the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 29 Jan 2021
Can the implementation of Low Impact Development reduce basin runoff?
Abstract. Low impact development (LID) was promoted as an alternative to conventional urban drainage methods. The effects of LID at site or urban scales have been widely evaluated. This project aims to investigate the impact of LID implementation on basin runoff at regional scale in a half urbanized catchment; especially the overlap of urban and rural sub-flows at peak times is concerned. A SUPERFLEX conceptual model framework was adapted as a semi-distributed model to simulate the rainfall-runoff relationship in the catchment for San Antonio, Texas as a case study. Scenario analyses of both urban development and LID implementation were conducted. Results show that (1) the infill urban development strategy benefits more from runoff control than the sprawl urban development strategy; (2) in non-flood season permeable pavements, bioretention cells, and vegetated swales decrease peak runoff forcefully and permeable pavements, bioretention cells, and green roofs are good at runoff volume retention; (3) contrary to the general opinion about the peak reduction effect of LID, for partly urbanized, partly rural basins and extremely wet conditions, the implementation of LID practices delays urban peak runoff and may cause stacking of rural and urban sub-flows, leading to larger basin peaks.
- Preprint
(1547 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on hess-2021-32', Konstantinos Soulis, 25 Feb 2021
Review of the manuscript entitled: “Can the implementation of Low Impact Development reduce basin runoff?” by Xinxin Sui and Frans van de Ven
In this paper a modeling study on the potential impact of various scenarios of future urban expansion and various scenarios of low impact development interventions in a case study site in the catchment of San Antonio in Texas, USA is presented. The concept of the study is really interesting and within the scopes of HESS journal. The study also involves laborious work and an interesting case study site.
The main novelty of the paper is related to the investigation of the case of a large urban area located at the downstream part of a rural watershed. Several urban expansion scenarios and LID practices application scenarios are investigated. Accordingly, the study could potentially provide interesting information for urban planning. However, the most prominent findings of this study; i.e. that LID may result in increased peaks and that a more compacted and denser urban area may result in decreased peak discharges; are somehow controversial. The reasoning behind them is logical and I can understand that under very specific conditions this could be possible. However, the key question is what is the probability of this to happen? Is it a rare coincidence or a typical outcome for big storm events in the case of a large urban area located at the downstream part of a rural watershed? I believe that the methodology followed in this study cannot safely answer this question for two main reasons.
The first reason is related to the model used in this study. The model in the manuscript is described as semi-distributed. As I can understand by reading the manuscript, in reality it is a simple, cascading buckets, lumped model applied in the entire watershed as a whole with the same lumped parameters values for the entire watershed. Probably the model is described as semi-distributed because it involves different representations for the various parts and processes (rural, urban, LID). The model seems to perform well anyway, even if the calibration and validation procedures are not adequately presented. Accordingly, this model could be adequate for other types of studies. The problem is though that the model (as I can understand) was calibrated and validated only for the current conditions (without LID). The model applied for the LID scenarios involves more components (lags and storages) and the involved parameters values were assumed from the literature, which is really challenging and uncertain. Therefore, given that the model is lumped and empirical, its performance for the future conditions and especially for the magnitude and the timing of the flow components is questionable. Accordingly, the key finding of the study that LID may result in increased discharge peaks cannot be justified by the model results as the concurrence of the runoff peaks coming from the rural and the urban parts of the watershed due to LID could be a coincidence considering the involved uncertainties. For example, in reality parts of the urban watershed may deliver runoff to the river faster and other parts slower depending on the location of the various LID components and the resulted peak will be the mix of spatially distributed processes.
I would also like to mention at this point that while for the justification of the presented mechanisms for peaks increases, the studied events hydrographs are presented in detail in some figures, the performance of the model for the same events isn’t clear as only some general performance metrics for the entire time series are presented. Considering the emphasis that is given to the timing and magnitude of sub-event peaks in this study, I would expect a more detailed presentation of the model performance on predicting these events hydrographs. Accordingly, model calibration, validation and performance at event scale should be described in more detail and a figure (and some metrics) presenting model accuracy at event scale is required.
The second reason is that the model is applied for a short period including a small number of flood events that are analyzed to provide the above-mentioned conclusions. Accordingly, it isn’t possible to understand if these conditions are typical or limited to the investigated events.
Considering that the title of this paper is “Can the implementation of Low Impact Development reduce basin runoff?” the reader expects that the applied methodology should be able to provide a clear and justified answer.
Another important weakness of this study is related to the presentation of the methodology. There are many unclear parts and some other parts are not described at all. I believe that the model structure and the procedure followed to decide it should be presented more clearly. Most important, the identification of the parameters values for the model describing LID practices should be presented in detail and justified. In the manuscript there is just a mention of some previous studies. It should be also described in detail how the model was calibrated and validated (data used, methods, locations etc). Finally, a paragraph describing the model application, the data used, the modeled events characteristics etc. and the methodology and the steps followed for the analysis of the results is required.
Finally, a general weakness of this paper is that the results and the main findings are not discussed in the light of previous relevant studies in the results section, and most importantly in the discussion section. There are numerous previous studies on the topics presented in this study but there aren’t any citations and real discussion in the discussion section or in the results section.
Apart from the above main comments the language includes many errors and should be improved. I am not a native English speaker though and therefore I avoided providing specific suggestions on this matter. The presentation, figures etc. are generally good. There is only a problem with the definition of symbols and abbreviations that makes a little difficult to understand some parts of the methodology especially the related tables.
I am providing some additional specific comments in a commented version of the manuscript.
Based on the fundamental weaknesses of the paper that I explained above I am not convinced that this paper can be adequately improved to be suitable for publication. However, I have some reservations because some parts of the methodology presentation were confusing or missing and therefore there is a chance that I wasn’t able to accurately understand what exactly has been done in reality and to effectively evaluate the manuscript. Accordingly, as the concept and the main idea behind this study are very interesting and it involves laborious work, I am recommending a major revision to provide the authors the opportunity to respond to my main concerns and to address the key limitations of this study.
(Please also see the attached pdf file)
Best Regards
Konstantinos X. Soulis
-
CC1: 'Reply on RC1', Xinxin Sui, 22 Mar 2021
Response to comments by referee #1 (Konstantinos X. Soulis)
We would like to thank the referee, Konstantinos X. Soulis, for the careful and thorough reading of this manuscript and the thoughtful comments and constructive suggestions, which help improve this manuscript's quality. Our detailed responses could be found in the attached letter.
Yours sincerely,
Xinxin Sui and Frans van de Ven
-
AC2: 'Reply on RC1', Frans van de Ven, 20 Jul 2021
Responses to comments by referee #1 (Konstantinos X. Soulis)
Please see the attached pdf, as Figures are missing in the text below and the lay-out whas crippled while copying from teh Word document.
We would like to thank the referee, Konstantinos X. Soulis, for the careful and thorough reading of this manuscript and the thoughtful comments and constructive suggestions, which help improve this manuscript's quality. Our response follows:
Reviewer #1:
- The hydrological model description isn't sufficient neither clear enough. The model used is described as semi-distributed. As I can understand by reading the manuscript, in reality it is a simple, cascading buckets, lumped model applied in the entire watershed as a whole with the same lumped parameters values for the entire watershed. Probably the model is described as semi-distributed because it involves different representations for the various parts and processes (rural, urban, LID) but this isn't accurate. (please see the relevant general comment) (page 6, line 189)
Response: We appreciate this feedback. An additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application. Besides, some supplementary descriptions about the model used in this research were added in Section 3, Methodology, in which the model was described as a semi-distributed model with two hydrological response units (HRUs) for urban and rural surfaces, respectively. Therefore, there are two different hydrological systems for urban and rural areas separately. That’s why it’s called as a semi-distributed model.
This model is similar to the FLEX-Topo model (which was introduced in the added Section 2.3 SUPERFLEX model framework). The FLEX-topo model is also a semi-distributed model with three HRUs (plateau, hillslope, and wetland). Inspired by FLEX-topo, this model was created by separating the hydrological systems in urban and rural surfaces of one catchment.
- A very short description of the main characteristics of this model would be helpful. (page 6, line 190)
Response: We thank you for this suggestion. As mentioned before, an additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application.
- was? (page 7, line 205)
Response: We thank the reviewer for pointing this out. It was revised.
- On what basis? How? (page 7, line 214)
Response: This is about using expert knowledge to increase the efficiency of parameter calibration. The sentence in the article was rewritten as "For parameter calibration, the initial range of each parameter was given roughly based on the experience and information about catchment properties according to the physical realism of these parameters (Breuer et al., 2003; Gharari et al., 2014)."
It should be mentioned that most of the parameters in SUPERFLEX model have physical meanings, just like the popular TOPMODEL. For example, D (Precipitation distribution factor for Su) indicates the area ratio of unsaturated zone in rural areas or green (permeable) surfaces in urban areas. Sumax (Maximum unsaturated storage) could be estimated from the depth of soil layer and porosity. Therefore, a pair of relatively broad preliminary limitations were given based on the empirical value for more efficient parameter calibration and to prevent unrealistic calibration results.
Breuer, L., Eckhardt, K., and Frede, H. G.: Plant parameter values for models in temperate climates, Ecological Modelling, 169(2-3), 237-293., 2003
Gharari, S., Hrachowitz, M., Fenicia, F., Gao, H., and Savenije, H. H.: Using expert knowledge to increase realism in environmental system models can dramatically reduce the need for calibration, Hydrology and Earth System Sciences, 18(12), 4839-4859, doi:10.5194/hess-18-4839-2014, 2014
- This isn't clear. I cannot understand what has been actually done. (page 7, line 214)
Response: This is about using Monte Carlo methods for parameter calibration. After the decision of calibration limitations (as mentioned in the last response), the Monte Carlo sampling method was used to generate random pair of parameters between the limitations. In this research, the semi-distributed model was created by testing the model performance from six-bucket models to eight-buckets models. As more complex models have more parameters, the same number of samples could be seen as "unfair". Therefore, more complex models with more parameters were tested with larger numbers of parameter sample sets to ensure the calibration scale as "fair" as possible.
