Round 1

Reviewer 1 Report

Introduction:

Lines 40-41: There seems to be a jump from the influences/contributions of land use change in peri-urban areas to the pollution control for improved water quality in cities. Is urban runoff impacting the hinterland or are the hinterlands flowing into urban areas? How does recognition (and don’t you mean quantification and identification of source areas) in peri-urban areas impact urban water quality; what is the connection?

Line 87-What is meant by “plain” peri-urban area? Do you mean the study location within the North China Plain?

Line 89-Describe at least the type of models that were combined such as distributed parameter models.

Lines 94-95-The sentence starting with the “The simulation results showed…” only has context if you define the type of models combined and thus the type of output products.

Study area and materials

Study Area

Line 102-Latitude and longitude don’t need to be indicated.

Line 103-What are the populations of the urban area and the suburbs?

Line 104-What is the range in elevation in the south and north that make them relatively higher or lower, respectively?

Line 108-What is the rainfall distribution, the month or months with the greatest rainfall? Annual temperature minimum and maximum?

If land use change is an important factor driving pollutant sources and runoff, then soil type and land use need to be discussed here at least to indicate soil types (indication of susceptibility to runoff) and to quantify the % distribution of land classes; rather than just as Figure 2.

Lines 109-114-This paragraph seems misplaced and would be better in the latter half of the introduction to provide a basis for the need for your study. It would also be appropriate to discuss the actual results and not just refer to “measured data”, particularly in terms of TN and TP which your study is investigating.

Data used in the study

Line 118-What do you mean by “prepared”?

Line 120-Are these the actual flow paths of the major rivers? Have they been channelized or is this a function of the grid analysis of the model?

Line 126-There needs to be more transparency and additional information as to the type of data used for the NPS pollutants. Were data available for TP and TN on an annual basis, monthly basis, event? Were these concentration data or load data in terms of water/runoff? Was the data specific to your study area in terms of soil nutrient content by soil type, by area? Which parameter values were taken from the literature?

Results

Line 251-You have not discussed parameter sensitivity analysis, performance analysis between observed and predicted values i.e. goodness of fit or bias. You have only indicated that the simulation results met the water balance.

Lines 260-261-What do you mean that the results were “basically consistent with the actual situation in the study area”. You need to demonstrate the model goodness of fit through measures such as the Nash-Sutcliffe coefficient. At the very least, you need a statistical analysis of the “actual” data and the simulated output data.

Line 270-Assessing model veracity must go beyond the box and whisker plot (Figure 5).

Line 273-What do you mean the simulation results were “basically reasonable”?

It is difficult to have any confidence in the temporal and spatial variation discussed in TN and TP load when the model has not been validated.

Author Response

For convenience, all the revisions in the revised manuscript are marked in red. However, changes to minor grammatical errors and improper expressions are not shown.

All the numbers of figures, pages and lines are the updated numbers in the revised manuscript.

Comments and Suggestions for Authors from Reviewer 1:

Introduction:

Point 1: Lines 40-41: There seems to be a jump from the influences/contributions of land use change in peri-urban areas to the pollution control for improved water quality in cities. Is urban runoff impacting the hinterland or are the hinterlands flowing into urban areas? How does recognition (and don’t you mean quantification and identification of source areas) in peri-urban areas impact urban water quality; what is the connection?

Response 1: We agree about this “jump” because we misused the phrase “peri-urban areas”; and what we wanted to emphasize are the typical cites of developing countries that have an underdeveloped infrastructure and therefore have relatively lowe sewage treatment capacities. Here, we substituted “peri-urban areas” with “developing cities”. Please see Page 1, Lines 36-38.

As to the title of the paper, we revised “a typical peri-urban area” to “a typical urbanized area”. The word “urbanized” means that more than 50% of the area consists of urban residents. For this case, we selected a typical rapidly developing urbanized area in China as our study area nearly 82.9% of which is composed of urban residents, and we focused on the sustainable development of this area in the future. In 2007, the global urbanization level exceeded 50% for the first time, and it is predicted that the global urbanization level will rise to 61% in 2025 and that the urban population will soar to 5 billion. In 2050, the urban population of developed and developing countries will exceed 86% and 64% respectively (United Nations 2014). Please see Pages 1 and 3, Lines 3-4, 15-17 and 90-93.

