Application of a Surrogate Model for a Groundwater Numerical Simulation Model for Determination of the Annual Control Index of the Groundwater Table in China

Round 1

Reviewer 1 Report

Dear authors,

After carefully reading of your manuscript I selected the following main comments:

Journal template was not used.

Citation style and reference list are not according instructions.

Title is not clear.

Define using of the word “abstraction” in your paper, it could be confusing for many readers. Also define “water table index”.

p. 4. “the uncertainty of hydrological factors will lead to the uncertainty of water demand”. It is not obvious statement and if you think so, please, better elaborate how needs for water supply are correlated with hydrology (and hydrogeology + rainfall).

“Hydrogeological uncertainty” – in this subchapter you did not describe hydrogeological variables at all.

p. 5. Eq. 4, what are L and T? Length and time? Units? We still do not know almost anything about your numerical model, inputs, outputs, size etc.

p. 7. “Latin Hypercube Sampling Technique (LHS)”. Explain the method and application in sampling for your model. What is “reasonably considered”?

Eq. 7. Still do not understand variables. Where is legend?

Eq. 8. What is deterministic drift?

Eq. 9. Legend?

Fig. 1 – quality is too low, text non-readable. Where is top map located on lower-left? Where is lower-right on the top? Where are other countries and islands on the global map (only Taiwan is there)?

p. 9. Geological section?

Fig. 2. Quality is too low. Where is Fig. 2. Located on Fig. 1.?

p. 12. “The calculated groundwater flow field matched the actual flow field reasonably” – such statement must be numerical, not textual.

p. 13. “A KRIGING model program was developed in MATLAB based on the principle of Kriging method” – did you develop algorithm alone? For which Kriging technique? No one map is shown.

I still do not understand what surrogate model included. How to check algorithm with 10 group of data? The set was enough large? Did you use 10 data or 10 data groups for validation?

p. 16. Table caption and content do not match.

p. 17. “in Minqin Basin showed that the technical method established in this study”. It seems to me that you applied analytical method, not technical.

Kind regards,

Reviewer

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The introduction is very well approached by providing a clear context and then defining the problem that needs to be addressed. It also includes a literature review that serves to justify the need for this research and how it contributes to filling a existing gap in groundwater modelling.

 

The authors should avoid using too informal expressions, such as "Generally speaking", "&" or "can't".

 

The methodology is well explained, with clear indications about how the formulas evolve due to simplifications and/or assumptions.

 

Please, provide the full meaning of acronyms the first time they are used (i.e. GMS = Groundwater Modeling System)

 

The level of English needs to be revised throughout the whole manuscript. There are some grammar and orthographic mistakes that should be amended (e.g. "Using the Latin Hypercube Sampling Technique (LHS) is used to sample q within the feasible pumping internal").

 

Why did you choose Kriging? Please, justify. Also, what type of Kriging? Ordinary Kriging?

 

The authors must provide more details about the quality of their data (e.g. resolution, scale, etc.). The accuracy of the results completely depend on the quality of the data used.

 

Should I assume that CC is correlation coefficient? If positive, which type of CC? Pearson, Spearman, Phi...? In any case, the p-value of such correlation should be provided. Otherwise, the result is meaningless.

 

Like the abstract, the conclusions are not well approached. They contain a lot of details and figures that are not suitable for this section. The conclusions should deal with the main general findings of the study, as well as their implications for its research field and, by extension, the society. The limitations of the investigation should be highlighted too and linked with some potential future lines of research.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear authors,

In this 2nd round your manuscript reached the quality to be reviewed. Some of mine suggestion had been applied, but I found some that could be useful to use in your work. They are the following:

Figure 1 – left. Too many and too small text that cannot be read. Taiwan is politically identity that could not be shown simple as mainland China province (coloured).

 

Line 285 – „fugure.2“?

 

Figure 2 – Superscript on Quaternary are non-readable. Also, you need to explain Q symbols in legend. Geological section of Fig. 1 must be cleary connected with Fig. 2 (e.g., A-A'). Fig. 2 unis on Y, scale on X??? Bore hole=borehole, Shiyang river=Shiyang River, Caiqi=town?

 

Line 309 – „the mining volume“ – the groundwater is not mined, but recovered or exploited or pumped or drained.

 

Line 348 - Hongyashan reservoir = Hongyashan Reservoir, Tengger desert = Tengger Desert (same rule as Qingtu Lake).

 

Lines 368/9 - We chose 5% of the total amount for qi and b 18% of the total amount for qi for each sub management area in these two irrigation areas.

 

Explain why 5 %?

 

Line 372 - The sampling dimension (n) is set as 100. Do you mean on number of samples?

 

Line 378 - fig 3 = Figure 3.

 

Fig. 4 – the legend is non-readable.

 

Lines 423-432 – ordinary Kriging=Ordinary Kriging. The Kriged mapi s still not given in paper, but it is obligatory. Eq. 8 is not the best choice for Kriging example. Please use simple matrix equation for the Ordinary Kriging. Regarding input dana what are „q“ and especially „h“??? What is difference among three mention scenarios? And the Kriging is not process that is trained (like neural networks), but it is process of mapping (deterministic, not stochastic). Did you perform 3D Kriging (block Kriging), if you used „h“?

 

Table 6 (as example) – I do not understand why some blanks cells are part of table on the right side? Some information would be added (N/A or similar).

 

Line 501 - A surrogate model based on the Kriging method = A surrogate model based on the Kriging maps… (surrogate model or models???).

 

Kind regards,

Reviewer