- Some of the symbols are not described (e.g. Su, Sf, Sh, Ss). Subscripts in the parameters symbols would be helpful (the symbols should be presented in the same way as in the mathematical expressions and parameters and components description (e.g. Italics with subscripts) (page 8, line 223)
Response: It is about some symbols in Table 2. Actually, Table 2 was introduced after Figure 3 Schematic figure of semi-distributed model structure. The bucket names and their symbolic expression (Su, Sf, Sh, Ss) were introduced under Figure 3. The name of Table 2 was revised as Parameters and model components in the semi-distributed model as shown in Figure 3, which may help to avoid confusion.
- Please make clear what was done (specific process / steps) in a simple way. At some point at the beginning, a short description of the model type and characteristics should be provided (e.g. semi-distributed, cascading bucket model, event based? etc.) (page 7, line 203)
Response: The first piece of advice is about model setup. An additional paragraph about the strategies used in model setup was added in Section 3.2.1 Semi-distributed model setup. For the second suggestion about the description of model, as mentioned before, an additional section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application.
- Also unclear. What do you mean by the term verified NSEs and what by the phrase "calculated as accuracy indexed" (page 7, line 209)
- How the "variance of the verified NSEs" is a measure of precision? (page 7, line 209)
Response to 8 and 9: These involve the specific procedure of model structure selection. As mentioned in the revised manuscript and the responses above, 1) the Monte Carlo sampling method was used to generate pairs of parameters, and 2) both the Nash-Sutcliffe Efficiency (NSE) and correlation coefficient (R2) were used to evaluate the model performance. Therefore, each pair of data would get two NSE and R2 values as calibration and validation results, respectively.
To obtain a model structure with both stable (high precision) and outstanding (high accuracy) performance, the optimal, mean, and variance of those verified NSE and R2 were examined to evaluate the accuracy and precision of model structures, as the table below.
Accuracy index
Precision index
Model Num.
The R2 of calibration
Optimal verified R2
The NSE of calibration
Optimal verified NSE
Mean of verified NSE
Variance of verified NSE
M01A
0.882
0.865
0.723
0.735
0.117
0.399
M01B
0.897
0.873
0.673
0.748
0.297
0.240
M02
0.863
0.871
0.712
0.749
0.155
0.424
M03A
0.902
0.850
0.721
0.706
-0.056
0.758
M03B
0.815
0.855
0.656
0.730
0.039
0.587
M04
0.878
0.883
0.714
0.744
0.036
0.606
Symbols: Boldface row indicates the selected model structure; Italic columns indicate the calibration results.
Since the trivial details is quite complex and of little importance to the main content of the article, the specific procedures was neglected in the manuscript to avoid unnecessary confusion of readers.
- Below, there is another figure (Fig. 4) with additional structures for various LID scenarios. It should be mentioned here that this structure corresponds to the current case without LID and that this isn't the general model. Please try to provide a more comprehensive description of the model. (page 9, line 229)
Response: We appreciate this insightful suggestion. An additional paragraph was added at the start of Section 3.2 Hydrological model, to distinguish the different model setup methods between current condition and assumed scenarios (urbanization and LID).
- Please explain this. In table 2, D is the precipitation distribution factor. It is unclear what you mean with that. (page 9, line 242)
Response: Thank you for pointing this out. The description of D was clarified as the Precipitation distribution factor for Su in Table 2.
- Do you mean the new strucrures presented in Figure 4? (page 10, line 250)
Response: Thanks for the feedback. It was revised by adding "of the original semi-distributed model" in the manuscript. At the end of this paragraph, Figure 4 was introduced as the final schematic model figure of four LID modules.
- Considering that this model structure is empirical, some more information on how the values of all these parameters that do not have a clear physical basis were specifically assumed, on what base and by which source (for each parameter). Transferring empirical parameters values between different models is challenging. (page 10, line 251)
Response: We thank you for this valuable suggestion. As mentioned in response (4) above, most of the parameters in SUPERFLEX model have physical meanings. And it is feasible to estimate those parameters based on relevant literature, realistic field test results, and data from local government files. Estimating parameter based on prior information has been applied in many studies, especially for this type of ungauged basins (assumed LID implementation). Due to space restrict, the specific parameter estimating procedure was deleted in the manuscript. But an example of parameter estimating for bioretention cells are provided here:
There are five parameters in the bioretention model component, the precipitation distribution factor (DLID), the ratio of drainage area to construction area (AR), the maximum interception depth (Imax,B), the maximum water storage depth in soil layer (Sumax), and time lag coefficient of bioretention cells (TlagB).
Both of the precipitation distribution factor (DLID) and the ratio of construction area to drainage area (AR) depends on the concrete LID implementation plan, which two could be adjusted to fit different LID scenarios.
The maximum interception depth (Imax,B) indicated the interception capacity of bioretention cells. Li et al. (2009) found that the intercept depths for six bioretention facilities ranged from 0.6 to 4.6 mm in Maryland, U.S. A relatively good vegetation condition of the bioretention is assumed in this project with the 3.5 mm interception capacity. To adapt this parameter into the urban module, the assumed interception capacity should be multiplied by the precipitation distribution factor (DLID) and be divided by the ratio of construction area to drainage area (AR), as the final parameter, Imax,B;
As for the maximum water storage depth Sumax,B, according to "SARA LID Guidance Manual", 2 to 5 feet (0.6 to 1.2 m) soil media depth is recommended for the bioretention design. And the average soil media depth of the six typical bioretention facilities in the Maryland is referred as 0.84 m (Li et al. 2009). Considering these two recommended values, the depth of soil media layer is presumed as 0.85 m in this project, and an empirical soil porosity is chosen as 0.35 since the moderately permeable condition of local soil. Therefore, the water storage capacity for the bioretention is supposed as 300 mm. Then, the water storage capacity should multiply with the precipitation distribution factor (DLID) and be divided by the ratio of construction area to drainage area (AR), which is the estimated value of parameter Sumax,B;
Finally, according to a field test by Hunt. (2008), the peak flow of the bioretention cell could be delayed by 3 hours. Therefore, Tlag, the number of the delayed time interval, was assumed as 13 ( 3 (hours) * 2 (intervals per hour) * 2 (symmetric equation) + 1 (starting point)), to fit in the mathematical expression of delay.
- I was expecting a paragraph concerning model calibration and validation methods, data, hydrometric stations used etc. and a paragraph describing the model application, the data used, modeled events characteristics etc. and the analysis of the results. (page 12, line 265)
Response: For model calibration, validation, and data used, an additional paragraph about the strategies used in model setup was added in Section 3.2.1 Semi-distributed model setup. For model application, an additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application. For the model performance, Figure 5 was updated as individual peak events were compared during both non-flood and flood seasons.
- Please check this! Is it exactly the same? Really strange as it concerns both the calibration and the validation periods. (page 12, line 272)
Response: We thank you for your careful reading. The typo was revised from 166 to 160 mm.
- peaks? (page 12, line 276)
Response: The word was revised according to your suggestion.
- Is this also a calibration parameter? Does it represent the permeable part of the watershed? It seems that a parameter like this could be estimated by the percentage of impermeable areas or something like this with reasonable accuracy in case that this parameter has the physical meaning that I assumed above. (page 12, line 278)
Response: It is about the parameter D (precipitation distribution factor for the unsaturated zone). Your perception is correct. But in this research, all the parameters for current condition (since hydrological data are available), including D were obtained from calibration. Those parameters in urbanization and LID implementation scenarios were estimated (since no measured hydrological data).
- In figure 5 seems that the difference is mostly on peak flows, base flow seems to be similar in both cases. (page 12, line 286)
Response: The difference in base flows was indeed insignificant in this figure. This sentence was revised.
- Isn't it relatively high for urban areas? (page 12, line 288)
Response: It is about the parameter D (precipitation distribution factor for the unsaturated zone) in urban areas, which is 0.83 according to the parameter calibration results. This result can be reasonable as there are some green spaces in urban areas, and there are more in large areas of sub-urban regions.
- It would be interesting to also have a figure presenting in detail a storm event, similar to figure 7, in order to be able to see the performance of the model at detailed time scales, as the following discussion puts emphasis on specific events. (page 13, line 290)
Response: A new version of this Figure 5 could be found in the revised manuscript.
- I cannot see this in the figure. It seems like peaks timing is the same in all scenarios. Also I cannot differentiate the rural runoff from scenario A and current situation. It seems like the lower peak in scenario A is the result of lower urban peak flow for some reason. (page 16, line 337)
Response: Section 4.2 Urbanization influences on basin runoff was rewritten, and Figure 6 was updated with the zoom-in figure. For the comments about the rural and urban sub-flows, it was true that urbanization scenario A and current situation share the same rural sub-flow. It was because the larger areas of urban grey (impermeable) surfaces and less urban green surfaces in urbanization scenario A. This change of urbanization scenario A moved part of the runoff of the latter summit II forward and superimposed on the faster summit I, which cause the decrease of the latter summit II (and increase of summit I). Therefore, the maximum peak decrease, which contributed by the latter summit II.
- This is controversial. It could be the case by coincidence in very specific situations. (page 16, line 357)
Response: It is about the urbanization influences in flood season. Section 4.2 Urbanization influences on basin runoff was rewrote, and Figure 6 was updated as another peak happening in flood season was added. It can be found, all three consecutive peaks during flood season were lower in urbanization scenario A than in current conditions. This phenomenon seems controversial at first glance. But the hydrological mechanism behind this was solid, as the description in last response. Because during flood season, the basin peaks were mainly contributed from large areas of rural and urban green areas. Therefore in urbanization A, faster urban sub-flow spread the peak over a longer period of time, hence reduce peaks in the total runoff.