Point 2: Line 87-What is meant by “plain” peri-urban area? Do you mean the study location within the North China Plain?

Response 2: I am sorry for the confusion due to the unclear description. We were referring to an urbanized area with relatively flat terrain and small relief. The study area is located in the lower reaches of the Huai River Basin, which flows into the Yellow Sea. Here we revised the text to “a typical urbanized area” and the detailed topographic information was added to the “Study area” subsection. Please see Page 3, Lines 112-116.

Point 3: Line 89-Describe at least the type of models that were combined such as distributed parameter models.

Response 3: We combined the lumped models such as the Xinanjiang model and the water tank model. Please see Page 3, Line 95.

Point 4: Lines 94-95-The sentence starting with the “The simulation results showed…” only has context if you define the type of models combined and thus the type of output products.

Response 4: We deleted it. Please see Page 3, Lines 101-102.

Study area and materials:

Point 5: Line 102-Latitude and longitude don’t need to be indicated.

Response 5: We modified following this suggestion.

Point 6: Line 103-What are the populations of the urban area and the suburbs?

Response 6: We added the total population and the proportion of urban residents. Please see Page 3, Lines 112-114.

Point 7: Line 104-What is the range in elevation in the south and north that make them relatively higher or lower, respectively?

Response 7: The elevation is relatively high at 10.24 m in the south and relatively low at 8.79 m in the north. Relevant information was added. Please see Page 3, Lines 115-116.

Point 8: Line 108-What is the rainfall distribution, the month or months with the greatest rainfall? Annual temperature minimum and maximum?

Response 8: We added relevant information in the revised version. Please see Page 4, Lines 122-127.

Point 9: If land use change is an important factor driving pollutant sources and runoff, then soil type and land use need to be discussed here at least to indicate soil types (indication of susceptibility to runoff) and to quantify the % distribution of land classes; rather than just as Figure 2.

Response 9: We added the distribution of soil types and land classes in the revised manuscript. Please see Page 3, Lines 116-120.

Point 10: Lines 109-114-This paragraph seems misplaced and would be better in the latter half of the introduction to provide a basis for the need for your study. It would also be appropriate to discuss the actual results and not just refer to “measured data”, particularly in terms of TN and TP which your study is investigating.

Response 10: We modified following this suggestion. Please see Page 3, Lines 90-93.

Data used in the study:

Point 11: Line 118-What do you mean by “prepared”?

Response 11: Here, we intended to indicate the process of “collecting”. To avoid ambiguity, we changed the text. Please see Page 5, Line 135.

Point 12: Line 120-Are these the actual flow paths of the major rivers? Have they been channelized or is this a function of the grid analysis of the model?

Response 12: Yes, they are the actual flow paths. Most rivers have been channelized in the study area. We modified in the revised version Please see Page 3, Lines 116-117.

Point 13: Line 126-There needs to be more transparency and additional information as to the type of data used for the NPS pollutants. Were data available for TP and TN on an annual basis, monthly basis, event? Were these concentration data or load data in terms of water/runoff? Was the data specific to your study area in terms of soil nutrient content by soil type, by area? Which parameter values were taken from the literature?

Response 13: We added relevant information in the revised manuscript. Please see Page 5, Lines 138-142 and Table 2.

Results

Point 14: Line 251-You have not discussed parameter sensitivity analysis, performance analysis between observed and predicted values i.e. goodness of fit or bias. You have only indicated that the simulation results met the water balance.

Point 15: Lines 260-261-What do you mean that the results were “basically consistent with the actual situation in the study area”. You need to demonstrate the model goodness of fit through measures such as the Nash-Sutcliffe coefficient. At the very least, you need a statistical analysis of the “actual” data and the simulated output data.

