and concentrations of dissolved inorganic nitrogen (DIN) played a significant role in determining the interannual variations (by 39 % and 49 %, respectively) and long-term trends (by 39 % and 56 %, respectively) in LOI. Due to the increasing nutrient loads and alterations in physical conditions (e.g. the long-term decreasing trend in wind speeds), coastal eutrophication was exaggerated and massive marine-sourced organic matter was subsequently produced, thereby resulting in an expansion of intensified low-oxygen conditions. It has also driven a shift in the dominant source of organic matter from terrestrial inputs to in situ primary production, which has probably led to an earlier onset of hypoxia in summer. In summary, the eastern PRE has undergone considerable deterioration of low-oxygen conditions in the context of substantial changes in anthropogenic eutrophication and external physical factors.
Zheng Chen et al.
Status: final response (author comments only)
-
RC1: 'Comment on bg-2021-358', Anonymous Referee #1, 04 Feb 2022
Review Chen et al. 2022, Interannual variabilities, long-term trends, and regulating factors of low-oxygen conditions in the eastern Pearl River Estuary, MS No.: bg-2021-358
General Comments :
Chen et al. evaluated long-term patterns in DO in the eastern Pearl River Estuary (PRE) across seasons and regions, computed an aggregated metric of low DO, and then tested possible controlling factors of it with multiple regression. They found dissolved nitrogen and wind speed were the most explanatory variables for interannual variations and long-term trends. They use additional water quality observations to evaluate the changes to the system over time and hypothesize shifts in the system dynamics. Overall, this is a very interesting study making good use of a long-term data set to evaluate long-term change. I appreciate their thorough treatment of the data both spatially and temporally. My major comments involve clarifying the methods and what is represented in some of the graphics. Clarification is needed throughout as to which months of data are included in different average results and how the data is aggregated to represent “summer”. In addition, more clarification is needed on the PCA approach as well as some re-organization of which information is presented in the Methods or Results.
Specific Comments:
- Lines 60 – 81: Within this section, please incorporate the reasoning for your focus on the Eastern PRE. Can you describe whether this region was selected from the larger PRE because this is where the longest term data is, or is it because this is where the lowest oxygen occurs? It would provide more context if you included some description in the Introduction about how water quality in this eastern region compares to the rest of the estuary.
- Line 104 – Please describe what spatial interpolation approach was used in MATLAB for the interpolations. Also, since you have land in between some of the stations, how did the method deal with that? It would be helpful to show what the region looks like in vertical cross-section as a 2nd panel of Figure 1 with the sample locations and depths indicated with dots. This would be like one of the panels of Figure A2, showing which depths each station is sampled at. This would be a helpful way to visualize the depths at each station.
- Lines 108-120: This discussion of the PCA needs modification. Please include a table of the variables used in the PCA. I kept having to look back in the text to see how “low oxygen”, “Area3,” etc, were defined. I’d suggest including just that table and a description of the approach here in the Methods section. The resulting equation (Line 117) and description of it should probably be moved to the Results section. Also, please summarize the rest of the PCA results (in an appendix table), such as what % of variance the other components had, and what their weights were.
- Line 125 – show an equation to describe this standardization
- Lines 123-134: There need to be some discussion of these different test results in the Results section.
- Figure 4a – can you describe the values plotted here more? Is the minimum, mean and range just from the bottom observations, or is it generated from the interpolation?
- Figure 4 (b) and (c)– We need information on the spatial interpolation to get the area and thickness. Also, if samples are collected every month, it is unclear what the bars in (b) and (c) represent. Are they the average of each month’s spatially-aggregated values? If so, please put range bars on each bar to show the range across the summer months. Or pick one month to show.
- Figure A2 – Similarly to Figure 4, specify which month of the summer these plots are for. If they are average of all the summer cruises, please justify that approach.
- Figure 5 – The min and mean DO symbols in legend seem switched.
- Figure 5a-h – I’d like to see the surface and bottom graphs with the same vertical scale (0 to 10). It can be confusing to have them different when they are right next to each other.
- Figure 5 – I’m unsure from the descriptions as to how the mean and minimum were calculated with multiple stations and months of the summer. Is the minimum the absolute minimum observed in that region in the summer, or an average of the lowest value across the stations? Also is the mean a spatial and temporal mean across the summer?
- Figure 6 – The really high values in the range in recent years in July are worth mentioning. Is that just one location that is causing that range to increase, or is it some indication of increased variability?
- Line 233 - A diagram or flow-chart that describes the sampling and cases used in the regression analysis to get to the results would help my understanding (and probably other readers) of the methods. This could go in the Appendix.
- Figure 9 – The wind speed decrease over time seems very large. Because the results indicate this is an important variable, this deserves more discussion or investigation. If the authors already know other work that has investigated decreasing wind speeds, please cite it and describe briefly. But if there is no other research explaining this wind decrease, it would be a good idea to double-check the data and be sure that it is not an artifact of sampling dates or density shifting or sensor height changing.
- Appendix Figure A1 is important b/c it doesn’t suffer from any possible aggregation or averaging bias. It might be useful to make an addition panel that shows how the bottom summer counts have changed over time – maybe make one for the first half and one for the 2nd half of the record. This could also show if there’s a spatial shift.
Technical Corrections:
- Abstract, Line 15 – change “was” to “were”
- Abstract, line 17 – suggest changing “through the principal component analysis” to something else. Maybe “as a result of a principal component analysis”
- Abstract, line 25 – It is unclear what “It” refers to in this sentence. Please re-write.