- A general weakness of this paper is that the results and the main findings are not discussed in the light of previous relevant studies in the results section, and most importantly in the discussion section. There are numerous studies on these topics. (page 20, line 435)
Response: Thank you for this valuable advice. Some comparisons about the performance of four LID practices between literatures and this research results were added in Section 4.3.1-4.3.4. Besides, another Section 5.3 Comparative analysis was created in the Section 5 Discussion to compare some arguments in this research refer to former studies.
- The reduction of peak reduction ability of LID in wet periods and for big storms is logical and was also discussed in previous studies. However, The coincidence of rural and LID peaks that leads to a small increase of the combined peak can be the result of a coincidence. Considering also the uncertainty of the model performance as regards LID hydrological response, I believe that more caution is required before drawing these conclusions. (page 18, line 463)
- Please see my previous comment. I believe that the obtained results cannot justify this as a general conclusion for combined urban and rural watersheds, considering also the modeling uncertainties. This could be the case in very specific cases, but not generally. (page 23, line 496)
Response to 24 and 25: We appreciate this suggestion. Although completely avoid of model uncertainty is impossible, more tests about the rainfall-runoff relationship are helpful. To further confirm LID influences in flood season, ten precipitation events with different rain intensities and durations were selected from the 600-day precipitation observation. These selected rain events were tested after 15th Sep. 2018 (flood season with the saturated subsurface soils in rural areas) in mixed LID scenario. Ten original peak runoffs corresponding to the selected rain events and ten test results were shown below.
Among ten test precipitation events, eight basin peak values were increased from 0.1 % to 3.2 % after the implementation of mixed LID practices, while only two peaks (a and c) were decreased by -0.2 % and 3.1 %. Especially for three extreme large rainstorms, the peak values increased by 0.75 % (b), 1.17% (g), and 1.82% (i). Even though the increase is small, it is to be concluded that during extremely wet conditions, the effect of implementing LID measures on peak flow reduction is negligible, if not negative in basins with combined urban and rural land use.
Based on this test, a paragraph was added in Section 4.4 LID performance in flood season. Considering the limitation of space, these two figures were not shown in the manuscript.
- As you mentioned in the previous sentence, this could be the result of a specific coincidence. By extending this conclusion, the natural watershed should present higher peaks than the urbanized one. In order to be able to justify this conclusion the model should be calibrated and validated separately for the urban and the rural parts of the watershed (this could be possible; as I can see in figure 1 there are hydrometric stations at the outlets of the urban and the rural watersheds). The LID modeling performance should be also justified, as any differences in the lag and the retention may change the obtained results. (page 23, line 500)
Response: Thanks for this comment. As you expect, the runoff data collected from the urban and rural sub-catchments were used. As mentioned in Section 3.2.1 Semi-distributed model setup, the hydrological model starts from two simple lumped pre-models, one for a rural and one for an urban sub-catchment, respectively. In this process, the data collected from two sub-catchments (rural and urban) were used to calibrate the two lumped pre-models, respectively. Then, the dominant water processes was identified from lumped models and inherited by semi-distributed models for the simulation of the whole study catchment. During the selection of semi-distributed model, the runoff characteristics of simulated urban and rural sub-flows were also compared with the runoff timeseries of two sub-catchments.
For the LID model module, the parameter uncertainty was admitted and discussed in the section 5.2 Limitations as "LID implementation scenarios presume optimistic LID implementation conditions by using favorable LID parameters, hence overlooking practical implementation, operation, and maintenance problems such as the damage of LID practices and the blockage in soil media."
Besides, since it is unrealistic to avoid model uncertainty completely, different LID implementation scenarios with various types of LID practices and different construction extents were designed to provide results as reliable as possible. According to the model results, for all the 5 LID scenarios and 2 precipitation events, 9 of 10 basin peaks were increased after the implementation of LID. This is admitted that the specific increase numbers may be fluctuated due to parameter uncertainty. But the risk of increasing basin flood can be proved. And also, the hydrological mechanism behind this was solid as the LID practices delay the urban sub-flows and cause more overlap of rural and urban peaks, which increase the basin peaks in the end.
- A key limitation is that the model was calibrated and validated only for the current conditions and not for LID. The model structure is different for the LID scenarios and new parameters are involved. Therefore, the ability of the model to effectively describe LID is questionable. I believe that a solution would be to use the same model structure in all scenarios and change some parameters corresponding to the areas covered by LID or something. As I also mentioned in the relative section, it isn't clear how the parameters values related to LID were obtained. (page 24, line 513)
Response: For the selection of LID parameters, it was answered by response 13. For the calibration of LID model module, as the response 17, since the LID implementation scenarios are designed, there is no measured data for calibration. In addition, for this kind of ungauged basins, that parameter calibration can be replaced by estimating with prior information has been proved by some studies (Sivapalan et al., 2003; Breuer et al., 2003; Hrachowitz et al., 2014; Gharari et al., 2014).
Breuer, L., Eckhardt, K., and Frede, H. G.: Plant parameter values for models in temperate climates, Ecological Modelling, 169(2-3), 237-293., 2003
Gharari, S., Hrachowitz, M., Fenicia, F., Gao, H., and Savenije, H. H.: Using expert knowledge to increase realism in environmental system models can dramatically reduce the need for calibration, Hydrology and Earth System Sciences, 18(12), 4839-4859, doi:10.5194/hess-18-4839-2014, 2014
Hrachowitz, M., Fovet, O., Ruiz, L., Euser, T., Gharari, S., Nijzink, R., Freer, J., Savenije, H. H., and Gascuel-Odoux, C.: Process consistency in models: The importance of system signatures, expert knowledge, and process complexity, Water Resources Research, 50(9), 7445-7469, doi:10.1002/2014wr015484, 2014
Sivapalan, M., Takeuchi, K., Franks, S. W., Gupta, V. K., Karambiri, H., Lakshmi, V., . . . Zehe, E. (2003). IAHS Decade on Predictions in Ungauged Basins (PUB), 2003–2012: Shaping an exciting future for the hydrological sciences. Hydrological Sciences Journal, 48(6), 857-880. doi:10.1623/hysj.48.6.857.51421
- However, the key question is what is the probability of this to happen? Is it a rare coincidence or a typical outcome for big storm events in the case of a large urban area located at the downstream part of a rural watershed? (This question was extracted from the review letter.)
Response: We thank you for this valuable question. The expression about the specific geographical condition sounds limited. However, it is quite common and may happen at every river segment downstream in the city. For example, if we selected one random river section at the urban downstream as one catchment outlet. In this catchment, the urban areas will be closer to the basin outlet than other rural areas.
Yours sincerely,
Xinxin Sui and Frans van de Ven
-
CC1: 'Reply on RC1', Xinxin Sui, 22 Mar 2021
-
RC2: 'Comment on hess-2021-32', Anonymous Referee #2, 29 Jun 2021
The hydrological performance of LID has been widely evaluated. The study focused on the basin runoff. It is interesting. And the article is well written. However, the utility and authenticity of conclusions are limited in view of this being largely a model-based simulation study wherein no observed data are used for model calibration with LID practices. The parameter values have a high impact on LID performance. Meanwhile, the model calibration (no LID practices) period is only 1 year without a warm period, which cannot represent the model performance in different climate conditions (e.g. dry and wet). A thorough analysis of the model results based on observed data is needed. In addition, the urban drainage system was not considered which has a great impact on the urban rainfall-runoff process.
1. The authors emphasized this study used a relatively simple semi-distributed model to reduce the model uncertainty caused by over-complex models. However, there is no comparison between the SUPERFLEX model and other models on model performance. Also, the NSE values of calibration and validation were 0.68 and 0.69, which are not high compared with some distributed physical models at some catchments. I suggest adding something to introduce what kind of phenomena this model can simulate and what are the limitations of its applicability. I also expect a short discussion on why this model is better than other models designed for similar purposes,e.g. SWMM.
2. The high peak flow is usually mainly contributed by surface runoff. In urban areas, due to the drainage system, the peak flow usually occurs quickly. In Figure 7, on 9th, November, the peak flow occurred quickly due to the heavy rainfall on the same day. But why the peak flow was higher on 11th, November while the rainfall was much smaller than the one on 9th, November? Is it due to parameter settings? Please explain it and show the observed values.
3. In Fig.7, why the peak flow at the rural catchment was much lower than the one at the urban catchment on 9-11, November, while it was opposite on 17th, November? Could you show the observed values hereï¼Is it caused by the model parameter setting? e.g. original soil water.
4. The authors discuss the effects of urbanization, single LID, and mix LID using different rainfall events. Please show the simulation results with the same rainfall event.
5. Since the total period for calibration and validation is not long. Please compare the simulations and observations at the outlets of rural and urban catchments. Check the model performance at these two sites.
6. The study area is a large basin with 4544 km2. There is only 1 precipitation station in the rural catchment. Please discuss its impact on this study and consider how to improve it.
7. Scenario A is 100% in current, Scenario B is 70% in current and Scenario C is 50% in current. The per capita living space for A, B, C is 0.9, 0.85, and 0. Is it an appropriate assumption?
8. Nowadays there is a tendency in writing Discussions making them longer with a greater number of comparisons with results reported in other publications. I'm not sure if I'm happy with such discussion, but please, consider extending this part.
Citation: https://doi.org/10.5194/hess-2021-32-RC2 -
AC1: 'Reply on RC2', Frans van de Ven, 20 Jul 2021
Responses to comments by referee #2
Please see the attached pdf, as a Figure is missing in the text below and the lay-out was crippled by copying
We would like to thank the referee for the careful and thorough reading of this manuscript and the thoughtful comments and constructive suggestions, which help improve this manuscript's quality. Our response follows:
Reviewer #1:
- The authors emphasized this study used a relatively simple semi-distributed model to reduce the model uncertainty caused by over-complex models. However, there is no comparison between the SUPERFLEX model and other models on model performance. Also, the NSE values of calibration and validation were 0.68 and 0.69, which are not high compared with some distributed physical models at some catchments. I suggest adding something to introduce what kind of phenomena this model can simulate and what are the limitations of its applicability. I also expect a short discussion on why this model is better than other models designed for similar purposes,e.g. SWMM.