Point 16: Line 270-Assessing model veracity must go beyond the box and whisker plot (Figure 5).

Point 17: Line 273-What do you mean the simulation results were “basically reasonable”? It is difficult to have any confidence in the temporal and spatial variation discussed in TN and TP load when the model has not been validated.

Responses 14-17: As the above four problems involve the validation of the model, our responses are grouped together. The study area is a man-made river basin, of which the boundary is controlled by water conservancy engineering, and the river flow is controlled artificially. There only have water level observations rather than the flow volume. Therefore, a comparison of the simulated runoff with measured data cannot be provided. Here, as an alternative, we performed water balance analysis and rationality analysis of the runoff coefficient to demonstrate the rationality of the simulation results. For such a rain-sewage confluence area, many sewage outlets lack of measured (let alone continuously monitored) data, it is very difficult to verify the results of the simulation directly using observations and therefore it brings some uncertainties in the simulations. In this paper, an on-site investigation was conducted for mining available information, including the layout of sewage networks and their operation situations, the numbers of overflow sewage outlets, the discharges of industrial wastewater and so on. The simulated NPS pollution results in addition to surveys of domestic pollution sources, on-line monitoring of industrial pollution sources, and laboratory tests of sediment contamination were compared with the measured water quality. The simulated average concentration was consistent with the measured water quality. In addition, we compared the simulated results with those of the export coefficient model; their relative errors were less than 10% of the simulation results for the total amount and for each component. Although it is impossible to provide further validation, these analyses still represent progress for a developing city with such a complex situation and limited data. Relevant explanations were added to the discussion section. Please see Pages 17-18, Lines 406-423. In addition, we mentioned the sensitive parameters in the discussion section. Please see Pages 18-19, Lines 417-433.

References:

United Nations, Department of Economic and Social Affairs, Population Division (2014). World Urbanization Prospects: The 2014 Revision, Highlights (ST/ESA/SER.A/352).

Author Response File: Author Response.docx

Reviewer 2 Report

Overall

This study has developed a model to estimate TN and TP annual loads in a Chinese peri-urban area that reflects hydrology and nutrient dynamics. I suspect that there are many useful outputs from this model, however the authors did not present the material in a manner that was convincing. 

For example, in several places the authors indicated that the model was “basically reasonable.”  It is very good that the authors did not oversell the results, however the tone regarding left a great deal of doubt regarding the utility of the modeled information.

Secondly, I suspect that there is a great deal more meaningful results from the model than are reflected in the current presentation.  For example, Figure 9 shows the results from a single storm, but this would be a more impactful discussion if it reflected all storms to releave the modeled results of storm intensity versus dissolved.  Similarly, the point estimates (averages; e.g. Table 6) are far less interesting that modeled distributions vs observed over the full range of data.  

In general, in many places if appears as though the data is presented without careful consideration for how best to present the data or what would be the most useful case for presentation of the model and the associated results.

Finally, in many cases it seems like there was not a careful review of the contents – e.g., in-text reference to figures that are incorrect, ambiguity in labeling, repetitious parameter symbols, inconsistency in labeling, etcs.

Line-by-line comments

L40 – Is there some level or greater specificity that can be included in “

 water pollution is relatively more serious in peri-urban areas.“  For example, are there particularly classes of pollutants that are notably problematic or considered to be the greatest cause of impairment?

L41 – Similarly, can you be more specific in this statement “the accurate recognition of pollutants is of great significance  .“ Do you mean pollutant sources? Classes or specific compounds?

L58 – “Accurate estimations of NPS  “ loads? Compounds? Concentrations?

Table 1 is a nice summary of some of the options – it would be more useful is some citations to representative and pertinent studies were included.

L109 – 114 – this paragraph includes good qualitative high-level information, but it would be nice a more details were included. E.g., some details on development extent (e.g., % land urbanized, population increase).  Another point on pollution – what are the most problematic pollutants and what are the trends in concentrations/loads, etc for those specific pollutants.

