Impacts of Climate Change and Variability on Precipitation and Maximum Flows in Devil’s Creek, Tacna, Peru

Round 1

Reviewer 1 Report

Dear Editor,

After reading the manuscript, I found that the research presented can be of interest to those interested in assessing the impact of climate change on the hydrological response of ephemeral streams.
The manuscript is correctly written, but some spelling errors were found in tables and figures.
The correction of precipitation data enlarges the manuscript and is tiring to read. Consider moving into the supplementary information.
Some methods are poorly described.
The quality of the figures is also very poor; I cannot visualize these data. In addition, some inconsistencies of data represented in the figures make it difficult to rely on that data.
The temperature analysis is poorly justified and unnecessarily enlarges the manuscript; please remove it.
The same data are represented in both tables and figures.
I regret to inform you that the overall quality of the manuscript is poor; I recommend rewriting before a new submission.

Please, read the notes in the commented manuscript.

Comments for author File: Comments.pdf

Author Response

Dear Editor,

After reading the manuscript, I found that the research presented can be of interest to those interested in assessing the impact of climate change on the hydrological response of ephemeral streams. The manuscript is correctly written, but some spelling errors were found in tables and figures.

• Line 73: Please, explain in a more scientific, precise, way.

Answer:

We correct the statement to:

The infrastructure design is based on the IDF precipitation curves with the assumption of stationarity, which means that the statistical properties of future events will be similar to those of the past [9,10].

• Some methods are poorly described.

Answer

We have incorporated further explanation of the methods used

• The quality of the figures is also very poor; I cannot visualize these data. In addition, some inconsistencies of data represented in the figures make it difficult to rely on that data.

Answer:

The following figures were corrected according to indication:

• Figure 2. Total daily precipitation - Jorge Basadre Station.
• Figure 3. Frequency analysis of the partial series of total daily precipitation at JORGE BASADRE raingauge (From 1993 to 2020)
• Figure 4. Comparison double-mass between the precipitation data of the PISCO product in JORGE BASADRE raingauge located versus the rainfall of the JORGE BASADRE raingauge, during 7780 days (JANUARY 01/1993 – APRIL 21/2014)
• Figure 5. Comparation double-mass between the precipitation data of the PISCO product in JORGE BASADRE raingauge located versus the precipitation data of the PISCO product in Devil´s Creek located, during 7780 days (JANUARY 01/1993 – APRIL 21/2014)
• Figure 6. Estimated hourly rainfall for the Devil’s Creek. A (Event on JANUARY 23-24/2020), B (Event on FEBRUARY 21/2020), and C (Event on DECEMBER 27/2020)
• Figure 8. Comparation between (a) Daily rainfall in JORGE BASADRE raingauge and Daily precipitation from satellites products: (b) IMERG EARLY, (c) IMERG FINAL, and (d) CHIRPS.
• Was modify Figure 9. Correlations were made between the total daily precipitation data between the JORGE BASADRE raingauge and the virtual raingauges, (EV01 and EV02) using the products: (a,b) PISCO, (c, d) IMERGE EARLY, (e, f) IMERGE FINAL, and (g, h) CHIRPS.
• Figure 10. Location of soil sampling points.
• Figure 32. Hydrograph generated by sub-basins
• Figure 33. Hydrograph of entry to the Paso Camiara informal embankment.
• Figure 34. Hydrograph of water height in the dam and flow discharged due to the collapse of the Paso Camiara dam
• Was include Figure 35. Gestion newspaper reports: Debris flow in Tacna left three people dead. February 22, 2020
• Figure 37 (before Figure 36). Flood hydrographs for RCP4.5 and RCP8.5 scenarios and for 2021 - 2050 and 2051 – 2080 terms
• The temperature analysis is poorly justified and unnecessarily enlarges the manuscript; please remove it.