Response: We appreciate this feedback. In the discussion, the limitation of the model is mentioned as the heterogeneity within the rural and urban areas is not represented. Besides, an additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application. As mentioned in Section 2.3, it is admitted that some complex distributed hydrological model may bring better NSE values involving more parameters and more complex model structure, but it will suffer from the drawbacks including high data-requirement for calibration, equifinality, and model uncertainty (Beven, 1993; Savenije, 2001; Hrachowitz et al., 2014). This research selected a relatively simple semi-distributed model to describe the different hydrological conditions in urban and rural areas and at the same time avoid the problem caused by complex distributed models.
Secondly, this study involves LID practices that seem suitable for the urban stormwater model, similar to SWMM with available LID modules. But different hydrological responses of urban and rural areas are also concerned. Especially, the different time lags of urban and rural peaks are closely related to the research topic. It makes the description of hydrological attenuation processes important. Therefore, an adapted version of SUPERFLEX with LID modules is used to reflect the hydrological conditions of the whole catchment, especially the difference between urban and rural areas.
Beven, K., 1993. Prophecy, reality and uncertainty in distributed hydrological modelling. Advances in water resources, 16(1), pp.41-51.
Savenije, H.H., 2001. Equifinality, a blessing in disguise?. Hydrological processes, 15(14), pp.2835-2838.
Hrachowitz, M., Fovet, O., Ruiz, L., Euser, T., Gharari, S., Nijzink, R., Freer, J., Savenije, H.H.G. and Gascuel‐Odoux, C., 2014. Process consistency in models: The importance of system signatures, expert knowledge, and process complexity. Water resources research, 50(9), pp.7445-7469.
- The high peak flow is usually mainly contributed by surface runoff. In urban areas, due to the drainage system, the peak flow usually occurs quickly. In Figure 7, on 9th, November, the peak flow occurred quickly due to the heavy rainfall on the same day. But why the peak flow was higher on 11th, November while the rainfall was much smaller than the one on 9th, November? Is it due to parameter settings? Please explain it and show the observed values.
- In Fig.7, why the peak flow at the rural catchment was much lower than the one at the urban catchment on 9-11, November, while it was opposite on 17th, November? Could you show the observed values hereï¼Is it caused by the model parameter setting? e.g. original soil water.
Response: The answer to this question is similar to the answer to question 2. It is influenced by the antecedent rains. Because the 9th, September is close to the end of the dry season, rural reservoirs are quite empty. However, the dense precipitation events in the rainy season fill the rural reservoirs gradually, making the rural areas more sensitive to precipitation events. Therefore, the rural areas start to generate large peaks from the 17th onward. Besides, Figure 5 of model performance was re-drawn in the latest version. It compares some peak events between the observation and simulation, including the peak events in September 2018.
- The authors discuss the effects of urbanization, single LID, and mix LID using different rainfall events. Please show the simulation results with the same rainfall event.
Response: Thank you for this suggestion. As mentioned in the last answer, Figure 5 of model performance was re-drawn, as it adds the comparison of some peak events between the observation and simulation.
- Since the total period for calibration and validation is not long. Please compare the simulations and observations at the outlets of rural and urban catchments. Check the model performance at these two sites.
Response: We thank you for this suggestion. However, because the urban and rural sub-catchments are not exactly the urban and rural areas of the whole catchment, the observed runoff data at the outlets of urban and rural sub-catchments could not be used directly to calibrate the semi-distributed model directly. But in order to take advantage of this information, the runoff data of urban and rural sub-catchments were used in two parts of this research.
Firstly, as mentioned in Section 3.2.1 Semi-distributed model setup, the hydrological model starts from two simple lumped pre-models, one for a rural and one for an urban sub-catchment, respectively. In this process, the data collected from two sub-catchments (rural and urban) are used to calibrate the two lumped pre-models, respectively. Secondly, during the selection of parameter set for the semi-distributed model, the runoff characteristics of simulated urban and rural sub-flows are also compared with the observed runoff time series of two sub-catchments.
- The study area is a large basin with 4544 km2. There is only 1 precipitation station in the rural catchment. Please discuss its impact on this study and consider how to improve it.
Response: We appreciate this insightful question. A map of the Thiessen polygon for the precipitation data processing is provided below. As it shows, four precipitation stations are located in the rural areas, and two are in the sub-urban areas. But more dense precipitation stations are located in the eastern than the western part of the basin. This unbalanced station distribution may cause the inaccuracy of precipitation estimation. An improvement of this precipitation dataset could be using the gridded precipitation products retrieved from radar and satellite or derived from numerical models. These sorts of the gridded dataset are suitable for those distributed models but also suffer from the uncertainty of retrieval algorithms, instrument design, and physical understanding and parameterization (Yong et al., 2015; Tan et al., 2016; Duan et al., 2016; Tang et al., 2020; Sui et al., 2020).
In this research, a robust semi-distributed model is used, which assumes a uniform precipitation pattern among the study catchment. It will balance the possible underestimation and overestimation in the whole catchment, and also more suitable for our semi-distributed model.
In general, considering the research objective and model used in this study, using the Thiessen polygons method to calculate the areal precipitation based on the precipitation gauges seems appropriate.
Yong, B., Liu, D., Gourley, J.J., Tian, Y., Huffman, G.J., Ren, L. and Hong, Y., 2015. Global view of real-time TRMM multisatellite precipitation analysis: Implications for its successor global precipitation measurement mission. Bulletin of the American Meteorological Society, 96(2), pp.283-296.
Tan, J., Petersen, W. A., Tokay, A., 2016. A novel approach to identify sources of errors in IMERG for GPM ground validation. Journal of Hydrometeorology, 17(9), 2477-2491.
Duan, Z., Liu, J., Tuo, Y., Chiogna, G., Disse, M., 2016. Evaluation of eight high spatial resolution gridded precipitation products in Adige Basin (Italy) at multiple temporal and spatial scales. Science of the Total Environment, 573, 1536-1553.
Tang, G., Clark, M. P., Papalexiou, S. M., Newman, A. J., Wood, A. W., Brunet, D., Whitfield, P. H., 2020. EMDNA: Ensemble Meteorological Dataset for North America. Earth System Science Data Discussions, 1-41.
Sui, X., Li, Z., Ma, Z., Xu, J., Zhu, S., Liu, H., 2020. Ground Validation and Error Sources Identification for GPM IMERG Product over the Southeast Coastal Regions of China. Remote Sensing, 12(24), 4154.
- Scenario A is 100% in current, Scenario B is 70% in current and Scenario C is 50% in current. The per capita living space for A, B, C is 0.9, 0.85, and 0. Is it an appropriate assumption?
Response: The first assumption is about the extent of urban infill development. Based on the urban development information provided by “City of San Antonio: Comprehensive Plan” (2010), it is known that the local government plans to terminate the unconstrained sprawl of the city but adopt the infill strategy. Due to the uncertainty of future urban development, three different extents of urban infill development are assumed, and a relatively broad range of values is given for the three scenarios from 100% to 50%.
For the second assumption about the compression of living space in the future, it is more difficult to predict, controlled by the economic policy and urban condition together. The accurate forecast of this number is difficult, and therefore the parameter uncertainty is admitted in section 5.2 Limitations. In addition, this study focuses on the hydrological influences of LID implementation rather than providing a future prediction for San Antonio City. Some degree of uncertainty on the social assumption could be acceptable in this hydrological discussion.
- Nowadays there is a tendency in writing Discussions making them longer with a greater number of comparisons with results reported in other publications. I'm not sure if I'm happy with such discussion, but please, consider extending this part.
Response: Response: Thank you for this valuable advice. Some comparisons about the performance of four LID practices between literature and this research results were added in Section 4.3.1-4.3.4. Besides, another Section 5.3 Comparative analysis was created in the Section 5 Discussion to compare some arguments in this research refer to former studies.
Yours sincerely,
Xinxin Sui and Frans van de Ven
-
AC1: 'Reply on RC2', Frans van de Ven, 20 Jul 2021
-
EC1: 'Comment on hess-2021-32', Fabrizio Fenicia, 28 Jul 2021
Dear Dr. Sui:
Thank you for submitting your manuscript entitled: “Can the implementation of Low Impact Development reduce basin runoff?” to HESS. I have received 2 reviews for your manuscript, both of which requested major revisions. The reviews highlight important points that should be addressed in the revised version.Please note that I have not found a detailed answer to the general comments of referee 1, but only to the edited PDF. If the authors believe there is duplication, they should highlight it.
A point of major concern is that although the model is defined as semidistributed, it is in fact intended to produce predictions only at the basin outlet. I would argue that a defining feature of a semidistributed model is the ability to provide predictions at internal subcatchments [Boyle et al., 2001]. The question is therefore how, by using data at a single outlet, one is able to disentangle the behavior of multiple HRUs, such as rural and urban. As the model is empirical, it requires calibration, and the disentanglement of the HRUs has to be based on data. With semidistributed models, the behavior of individual HRUs is identified by calibrating of multiple subcatchments, ideally with different HRU proportions and different responses. I think in this study, by calibrating multiple HRUs on a single catchment, there is a strong risk of equifinality.
Complementing the reviewers’ assessment, I also think that the methodology should contain more details on calibration and validation, and generally should anticipate the structure of the results sections. Currently, many analyses in the results are presented without being anticipated in the methods section, which makes the results difficult to read.