Figures 1-3 – lots of really useful info.  Thanks,

Figure 3 – Seems like most of the blue “river” lines are very straight, which makes me think that this might be sewerage system or channals/channels.

Water quantity module – It would be helpful if the input data sources were listed when the parameters are included in the formula (or nearby).  For example, P and EI

L179 on  -- rather than “adsorption state,” I would suggest “sorbed” or “solid-associated” which is more consistent with the published literature with which I am familiar.

3.1 and 3.2 – units are included sporadically (e.g., not included in L152), but it would be more helpful if they were consistently included, and that they manner in which they are indicated is consistent.  E.g., L 147-148, the unit mm is shown after a comma, but shown parenthetically in 169. 

L183 – I believe that “(t)” is to show that it is mass unit of tons, but this is unclear.  Especially since t is also used to indicate time, it would be useful to spell out and be explicit.

L184 – what are the units for soil nutrient content, C, and what is the basis for this?  Can’t verify equation 7 without this info.

L192 – using “C” for two purposes: soil nutrient content and vegetation cover.  One of them should be changed.

L 197+ -- dissolved pollution is discussed, as was solid associated concentrations (C).  Commonly, people will consider an exchange between these two phases, often described by an equilibrium linear distribution coefficient (Kd).  This model may not include this exchange potential, but this should be directly recognized as a simplification.

L203 – lower case k is used in equation 10 but is shown as a capital in text.

L212 – it seems like the pollutant replenishment is only occurring during the precipitation event according to equation 11.  However, I would think that the pollutant would be replenished at all times, that the between storm events would be an important consideration (i.e., this is why some studies have tried to link antecedent dry period to washoff).

Table 6 – is “a” mean annum (or year)?  Better to state more clearly the units. Additionally, I think that there is likely much more interesting information that comes from this analysis then merely the averages.  Perhaps a box-and-whisker plot or something similar to show the ranges/distributions of the data from the 18 year period.

Table 7 – units are also tons per year, but for the specified year of 2018.  In other words, the data produced is directly comparable in Table 7 (specified single year = 2018) to the averages shown in Table 6.

I am also assuming that the TN is the summation of sorbed pollutant and dissolved pollutant.

What is “Life pollution?”

Figure 5 and associated text – I may have misunderstood the model, but do you only have a single estimate for the full year?  (From the model description, I thought that there were several points per storm.)  If you have more than 1 point for the year, I think that it would be more useful to present the range of results from the simulation to the range of measured results.  This could be done as a box-and-whisker plot, but would be even more helpful as a scatter plot so that paired points could be directly compared.

L289 - 292 – I don’t see a Figure 6a/b.  Maybe you mean Figure 7b?

Figure 11 – unclear what the different graphs in the tryptic represent.  E.g., a is labeled as average but is referred to as “river” in the text.  B and c are both labeled “sub-basin.”

Author Response

For convenience, all the revisions in the revised manuscript are marked in red. However, changes to minor grammatical errors and improper expressions are not shown.

All the numbers of figures, pages and lines are the updated numbers in the revised manuscript.

Comments and Suggestions for Authors from Reviewer 2:

Overall:

This study has developed a model to estimate TN and TP annual loads in a Chinese peri-urban area that reflects hydrology and nutrient dynamics. I suspect that there are many useful outputs from this model, however the authors did not present the material in a manner that was convincing. For example, in several places the authors indicated that the model was “basically reasonable.”  It is very good that the authors did not oversell the results, however the tone regarding left a great deal of doubt regarding the utility of the modeled information.

Secondly, I suspect that there is a great deal more meaningful results from the model than are reflected in the current presentation. For example, Figure 9 shows the results from a single storm, but this would be a more impactful discussion if it reflected all storms to releave the modeled results of storm intensity versus dissolved. Similarly, the point estimates (averages; e.g. Table 6) are far less interesting that modeled distributions vs observed over the full range of data. 

In general, in many places if appears as though the data is presented without careful consideration for how best to present the data or what would be the most useful case for presentation of the model and the associated results.