Answer:

Values and temperature changes of the Ensemble of the 15 climate models were used in the RS Minerve hydrological model, for this reason it is not appropriate to take out the temperature from the article.

• The same data are represented in both tables and figures.
  I regret to inform you that the overall quality of the manuscript is poor; I recommend rewriting before a new submission.

Answer:

This has been corrected in the text

• Line 422: These values do not correspond with those found in the Figure 12.

Answer:

Due to a scale problem the values were not shown, however, in this version the Figures were updated.

• Figure 13: Outliers do not match the maximum values.

Answer:

Figure 13 was updated

• Line 446: The boxplot explanation is not correct. "minimum" is the Q1 (25% percentile) minus 1.5 times the interquartile range. "maximum" is Q1, and "maximum is Q3 (75% pertentile)  plus 1.5 times the interquartile range (Q3-Q1).

Answer:

In this case, the minimum value corresponds to the lower extended line, and the maximum value to the upper extended line. Therefore, the values in the box represent 25%, 50% and 75%, respectively.

Please, read the notes in the commented manuscript.

Reviewer 2 Report

This is an interesting article presenting projections of precipitation and temperature changes in the Devil’s Creek, located  in the Tacna region in Peru, according to two selected emission scenarios RCP4.5 and RCP8.5  and in two time slices, for the period 2021-2050 and 2051-2080. However, it requires significant corrections before being published.The data is poorly described. Their temporal resolution has not been specified, are they daily or monthly totals? What does the abbreviation SENAMHI mean? The text repeatedly mentions the annual precipitation maximum. Is it about the maximum daily total rainfall in a year?table 1: % empty, is this a percentage of gaps?Twenty why the value was entered in words, whether it was 20% or 20 of the value?

In the description of the data taken from the models, it was stated from which regional models they came from, but nothing is known about the global models. They also affect the results, and it is particularly important to know whether it was one global model or many different ones.

Are the cumulative distribution functions in the formula in line 161 empirical distribution functions calculated from observations or theoretical (distribution functions of theoretical distributions fitted to the observations)?

Figure 2: The description of the timeline is illegible. Are these the daily precipitation totals for 28 years of observation?figure 4 and 5 symbol descriptions in the drawing should agree with the caption under the drawing

The Nash Sutcliffe efficiency index is not explained anywhere (line 233).

sentence: Likewise, it was found  that the difference between the total daily precipitation discharged from the PISCO prod-uct for the coordinates of the Jorge Basadre station and the total daily precipitation dis-charged from the Jorge Basadre station is -31%. (lines 234-236) 31% of the first one or of the second one?

What does it mean differs by -38% (1.31 * 1.06) (line 251). Table 3: Is PISCO a satalite product? The abbrevations POT and USCB are nowhere explained.Is the precipitation in the Jorge Basadre station, the same as the precipitation in the Jorge Basadre station area (or zone). If not, the differenvce should be clearly explained. (lines 312-319). Figure 9, table 4 and their description: Pearson linear correlation only makes sense if both of the variables are normally distributed, which I don't think is the case here. What does the term “prom” mean? (table 4)Figure 11. All fluxes should be explained in the figure description. Does ETP realy mean potential evapotranspiration? The direction of the arrow indicates that this is the ETR, and the ETP is some water supply, what?On figures 12, 14, 16 and 18 blue dashed line is described as a red line. Why the ensemble value  is sometimes equal to the highest value among models? On figures 13, 15, 17 and 19 The linear extensions represent the highest and lowest values; the upper, middle and lower 445 limits of the box represent the percentiles of 75%, 50%, and 25% respectively. Why some values are higher than the highest value? How the outliers are defined? 

If I understand correctly, the model was calibrated on the day the dam broke. Then a large amount of water stored above the dam entered the system below the dam. Is this a good case to calibrate? There is no justification as to why this particular day was chosen.