In summary, the papers requires major revision in order to address all the reviewers concerns. In submitting the revised version, please provide a point by point response to all the raised comments, therefore complementing the replies already submitted.
References
Boyle, D. P., H. V. Gupta, S. Sorooshian, V. Koren, Z. Y. Zhang, and M. Smith (2001), Toward improved streamflow forecasts: Value of semidistributed modeling, Water Resour Res, 37(11), 2749-2759.
Citation: https://doi.org/10.5194/hess-2021-32-EC1
Status: closed
-
RC1: 'Comment on hess-2021-32', Konstantinos Soulis, 25 Feb 2021
Review of the manuscript entitled: “Can the implementation of Low Impact Development reduce basin runoff?” by Xinxin Sui and Frans van de Ven
In this paper a modeling study on the potential impact of various scenarios of future urban expansion and various scenarios of low impact development interventions in a case study site in the catchment of San Antonio in Texas, USA is presented. The concept of the study is really interesting and within the scopes of HESS journal. The study also involves laborious work and an interesting case study site.
The main novelty of the paper is related to the investigation of the case of a large urban area located at the downstream part of a rural watershed. Several urban expansion scenarios and LID practices application scenarios are investigated. Accordingly, the study could potentially provide interesting information for urban planning. However, the most prominent findings of this study; i.e. that LID may result in increased peaks and that a more compacted and denser urban area may result in decreased peak discharges; are somehow controversial. The reasoning behind them is logical and I can understand that under very specific conditions this could be possible. However, the key question is what is the probability of this to happen? Is it a rare coincidence or a typical outcome for big storm events in the case of a large urban area located at the downstream part of a rural watershed? I believe that the methodology followed in this study cannot safely answer this question for two main reasons.
The first reason is related to the model used in this study. The model in the manuscript is described as semi-distributed. As I can understand by reading the manuscript, in reality it is a simple, cascading buckets, lumped model applied in the entire watershed as a whole with the same lumped parameters values for the entire watershed. Probably the model is described as semi-distributed because it involves different representations for the various parts and processes (rural, urban, LID). The model seems to perform well anyway, even if the calibration and validation procedures are not adequately presented. Accordingly, this model could be adequate for other types of studies. The problem is though that the model (as I can understand) was calibrated and validated only for the current conditions (without LID). The model applied for the LID scenarios involves more components (lags and storages) and the involved parameters values were assumed from the literature, which is really challenging and uncertain. Therefore, given that the model is lumped and empirical, its performance for the future conditions and especially for the magnitude and the timing of the flow components is questionable. Accordingly, the key finding of the study that LID may result in increased discharge peaks cannot be justified by the model results as the concurrence of the runoff peaks coming from the rural and the urban parts of the watershed due to LID could be a coincidence considering the involved uncertainties. For example, in reality parts of the urban watershed may deliver runoff to the river faster and other parts slower depending on the location of the various LID components and the resulted peak will be the mix of spatially distributed processes.
I would also like to mention at this point that while for the justification of the presented mechanisms for peaks increases, the studied events hydrographs are presented in detail in some figures, the performance of the model for the same events isn’t clear as only some general performance metrics for the entire time series are presented. Considering the emphasis that is given to the timing and magnitude of sub-event peaks in this study, I would expect a more detailed presentation of the model performance on predicting these events hydrographs. Accordingly, model calibration, validation and performance at event scale should be described in more detail and a figure (and some metrics) presenting model accuracy at event scale is required.
The second reason is that the model is applied for a short period including a small number of flood events that are analyzed to provide the above-mentioned conclusions. Accordingly, it isn’t possible to understand if these conditions are typical or limited to the investigated events.
Considering that the title of this paper is “Can the implementation of Low Impact Development reduce basin runoff?” the reader expects that the applied methodology should be able to provide a clear and justified answer.
Another important weakness of this study is related to the presentation of the methodology. There are many unclear parts and some other parts are not described at all. I believe that the model structure and the procedure followed to decide it should be presented more clearly. Most important, the identification of the parameters values for the model describing LID practices should be presented in detail and justified. In the manuscript there is just a mention of some previous studies. It should be also described in detail how the model was calibrated and validated (data used, methods, locations etc). Finally, a paragraph describing the model application, the data used, the modeled events characteristics etc. and the methodology and the steps followed for the analysis of the results is required.
Finally, a general weakness of this paper is that the results and the main findings are not discussed in the light of previous relevant studies in the results section, and most importantly in the discussion section. There are numerous previous studies on the topics presented in this study but there aren’t any citations and real discussion in the discussion section or in the results section.
Apart from the above main comments the language includes many errors and should be improved. I am not a native English speaker though and therefore I avoided providing specific suggestions on this matter. The presentation, figures etc. are generally good. There is only a problem with the definition of symbols and abbreviations that makes a little difficult to understand some parts of the methodology especially the related tables.
I am providing some additional specific comments in a commented version of the manuscript.
Based on the fundamental weaknesses of the paper that I explained above I am not convinced that this paper can be adequately improved to be suitable for publication. However, I have some reservations because some parts of the methodology presentation were confusing or missing and therefore there is a chance that I wasn’t able to accurately understand what exactly has been done in reality and to effectively evaluate the manuscript. Accordingly, as the concept and the main idea behind this study are very interesting and it involves laborious work, I am recommending a major revision to provide the authors the opportunity to respond to my main concerns and to address the key limitations of this study.
(Please also see the attached pdf file)
Best Regards
Konstantinos X. Soulis
-
CC1: 'Reply on RC1', Xinxin Sui, 22 Mar 2021
Response to comments by referee #1 (Konstantinos X. Soulis)
We would like to thank the referee, Konstantinos X. Soulis, for the careful and thorough reading of this manuscript and the thoughtful comments and constructive suggestions, which help improve this manuscript's quality. Our detailed responses could be found in the attached letter.
Yours sincerely,
Xinxin Sui and Frans van de Ven
-
AC2: 'Reply on RC1', Frans van de Ven, 20 Jul 2021
Responses to comments by referee #1 (Konstantinos X. Soulis)
Please see the attached pdf, as Figures are missing in the text below and the lay-out whas crippled while copying from teh Word document.
We would like to thank the referee, Konstantinos X. Soulis, for the careful and thorough reading of this manuscript and the thoughtful comments and constructive suggestions, which help improve this manuscript's quality. Our response follows:
Reviewer #1:
- The hydrological model description isn't sufficient neither clear enough. The model used is described as semi-distributed. As I can understand by reading the manuscript, in reality it is a simple, cascading buckets, lumped model applied in the entire watershed as a whole with the same lumped parameters values for the entire watershed. Probably the model is described as semi-distributed because it involves different representations for the various parts and processes (rural, urban, LID) but this isn't accurate. (please see the relevant general comment) (page 6, line 189)
Response: We appreciate this feedback. An additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application. Besides, some supplementary descriptions about the model used in this research were added in Section 3, Methodology, in which the model was described as a semi-distributed model with two hydrological response units (HRUs) for urban and rural surfaces, respectively. Therefore, there are two different hydrological systems for urban and rural areas separately. That’s why it’s called as a semi-distributed model.
This model is similar to the FLEX-Topo model (which was introduced in the added Section 2.3 SUPERFLEX model framework). The FLEX-topo model is also a semi-distributed model with three HRUs (plateau, hillslope, and wetland). Inspired by FLEX-topo, this model was created by separating the hydrological systems in urban and rural surfaces of one catchment.
- A very short description of the main characteristics of this model would be helpful. (page 6, line 190)
Response: We thank you for this suggestion. As mentioned before, an additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application.
- was? (page 7, line 205)
Response: We thank the reviewer for pointing this out. It was revised.
- On what basis? How? (page 7, line 214)
Response: This is about using expert knowledge to increase the efficiency of parameter calibration. The sentence in the article was rewritten as "For parameter calibration, the initial range of each parameter was given roughly based on the experience and information about catchment properties according to the physical realism of these parameters (Breuer et al., 2003; Gharari et al., 2014)."
It should be mentioned that most of the parameters in SUPERFLEX model have physical meanings, just like the popular TOPMODEL. For example, D (Precipitation distribution factor for Su) indicates the area ratio of unsaturated zone in rural areas or green (permeable) surfaces in urban areas. Sumax (Maximum unsaturated storage) could be estimated from the depth of soil layer and porosity. Therefore, a pair of relatively broad preliminary limitations were given based on the empirical value for more efficient parameter calibration and to prevent unrealistic calibration results.
Breuer, L., Eckhardt, K., and Frede, H. G.: Plant parameter values for models in temperate climates, Ecological Modelling, 169(2-3), 237-293., 2003
Gharari, S., Hrachowitz, M., Fenicia, F., Gao, H., and Savenije, H. H.: Using expert knowledge to increase realism in environmental system models can dramatically reduce the need for calibration, Hydrology and Earth System Sciences, 18(12), 4839-4859, doi:10.5194/hess-18-4839-2014, 2014
- This isn't clear. I cannot understand what has been actually done. (page 7, line 214)
Response: This is about using Monte Carlo methods for parameter calibration. After the decision of calibration limitations (as mentioned in the last response), the Monte Carlo sampling method was used to generate random pair of parameters between the limitations. In this research, the semi-distributed model was created by testing the model performance from six-bucket models to eight-buckets models. As more complex models have more parameters, the same number of samples could be seen as "unfair". Therefore, more complex models with more parameters were tested with larger numbers of parameter sample sets to ensure the calibration scale as "fair" as possible.
- Some of the symbols are not described (e.g. Su, Sf, Sh, Ss). Subscripts in the parameters symbols would be helpful (the symbols should be presented in the same way as in the mathematical expressions and parameters and components description (e.g. Italics with subscripts) (page 8, line 223)
Response: It is about some symbols in Table 2. Actually, Table 2 was introduced after Figure 3 Schematic figure of semi-distributed model structure. The bucket names and their symbolic expression (Su, Sf, Sh, Ss) were introduced under Figure 3. The name of Table 2 was revised as Parameters and model components in the semi-distributed model as shown in Figure 3, which may help to avoid confusion.