Finally, in many cases it seems like there was not a careful review of the contents – e.g., in-text reference to figures that are incorrect, ambiguity in labeling, repetitious parameter symbols, inconsistency in labeling, etcs.

Response: We thank for your valuable comments. We revised our manuscript following the suggestions.

Line-by-line comments:

Point 1: L40 – Is there some level or greater specificity that can be included in “ water pollution is relatively more serious in peri-urban areas.“  For example, are there particularly classes of pollutants that are notably problematic or considered to be the greatest cause of impairment?

Response 1: What we wanted to emphasize is the typical cities of developing countries, that have an underdeveloped infrastructure and therefore have relatively low sewage treatment capacities. Here, we changed “peri-urban areas” to “developing cities”. Please see Page 1, Lines 36-40.

Point 2: L41 – Similarly, can you be more specific in this statement “the accurate recognition of pollutants is of great significance. “ Do you mean pollutant sources? Classes or specific compounds?

Response 2: Yes, we intended to indicate the classes of the pollutant sources. We modified in the revised version. Please see Page 1, Lines 38-40.

Point 3: L58 – “Accurate estimations of NPS “ loads? Compounds? Concentrations?

Response 3: The NPS loads refer to compounds.

Point 4: Table 1 is a nice summary of some of the options – it would be more useful is some citations to representative and pertinent studies were included.

Response 4: We added citations to the revised text. Please see Table 1.

Point 5: L109 – 114 – this paragraph includes good qualitative high-level information, but it would be nice a more details were included. E.g., some details on development extent (e.g., % land urbanized, population increase).  Another point on pollution – what are the most problematic pollutants and what are the trends in concentrations/loads, etc for those specific pollutants.

Response 5: We added relevant information to Page 3, Lines 90-95, 112-114.

Point 6: Figures 1-3 – lots of really useful info. Thanks.

Response 6: Thanks.

Point 7: Figure 3 – Seems like most of the blue “river” lines are very straight, which makes me think that this might be sewerage system or channals/channels.

Response 7: Most rivers have been channelized. We added this information to the revised version. Please see Page 3, Lines 116-117.

Point 8: Water quantity module – It would be helpful if the input data sources were listed when the parameters are included in the formula (or nearby).  For example, P and EI

Response 8: We added information to the “Data used in this study” subsection. Please see Page 5, Lines 135-142 and Table 2.

Point 9: L179 on  -- rather than “adsorption state,” I would suggest “sorbed” or “solid-associated” which is more consistent with the published literature with which I am familiar.

Response 9: We changed as suggestion.

Point 10: 3.1 and 3.2 – units are included sporadically (e.g., not included in L152), but it would be more helpful if they were consistently included, and that they manner in which they are indicated is consistent.  E.g., L 147-148, the unit mm is shown after a comma, but shown parenthetically in 169.

Response 10: We changed following your suggestion.

Point 11: L183 – I believe that “(t)” is to show that it is mass unit of tons, but this is unclear.  Especially since t is also used to indicate time, it would be useful to spell out and be explicit.

Response 11: Yes, you are correct. In that “(t)” indicates tons. We corrected in the revised version.

Point 12: L184 – what are the units for soil nutrient content, C, and what is the basis for this?  Can’t verify equation 7 without this info.

Response 12: The units of C are g/g, which is the mass of N or P per unit mass of soil from soil data. We added the units in the revised version. Please see Page 8, Line 200.

Point 13: L192 – using “C” for two purposes: soil nutrient content and vegetation cover. One of them should be changed.

Response 13: We changed the abbreviation of vegetation cover to C_vm. Please see Page 8, Lines 206-207.

Point 14: L 197+ -- dissolved pollution is discussed, as was solid associated concentrations (C).  Commonly, people will consider an exchange between these two phases, often described by an equilibrium linear distribution coefficient (Kd).  This model may not include this exchange potential, but this should be directly recognized as a simplification.