- Abstract, last sentence – the phrase “in the context of” is fairly awkward. Consider re-wording this sentence to make your summary stronger.
- Intro, Line 33 – suggestion you use “organisms” instead of “creature”
- Intro, Line 43-45 – Simplify (or remove) this sentence since the next few sentences cover a lot about oxygen depletion. I’d suggest just “Terrestrial organic matter discharged to estuaries can lead to intense microbial respiration.”
- Intro, Line 54: For the Ni et al. 2020 paper, it is important to change “ocean” to “estuary.” They did not study the external impact of the Atlantic Ocean warming on the Chesapeake Bay.
- Methods, Lines 84-93 – Who collected this data?
- Results, Line 144 – I do not think the word “varied” is correct here.
- Results, Line 168 – wording like this sentence can be simplified. You could just start with “DO concentrations exhibited significant…”
- Results, Line 192, and other places – The phrase “DO content” is not something I’ve seen very much before in the hypoxia literature. I’d suggest using “DO concentrations” or just “DO”.
- Figure 6 – It would be helpful to use the same open circles for the blue symbols as in Figure 5.
- Discussion, Lines 297-298: Please revise the sentence that starts with “As quantified by statistic methods…” to work on the wording. Maybe “Our analysis showed that increasing DIN…”
-
AC1: 'Reply on RC1', Jiatang Hu, 01 Apr 2022
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2021-358/bg-2021-358-AC1-supplement.pdf
-
RC2: 'Comment on bg-2021-358', Anonymous Referee #2, 22 Feb 2022
The manuscript by Chen et al. provides data analysis (mostly statistical) of the water quality observations in the southeastern part of the Pearl River Estuary (PRE). Several recent studies, such as Hu et al., 2021, and Li et al., 2021, about the de-oxygenation problems in this region, provide background and justification of hypoxia-related study in PRE. Nevertheless, I feel this manuscript fails to connect the new data with the findings of these existing studies and covers only a small portion of the PRE; thus, its potential to convey what can be learned from a regional study to a broader audience is limited.
A critical flaw of this study is the representativeness of the stations that all statistical analyses are built upon. I acknowledge these are valuable (30 yr) monthly data covered by these stations, yet their spatial coverage is mainly surrounding the Hongkong island; it is a challenge to draw any solid conclusion regarding the PRE, even the east part, without a throughout cross-reference with the model and data by the previously mentioned two studies. Instead of using these stations to refer to east PRE, I would say the author should do the opposite—to provide a possible water quality study of the Hongkong coast with impacts from the PRE.
Built on the above point, I see in line 103 (page 4), “the observed DO profiles were interpolated by MATLAB along with the three subregions with a grid resolution of 600 m (length) â0.3 m (depth).” But take a look at the distribution of the stations; they are concentrated mainly in the nearshore area surrounding Hongkong; I doubt they can be representative of the condition in the east part of the PRE (e.g., the ECTZ defined by Li et al., 2021). And any conclusion based on such “hypoxia area” analysis (e.g., Fig. 4 ) is thus questionable. In addition, the water depth of these stations varies from 5 m to more than 30m, and the analysis in this manuscript concentrated mostly surface and bottom, which impair the reliability of such analysis in the de-oxygenation study, which is very sensitive to water depth and vertical distribution of variables.
In the abstract, the author indicated that “there is still a lack of quantitative understanding of the long-term trends and interannual variabilities in oxygen conditions in the PRE as well as the driving factors, which was comprehensively investigated in this study,” which I could not agree, I think Hu et al. and Li et al. provided good studies about the mechanism of oxygen dynamics in the PRE. Yet, this manuscript fails to connect what is observed by the stations surrounding Hongkong to what has been reported in a larger geospatial content (PRE and coastal/shelf water).
The author uses wind speed in their statistic analysis which is also questionable. How about wind direction, and does the wind play the same role in low oxygen development over a year? I am asking because Li et al. (2021) indicated that both wind direction and intensity influenced the circulation nutrient flux, detritus, and vertical mixing. Also, as suggested by Li et al., 2021, what is the role of shelf circulation in physics (mixing, etc.) and nutrient and sediment delivery? Li et al. (2021) and Feng et al. (2014) show that the upwelling and downwelling favorable wind condition has different impacts on the low-oxygen development in this area. It is problematic to use the monthly mean wind speed as a predictor without looking into wind’s detailed role in this environment.
The author’s conclusion that the eastern PRE would “develop into a severe hypoxic state within the next two decades” is too strong to be supported by the analysis provided by this study. For instance, what do the wind, large-scale circulation (cause it affects lateral delivery of water and nutrient, etc.), and river (Pearl plus wastewater from the city) look like in the next two decades? The Pearl River discharged into the PRE from the north, yet if we focus on the spatial scale covered by the stations in this study, what is Pearl River’s role in MM stations? Also, what is the impact of overland runoff from Hongkong, such as wastewater discharge, which is also indicated by Hu et al. (2021), and the author briefly mentioned this in Line 250 of page 8.
-
AC2: 'Reply on RC2', Jiatang Hu, 01 Apr 2022
The comment was uploaded in the form of a supplement: https://bg.copernicus.org/preprints/bg-2021-358/bg-2021-358-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Jiatang Hu, 01 Apr 2022
Zheng Chen et al.
Zheng Chen et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 458 | 66 | 18 | 542 | 5 | 8 |
- HTML: 458
- PDF: 66
- XML: 18
- Total: 542
- BibTeX: 5
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1