- Please make clear what was done (specific process / steps) in a simple way. At some point at the beginning, a short description of the model type and characteristics should be provided (e.g. semi-distributed, cascading bucket model, event based? etc.) (page 7, line 203)
Response: The first piece of advice is about model setup. An additional paragraph about the strategies used in model setup was added in Section 3.2.1 Semi-distributed model setup. For the second suggestion about the description of model, as mentioned before, an additional section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application.
- Also unclear. What do you mean by the term verified NSEs and what by the phrase "calculated as accuracy indexed" (page 7, line 209)
- How the "variance of the verified NSEs" is a measure of precision? (page 7, line 209)
Response to 8 and 9: These involve the specific procedure of model structure selection. As mentioned in the revised manuscript and the responses above, 1) the Monte Carlo sampling method was used to generate pairs of parameters, and 2) both the Nash-Sutcliffe Efficiency (NSE) and correlation coefficient (R2) were used to evaluate the model performance. Therefore, each pair of data would get two NSE and R2 values as calibration and validation results, respectively.
To obtain a model structure with both stable (high precision) and outstanding (high accuracy) performance, the optimal, mean, and variance of those verified NSE and R2 were examined to evaluate the accuracy and precision of model structures, as the table below.
Accuracy index
Precision index
Model Num.
The R2 of calibration
Optimal verified R2
The NSE of calibration
Optimal verified NSE
Mean of verified NSE
Variance of verified NSE
M01A
0.882
0.865
0.723
0.735
0.117
0.399
M01B
0.897
0.873
0.673
0.748
0.297
0.240
M02
0.863
0.871
0.712
0.749
0.155
0.424
M03A
0.902
0.850
0.721
0.706
-0.056
0.758
M03B
0.815
0.855
0.656
0.730
0.039
0.587
M04
0.878
0.883
0.714
0.744
0.036
0.606
Symbols: Boldface row indicates the selected model structure; Italic columns indicate the calibration results.
Since the trivial details is quite complex and of little importance to the main content of the article, the specific procedures was neglected in the manuscript to avoid unnecessary confusion of readers.
- Below, there is another figure (Fig. 4) with additional structures for various LID scenarios. It should be mentioned here that this structure corresponds to the current case without LID and that this isn't the general model. Please try to provide a more comprehensive description of the model. (page 9, line 229)
Response: We appreciate this insightful suggestion. An additional paragraph was added at the start of Section 3.2 Hydrological model, to distinguish the different model setup methods between current condition and assumed scenarios (urbanization and LID).
- Please explain this. In table 2, D is the precipitation distribution factor. It is unclear what you mean with that. (page 9, line 242)
Response: Thank you for pointing this out. The description of D was clarified as the Precipitation distribution factor for Su in Table 2.
- Do you mean the new strucrures presented in Figure 4? (page 10, line 250)
Response: Thanks for the feedback. It was revised by adding "of the original semi-distributed model" in the manuscript. At the end of this paragraph, Figure 4 was introduced as the final schematic model figure of four LID modules.
- Considering that this model structure is empirical, some more information on how the values of all these parameters that do not have a clear physical basis were specifically assumed, on what base and by which source (for each parameter). Transferring empirical parameters values between different models is challenging. (page 10, line 251)
Response: We thank you for this valuable suggestion. As mentioned in response (4) above, most of the parameters in SUPERFLEX model have physical meanings. And it is feasible to estimate those parameters based on relevant literature, realistic field test results, and data from local government files. Estimating parameter based on prior information has been applied in many studies, especially for this type of ungauged basins (assumed LID implementation). Due to space restrict, the specific parameter estimating procedure was deleted in the manuscript. But an example of parameter estimating for bioretention cells are provided here:
There are five parameters in the bioretention model component, the precipitation distribution factor (DLID), the ratio of drainage area to construction area (AR), the maximum interception depth (Imax,B), the maximum water storage depth in soil layer (Sumax), and time lag coefficient of bioretention cells (TlagB).
Both of the precipitation distribution factor (DLID) and the ratio of construction area to drainage area (AR) depends on the concrete LID implementation plan, which two could be adjusted to fit different LID scenarios.
The maximum interception depth (Imax,B) indicated the interception capacity of bioretention cells. Li et al. (2009) found that the intercept depths for six bioretention facilities ranged from 0.6 to 4.6 mm in Maryland, U.S. A relatively good vegetation condition of the bioretention is assumed in this project with the 3.5 mm interception capacity. To adapt this parameter into the urban module, the assumed interception capacity should be multiplied by the precipitation distribution factor (DLID) and be divided by the ratio of construction area to drainage area (AR), as the final parameter, Imax,B;
As for the maximum water storage depth Sumax,B, according to "SARA LID Guidance Manual", 2 to 5 feet (0.6 to 1.2 m) soil media depth is recommended for the bioretention design. And the average soil media depth of the six typical bioretention facilities in the Maryland is referred as 0.84 m (Li et al. 2009). Considering these two recommended values, the depth of soil media layer is presumed as 0.85 m in this project, and an empirical soil porosity is chosen as 0.35 since the moderately permeable condition of local soil. Therefore, the water storage capacity for the bioretention is supposed as 300 mm. Then, the water storage capacity should multiply with the precipitation distribution factor (DLID) and be divided by the ratio of construction area to drainage area (AR), which is the estimated value of parameter Sumax,B;
Finally, according to a field test by Hunt. (2008), the peak flow of the bioretention cell could be delayed by 3 hours. Therefore, Tlag, the number of the delayed time interval, was assumed as 13 ( 3 (hours) * 2 (intervals per hour) * 2 (symmetric equation) + 1 (starting point)), to fit in the mathematical expression of delay.
- I was expecting a paragraph concerning model calibration and validation methods, data, hydrometric stations used etc. and a paragraph describing the model application, the data used, modeled events characteristics etc. and the analysis of the results. (page 12, line 265)
Response: For model calibration, validation, and data used, an additional paragraph about the strategies used in model setup was added in Section 3.2.1 Semi-distributed model setup. For model application, an additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application. For the model performance, Figure 5 was updated as individual peak events were compared during both non-flood and flood seasons.
- Please check this! Is it exactly the same? Really strange as it concerns both the calibration and the validation periods. (page 12, line 272)
Response: We thank you for your careful reading. The typo was revised from 166 to 160 mm.
- peaks? (page 12, line 276)
Response: The word was revised according to your suggestion.
- Is this also a calibration parameter? Does it represent the permeable part of the watershed? It seems that a parameter like this could be estimated by the percentage of impermeable areas or something like this with reasonable accuracy in case that this parameter has the physical meaning that I assumed above. (page 12, line 278)
Response: It is about the parameter D (precipitation distribution factor for the unsaturated zone). Your perception is correct. But in this research, all the parameters for current condition (since hydrological data are available), including D were obtained from calibration. Those parameters in urbanization and LID implementation scenarios were estimated (since no measured hydrological data).
- In figure 5 seems that the difference is mostly on peak flows, base flow seems to be similar in both cases. (page 12, line 286)
Response: The difference in base flows was indeed insignificant in this figure. This sentence was revised.
- Isn't it relatively high for urban areas? (page 12, line 288)
Response: It is about the parameter D (precipitation distribution factor for the unsaturated zone) in urban areas, which is 0.83 according to the parameter calibration results. This result can be reasonable as there are some green spaces in urban areas, and there are more in large areas of sub-urban regions.
- It would be interesting to also have a figure presenting in detail a storm event, similar to figure 7, in order to be able to see the performance of the model at detailed time scales, as the following discussion puts emphasis on specific events. (page 13, line 290)
Response: A new version of this Figure 5 could be found in the revised manuscript.
- I cannot see this in the figure. It seems like peaks timing is the same in all scenarios. Also I cannot differentiate the rural runoff from scenario A and current situation. It seems like the lower peak in scenario A is the result of lower urban peak flow for some reason. (page 16, line 337)
Response: Section 4.2 Urbanization influences on basin runoff was rewritten, and Figure 6 was updated with the zoom-in figure. For the comments about the rural and urban sub-flows, it was true that urbanization scenario A and current situation share the same rural sub-flow. It was because the larger areas of urban grey (impermeable) surfaces and less urban green surfaces in urbanization scenario A. This change of urbanization scenario A moved part of the runoff of the latter summit II forward and superimposed on the faster summit I, which cause the decrease of the latter summit II (and increase of summit I). Therefore, the maximum peak decrease, which contributed by the latter summit II.
- This is controversial. It could be the case by coincidence in very specific situations. (page 16, line 357)
Response: It is about the urbanization influences in flood season. Section 4.2 Urbanization influences on basin runoff was rewrote, and Figure 6 was updated as another peak happening in flood season was added. It can be found, all three consecutive peaks during flood season were lower in urbanization scenario A than in current conditions. This phenomenon seems controversial at first glance. But the hydrological mechanism behind this was solid, as the description in last response. Because during flood season, the basin peaks were mainly contributed from large areas of rural and urban green areas. Therefore in urbanization A, faster urban sub-flow spread the peak over a longer period of time, hence reduce peaks in the total runoff.
- A general weakness of this paper is that the results and the main findings are not discussed in the light of previous relevant studies in the results section, and most importantly in the discussion section. There are numerous studies on these topics. (page 20, line 435)
Response: Thank you for this valuable advice. Some comparisons about the performance of four LID practices between literatures and this research results were added in Section 4.3.1-4.3.4. Besides, another Section 5.3 Comparative analysis was created in the Section 5 Discussion to compare some arguments in this research refer to former studies.