Response 14: We agree; thus, we simplified the exchange potential. We added a relevant explanation in the context. Please see Page 8, Line 195-196.

Point 15: L203 – lower case k is used in equation 10 but is shown as a capital in text.

Response 15: We corrected this discrepancy.

Point 16: L212 – it seems like the pollutant replenishment is only occurring during the precipitation event according to equation 11.  However, I would think that the pollutant would be replenished at all times, that the between storm events would be an important consideration (i.e., this is why some studies have tried to link antecedent dry period to washoff).

Response 16: In this paper, pollutant replenishment in the pollution generation stage was considered at all times. Equation 11 concerns the emission progress. During irrigation and drainage, farmlands can release pollution during dry periods. This process is very important and worth considering. However, it is also very complicated. In view of the dominant pollution load discharged during storm events, we ignore the dry period in this paper. In addition, there is no need to consider pollutant replenishment emissions during the dry period for urban areas or livestock and poultry breeding. Please see Page 8, Lines 223-227.

Point 17: Table 6 – is “a” mean annum (or year)?  Better to state more clearly the units. Additionally, I think that there is likely much more interesting information that comes from this analysis then merely the averages. Perhaps a box-and-whisker plot or something similar to show the ranges/distributions of the data from the 18 year period.

Response 17: Thank you. We revised the unit to “yr”. In this part, we mainly carried out a rational analysis. In section 4.3, we analyzed the variations of annual pollution load. In section 5.1, we compared the unit load with other relevant studies in China.

Point 18: Table 7 – units are also tons per year, but for the specified year of 2018.  In other words, the data produced is directly comparable in Table 7 (specified single year = 2018) to the averages shown in Table 6. I am also assuming that the TN is the summation of sorbed pollutant and dissolved pollutant.

Response 18: Figures 7 and 8 show the annual average TN and TP load production, respectively, from 2000 to 2018. TN is the summation of sorbed pollutants and dissolved pollutants.

Point 19: What is “Life pollution?”

Response 19: We intended to indicate “Domestic pollution”. We have revised the text.

Point 20: Figure 5 and associated text – I may have misunderstood the model, but do you only have a single estimate for the full year?  (From the model description, I thought that there were several points per storm.)  If you have more than 1 point for the year, I think that it would be more useful to present the range of results from the simulation to the range of measured results.  This could be done as a box-and-whisker plot, but would be even more helpful as a scatter plot so that paired points could be directly compared.

Response 20: We only have one year of effective water quality data because in many other years, the concentration of pollutants in the rivers of this basin had been diluted through an irregular schedule. Thus, it would be difficult to use this data to demonstrate our results.

Point 21: L289 - 292 – I don’t see a Figure 6a/b.  Maybe you mean Figure 7b?

Response 21: Yes, we have corrected the figure. Thank you.

Point 22: Figure 11 – unclear what the different graphs in the tryptic represent.  E.g., a is labeled as average but is referred to as “river” in the text.  B and c are both labeled “sub-basin.”

Response 22: We revised it. Please see Page 15, Line 352-358 and Figure 11.

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript simulate NPS in a developed area. Overall, authors have used a novel method and put too much efforts in this work. For a better read, I suggest that the authors should improve the quality of their images, decrease number of abbreviations, and clearly define the applied units. Proof-read of the manuscript by a native English speaker may also help to improve the quality of the paper. 

Author Response

For convenience, all the revisions in the revised manuscript are marked in red. However, changes to minor grammatical errors and improper expressions are not shown.

All the numbers of figures, pages and lines are the updated numbers in the revised manuscript.

Comments and Suggestions for Authors of Reviewer 3:

The manuscript simulate NPS in a developed area. Overall, authors have used a novel method and put too much efforts in this work. For a better read, I suggest that the authors should improve the quality of their images, decrease number of abbreviations, and clearly define the applied units. Proof-read of the manuscript by a native English speaker may also help to improve the quality of the paper.

Response: Thank you for your comments and suggestions. We have improved our manuscript accordingly.

Author Response File: Author Response.docx