- The reduction of peak reduction ability of LID in wet periods and for big storms is logical and was also discussed in previous studies. However, The coincidence of rural and LID peaks that leads to a small increase of the combined peak can be the result of a coincidence. Considering also the uncertainty of the model performance as regards LID hydrological response, I believe that more caution is required before drawing these conclusions. (page 18, line 463)
- Please see my previous comment. I believe that the obtained results cannot justify this as a general conclusion for combined urban and rural watersheds, considering also the modeling uncertainties. This could be the case in very specific cases, but not generally. (page 23, line 496)
Response to 24 and 25: We appreciate this suggestion. Although completely avoid of model uncertainty is impossible, more tests about the rainfall-runoff relationship are helpful. To further confirm LID influences in flood season, ten precipitation events with different rain intensities and durations were selected from the 600-day precipitation observation. These selected rain events were tested after 15th Sep. 2018 (flood season with the saturated subsurface soils in rural areas) in mixed LID scenario. Ten original peak runoffs corresponding to the selected rain events and ten test results were shown below.
Among ten test precipitation events, eight basin peak values were increased from 0.1 % to 3.2 % after the implementation of mixed LID practices, while only two peaks (a and c) were decreased by -0.2 % and 3.1 %. Especially for three extreme large rainstorms, the peak values increased by 0.75 % (b), 1.17% (g), and 1.82% (i). Even though the increase is small, it is to be concluded that during extremely wet conditions, the effect of implementing LID measures on peak flow reduction is negligible, if not negative in basins with combined urban and rural land use.
Based on this test, a paragraph was added in Section 4.4 LID performance in flood season. Considering the limitation of space, these two figures were not shown in the manuscript.
- As you mentioned in the previous sentence, this could be the result of a specific coincidence. By extending this conclusion, the natural watershed should present higher peaks than the urbanized one. In order to be able to justify this conclusion the model should be calibrated and validated separately for the urban and the rural parts of the watershed (this could be possible; as I can see in figure 1 there are hydrometric stations at the outlets of the urban and the rural watersheds). The LID modeling performance should be also justified, as any differences in the lag and the retention may change the obtained results. (page 23, line 500)
Response: Thanks for this comment. As you expect, the runoff data collected from the urban and rural sub-catchments were used. As mentioned in Section 3.2.1 Semi-distributed model setup, the hydrological model starts from two simple lumped pre-models, one for a rural and one for an urban sub-catchment, respectively. In this process, the data collected from two sub-catchments (rural and urban) were used to calibrate the two lumped pre-models, respectively. Then, the dominant water processes was identified from lumped models and inherited by semi-distributed models for the simulation of the whole study catchment. During the selection of semi-distributed model, the runoff characteristics of simulated urban and rural sub-flows were also compared with the runoff timeseries of two sub-catchments.
For the LID model module, the parameter uncertainty was admitted and discussed in the section 5.2 Limitations as "LID implementation scenarios presume optimistic LID implementation conditions by using favorable LID parameters, hence overlooking practical implementation, operation, and maintenance problems such as the damage of LID practices and the blockage in soil media."
Besides, since it is unrealistic to avoid model uncertainty completely, different LID implementation scenarios with various types of LID practices and different construction extents were designed to provide results as reliable as possible. According to the model results, for all the 5 LID scenarios and 2 precipitation events, 9 of 10 basin peaks were increased after the implementation of LID. This is admitted that the specific increase numbers may be fluctuated due to parameter uncertainty. But the risk of increasing basin flood can be proved. And also, the hydrological mechanism behind this was solid as the LID practices delay the urban sub-flows and cause more overlap of rural and urban peaks, which increase the basin peaks in the end.
- A key limitation is that the model was calibrated and validated only for the current conditions and not for LID. The model structure is different for the LID scenarios and new parameters are involved. Therefore, the ability of the model to effectively describe LID is questionable. I believe that a solution would be to use the same model structure in all scenarios and change some parameters corresponding to the areas covered by LID or something. As I also mentioned in the relative section, it isn't clear how the parameters values related to LID were obtained. (page 24, line 513)
Response: For the selection of LID parameters, it was answered by response 13. For the calibration of LID model module, as the response 17, since the LID implementation scenarios are designed, there is no measured data for calibration. In addition, for this kind of ungauged basins, that parameter calibration can be replaced by estimating with prior information has been proved by some studies (Sivapalan et al., 2003; Breuer et al., 2003; Hrachowitz et al., 2014; Gharari et al., 2014).
Breuer, L., Eckhardt, K., and Frede, H. G.: Plant parameter values for models in temperate climates, Ecological Modelling, 169(2-3), 237-293., 2003
Gharari, S., Hrachowitz, M., Fenicia, F., Gao, H., and Savenije, H. H.: Using expert knowledge to increase realism in environmental system models can dramatically reduce the need for calibration, Hydrology and Earth System Sciences, 18(12), 4839-4859, doi:10.5194/hess-18-4839-2014, 2014
Hrachowitz, M., Fovet, O., Ruiz, L., Euser, T., Gharari, S., Nijzink, R., Freer, J., Savenije, H. H., and Gascuel-Odoux, C.: Process consistency in models: The importance of system signatures, expert knowledge, and process complexity, Water Resources Research, 50(9), 7445-7469, doi:10.1002/2014wr015484, 2014
Sivapalan, M., Takeuchi, K., Franks, S. W., Gupta, V. K., Karambiri, H., Lakshmi, V., . . . Zehe, E. (2003). IAHS Decade on Predictions in Ungauged Basins (PUB), 2003–2012: Shaping an exciting future for the hydrological sciences. Hydrological Sciences Journal, 48(6), 857-880. doi:10.1623/hysj.48.6.857.51421
- However, the key question is what is the probability of this to happen? Is it a rare coincidence or a typical outcome for big storm events in the case of a large urban area located at the downstream part of a rural watershed? (This question was extracted from the review letter.)
Response: We thank you for this valuable question. The expression about the specific geographical condition sounds limited. However, it is quite common and may happen at every river segment downstream in the city. For example, if we selected one random river section at the urban downstream as one catchment outlet. In this catchment, the urban areas will be closer to the basin outlet than other rural areas.
Yours sincerely,
Xinxin Sui and Frans van de Ven
-
CC1: 'Reply on RC1', Xinxin Sui, 22 Mar 2021
-
RC2: 'Comment on hess-2021-32', Anonymous Referee #2, 29 Jun 2021
The hydrological performance of LID has been widely evaluated. The study focused on the basin runoff. It is interesting. And the article is well written. However, the utility and authenticity of conclusions are limited in view of this being largely a model-based simulation study wherein no observed data are used for model calibration with LID practices. The parameter values have a high impact on LID performance. Meanwhile, the model calibration (no LID practices) period is only 1 year without a warm period, which cannot represent the model performance in different climate conditions (e.g. dry and wet). A thorough analysis of the model results based on observed data is needed. In addition, the urban drainage system was not considered which has a great impact on the urban rainfall-runoff process.
1. The authors emphasized this study used a relatively simple semi-distributed model to reduce the model uncertainty caused by over-complex models. However, there is no comparison between the SUPERFLEX model and other models on model performance. Also, the NSE values of calibration and validation were 0.68 and 0.69, which are not high compared with some distributed physical models at some catchments. I suggest adding something to introduce what kind of phenomena this model can simulate and what are the limitations of its applicability. I also expect a short discussion on why this model is better than other models designed for similar purposes,e.g. SWMM.
2. The high peak flow is usually mainly contributed by surface runoff. In urban areas, due to the drainage system, the peak flow usually occurs quickly. In Figure 7, on 9th, November, the peak flow occurred quickly due to the heavy rainfall on the same day. But why the peak flow was higher on 11th, November while the rainfall was much smaller than the one on 9th, November? Is it due to parameter settings? Please explain it and show the observed values.
3. In Fig.7, why the peak flow at the rural catchment was much lower than the one at the urban catchment on 9-11, November, while it was opposite on 17th, November? Could you show the observed values hereï¼Is it caused by the model parameter setting? e.g. original soil water.
4. The authors discuss the effects of urbanization, single LID, and mix LID using different rainfall events. Please show the simulation results with the same rainfall event.
5. Since the total period for calibration and validation is not long. Please compare the simulations and observations at the outlets of rural and urban catchments. Check the model performance at these two sites.
6. The study area is a large basin with 4544 km2. There is only 1 precipitation station in the rural catchment. Please discuss its impact on this study and consider how to improve it.
7. Scenario A is 100% in current, Scenario B is 70% in current and Scenario C is 50% in current. The per capita living space for A, B, C is 0.9, 0.85, and 0. Is it an appropriate assumption?
8. Nowadays there is a tendency in writing Discussions making them longer with a greater number of comparisons with results reported in other publications. I'm not sure if I'm happy with such discussion, but please, consider extending this part.
Citation: https://doi.org/10.5194/hess-2021-32-RC2 -
AC1: 'Reply on RC2', Frans van de Ven, 20 Jul 2021
Responses to comments by referee #2
Please see the attached pdf, as a Figure is missing in the text below and the lay-out was crippled by copying
We would like to thank the referee for the careful and thorough reading of this manuscript and the thoughtful comments and constructive suggestions, which help improve this manuscript's quality. Our response follows:
Reviewer #1:
- The authors emphasized this study used a relatively simple semi-distributed model to reduce the model uncertainty caused by over-complex models. However, there is no comparison between the SUPERFLEX model and other models on model performance. Also, the NSE values of calibration and validation were 0.68 and 0.69, which are not high compared with some distributed physical models at some catchments. I suggest adding something to introduce what kind of phenomena this model can simulate and what are the limitations of its applicability. I also expect a short discussion on why this model is better than other models designed for similar purposes,e.g. SWMM.
Response: We appreciate this feedback. In the discussion, the limitation of the model is mentioned as the heterogeneity within the rural and urban areas is not represented. Besides, an additional Section 2.3 SUPERFLEX model framework was created for a more detailed introduction about SUPERFLEX and its application. As mentioned in Section 2.3, it is admitted that some complex distributed hydrological model may bring better NSE values involving more parameters and more complex model structure, but it will suffer from the drawbacks including high data-requirement for calibration, equifinality, and model uncertainty (Beven, 1993; Savenije, 2001; Hrachowitz et al., 2014). This research selected a relatively simple semi-distributed model to describe the different hydrological conditions in urban and rural areas and at the same time avoid the problem caused by complex distributed models.
Secondly, this study involves LID practices that seem suitable for the urban stormwater model, similar to SWMM with available LID modules. But different hydrological responses of urban and rural areas are also concerned. Especially, the different time lags of urban and rural peaks are closely related to the research topic. It makes the description of hydrological attenuation processes important. Therefore, an adapted version of SUPERFLEX with LID modules is used to reflect the hydrological conditions of the whole catchment, especially the difference between urban and rural areas.
Beven, K., 1993. Prophecy, reality and uncertainty in distributed hydrological modelling. Advances in water resources, 16(1), pp.41-51.
Savenije, H.H., 2001. Equifinality, a blessing in disguise?. Hydrological processes, 15(14), pp.2835-2838.
Hrachowitz, M., Fovet, O., Ruiz, L., Euser, T., Gharari, S., Nijzink, R., Freer, J., Savenije, H.H.G. and Gascuel‐Odoux, C., 2014. Process consistency in models: The importance of system signatures, expert knowledge, and process complexity. Water resources research, 50(9), pp.7445-7469.
- The high peak flow is usually mainly contributed by surface runoff. In urban areas, due to the drainage system, the peak flow usually occurs quickly. In Figure 7, on 9th, November, the peak flow occurred quickly due to the heavy rainfall on the same day. But why the peak flow was higher on 11th, November while the rainfall was much smaller than the one on 9th, November? Is it due to parameter settings? Please explain it and show the observed values.
- In Fig.7, why the peak flow at the rural catchment was much lower than the one at the urban catchment on 9-11, November, while it was opposite on 17th, November? Could you show the observed values hereï¼Is it caused by the model parameter setting? e.g. original soil water.
Response: The answer to this question is similar to the answer to question 2. It is influenced by the antecedent rains. Because the 9th, September is close to the end of the dry season, rural reservoirs are quite empty. However, the dense precipitation events in the rainy season fill the rural reservoirs gradually, making the rural areas more sensitive to precipitation events. Therefore, the rural areas start to generate large peaks from the 17th onward. Besides, Figure 5 of model performance was re-drawn in the latest version. It compares some peak events between the observation and simulation, including the peak events in September 2018.
- The authors discuss the effects of urbanization, single LID, and mix LID using different rainfall events. Please show the simulation results with the same rainfall event.
Response: Thank you for this suggestion. As mentioned in the last answer, Figure 5 of model performance was re-drawn, as it adds the comparison of some peak events between the observation and simulation.
- Since the total period for calibration and validation is not long. Please compare the simulations and observations at the outlets of rural and urban catchments. Check the model performance at these two sites.
Response: We thank you for this suggestion. However, because the urban and rural sub-catchments are not exactly the urban and rural areas of the whole catchment, the observed runoff data at the outlets of urban and rural sub-catchments could not be used directly to calibrate the semi-distributed model directly. But in order to take advantage of this information, the runoff data of urban and rural sub-catchments were used in two parts of this research.
Firstly, as mentioned in Section 3.2.1 Semi-distributed model setup, the hydrological model starts from two simple lumped pre-models, one for a rural and one for an urban sub-catchment, respectively. In this process, the data collected from two sub-catchments (rural and urban) are used to calibrate the two lumped pre-models, respectively. Secondly, during the selection of parameter set for the semi-distributed model, the runoff characteristics of simulated urban and rural sub-flows are also compared with the observed runoff time series of two sub-catchments.
- The study area is a large basin with 4544 km2. There is only 1 precipitation station in the rural catchment. Please discuss its impact on this study and consider how to improve it.
Response: We appreciate this insightful question. A map of the Thiessen polygon for the precipitation data processing is provided below. As it shows, four precipitation stations are located in the rural areas, and two are in the sub-urban areas. But more dense precipitation stations are located in the eastern than the western part of the basin. This unbalanced station distribution may cause the inaccuracy of precipitation estimation. An improvement of this precipitation dataset could be using the gridded precipitation products retrieved from radar and satellite or derived from numerical models. These sorts of the gridded dataset are suitable for those distributed models but also suffer from the uncertainty of retrieval algorithms, instrument design, and physical understanding and parameterization (Yong et al., 2015; Tan et al., 2016; Duan et al., 2016; Tang et al., 2020; Sui et al., 2020).
In this research, a robust semi-distributed model is used, which assumes a uniform precipitation pattern among the study catchment. It will balance the possible underestimation and overestimation in the whole catchment, and also more suitable for our semi-distributed model.
In general, considering the research objective and model used in this study, using the Thiessen polygons method to calculate the areal precipitation based on the precipitation gauges seems appropriate.
Yong, B., Liu, D., Gourley, J.J., Tian, Y., Huffman, G.J., Ren, L. and Hong, Y., 2015. Global view of real-time TRMM multisatellite precipitation analysis: Implications for its successor global precipitation measurement mission. Bulletin of the American Meteorological Society, 96(2), pp.283-296.
Tan, J., Petersen, W. A., Tokay, A., 2016. A novel approach to identify sources of errors in IMERG for GPM ground validation. Journal of Hydrometeorology, 17(9), 2477-2491.
Duan, Z., Liu, J., Tuo, Y., Chiogna, G., Disse, M., 2016. Evaluation of eight high spatial resolution gridded precipitation products in Adige Basin (Italy) at multiple temporal and spatial scales. Science of the Total Environment, 573, 1536-1553.
Tang, G., Clark, M. P., Papalexiou, S. M., Newman, A. J., Wood, A. W., Brunet, D., Whitfield, P. H., 2020. EMDNA: Ensemble Meteorological Dataset for North America. Earth System Science Data Discussions, 1-41.
Sui, X., Li, Z., Ma, Z., Xu, J., Zhu, S., Liu, H., 2020. Ground Validation and Error Sources Identification for GPM IMERG Product over the Southeast Coastal Regions of China. Remote Sensing, 12(24), 4154.
- Scenario A is 100% in current, Scenario B is 70% in current and Scenario C is 50% in current. The per capita living space for A, B, C is 0.9, 0.85, and 0. Is it an appropriate assumption?
Response: The first assumption is about the extent of urban infill development. Based on the urban development information provided by “City of San Antonio: Comprehensive Plan” (2010), it is known that the local government plans to terminate the unconstrained sprawl of the city but adopt the infill strategy. Due to the uncertainty of future urban development, three different extents of urban infill development are assumed, and a relatively broad range of values is given for the three scenarios from 100% to 50%.
For the second assumption about the compression of living space in the future, it is more difficult to predict, controlled by the economic policy and urban condition together. The accurate forecast of this number is difficult, and therefore the parameter uncertainty is admitted in section 5.2 Limitations. In addition, this study focuses on the hydrological influences of LID implementation rather than providing a future prediction for San Antonio City. Some degree of uncertainty on the social assumption could be acceptable in this hydrological discussion.
- Nowadays there is a tendency in writing Discussions making them longer with a greater number of comparisons with results reported in other publications. I'm not sure if I'm happy with such discussion, but please, consider extending this part.
Response: Response: Thank you for this valuable advice. Some comparisons about the performance of four LID practices between literature and this research results were added in Section 4.3.1-4.3.4. Besides, another Section 5.3 Comparative analysis was created in the Section 5 Discussion to compare some arguments in this research refer to former studies.
Yours sincerely,
Xinxin Sui and Frans van de Ven
-
AC1: 'Reply on RC2', Frans van de Ven, 20 Jul 2021
-
EC1: 'Comment on hess-2021-32', Fabrizio Fenicia, 28 Jul 2021
Dear Dr. Sui:
Thank you for submitting your manuscript entitled: “Can the implementation of Low Impact Development reduce basin runoff?” to HESS. I have received 2 reviews for your manuscript, both of which requested major revisions. The reviews highlight important points that should be addressed in the revised version.Please note that I have not found a detailed answer to the general comments of referee 1, but only to the edited PDF. If the authors believe there is duplication, they should highlight it.
A point of major concern is that although the model is defined as semidistributed, it is in fact intended to produce predictions only at the basin outlet. I would argue that a defining feature of a semidistributed model is the ability to provide predictions at internal subcatchments [Boyle et al., 2001]. The question is therefore how, by using data at a single outlet, one is able to disentangle the behavior of multiple HRUs, such as rural and urban. As the model is empirical, it requires calibration, and the disentanglement of the HRUs has to be based on data. With semidistributed models, the behavior of individual HRUs is identified by calibrating of multiple subcatchments, ideally with different HRU proportions and different responses. I think in this study, by calibrating multiple HRUs on a single catchment, there is a strong risk of equifinality.
Complementing the reviewers’ assessment, I also think that the methodology should contain more details on calibration and validation, and generally should anticipate the structure of the results sections. Currently, many analyses in the results are presented without being anticipated in the methods section, which makes the results difficult to read.
In summary, the papers requires major revision in order to address all the reviewers concerns. In submitting the revised version, please provide a point by point response to all the raised comments, therefore complementing the replies already submitted.
References
Boyle, D. P., H. V. Gupta, S. Sorooshian, V. Koren, Z. Y. Zhang, and M. Smith (2001), Toward improved streamflow forecasts: Value of semidistributed modeling, Water Resour Res, 37(11), 2749-2759.
Citation: https://doi.org/10.5194/hess-2021-32-EC1
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 955 | 466 | 42 | 1,463 | 41 | 33 |
- HTML: 955
- PDF: 466
- XML: 42
- Total: 1,463
- BibTeX: 41
- EndNote: 33
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1