 

In case of temperature changes, Figures 22 and 23 show a clear increase in the annual maximum in both emission scenarios.Figures 24 and 25 show that both the minimum and maximum temperature predicted in both scenarios for both forecast periods is a few degrees higher than the cumulative temperature for the historical period 1981-2005. In this context, the negative anomalies of the minimum temperature shown in Figure 26 and the information about the projected minimum temperature drop in winter (June, July, August) are completely incomprehensible and require a thorough explanation. 

Author Response

This is an interesting article presenting projections of precipitation and temperature changes in the Devil’s Creek, located in the Tacna region in Peru, according to two selected emission scenarios RCP4.5 and RCP8.5  and in two time slices, for the period 2021-2050 and 2051-2080. However, it requires significant corrections before being published.

• The data is poorly described.

Answer:

What data, what page o line?

• Their temporal resolution has not been specified, are they daily or monthly totals? What does the abbreviation SENAMHI mean?

Answer:

The temporal resolution has been specified: “daily”. On the other hand, we insert the meaning of SENAMHI (The Peruvian National Meteorological and Hydrological Service).

• The text repeatedly mentions the annual precipitation maximum. Is it about the maximum daily total rainfall in a year?

Answer:

Yes. It is the annual maximum daily precipitation

• table 1: % empty, is this a percentage of gaps? Twenty why the value was entered in words, whether it was 20% or 20 of the value?

Answer:

Yes. It is the percentage of gaps. It was considered in the Table. Moreover, it the word twenty was changed to its numerical value of 20.

• In the description of the data taken from the models, it was stated from which regional models they came from, but nothing is known about the global models. They also affect the results, and it is particularly important to know whether it was one global model or many different ones.

Answer:

This research uses downscaled climate scenarios for the globe from the NASA NEX-GDDP dataset which are derived from the General Circulation Model (GCM) runs conducted under the Coupled Model Intercomparison Project Phase 5 (CMIP5). This was clarified in the text.

• Are the cumulative distribution functions in the formula in line 161 empirical distribution functions calculated from observations or theoretical (distribution functions of theoretical distributions fitted to the observations)?

Answer:

Yes, it is the empirical cumulative distribution function of observed and modelled time series. The word “empirical” was added to the text.

• Figure 2: The description of the timeline is illegible. Are these the daily precipitation totals for 28 years of observation? figure 4 and 5 symbol descriptions in the drawing should agree with the caption under the drawing

Answer:

The Figure 2 was corrected. Yes, The JORGE BASADRE raingauge has a continuous record of 28 years (1993 to 2020)

The Figure 4 and Figure 5. For comparison we use double-mass plot.

• The Nash Sutcliffe efficiency index is not explained anywhere (line 233).

Answer:

We explained of The Nash Sutcliffe efficiency index. The equation and the efficiencies reached according the values.

• Sentence: Likewise, it was found that the difference between the total daily precipitation discharged from the PISCO product for the coordinates of the Jorge Basadre station and the total daily precipitation dis-charged from the Jorge Basadre station is -31%. (lines 234-236) 31% of the first one or of the second one?

Answer:

Of the second one.

• What does it mean differs by -38% (1.31 * 1.06) (line 251).

Answer:

We explained the calculation of the differ by -38%

• Table 3: Is PISCO a satelite product? The abbrevations POT and USCB are nowhere explained.

Answer:

No, PISCO used satellite products. PISCO is a data base product ‘Peruvian Interpolated data of the SENAMHI Climatological and hydrological Observations’

The word POT was a spelling mistake. The right word is NASA. UCSB mean University of California Santa Barbara.

• Is the precipitation in the Jorge Basadre station, the same as the precipitation in the Jorge Basadre station area (or zone). If not, the difference should be clearly explained. (lines 312-319).

Answer:

The precipitation in the Jorge Basadre raingauge is greater than the precipitation at the lower and upper part of the Devil’s Creek. Satellite products were used to demonstrate this.

• Figure 9, table 4 and their description: Pearson linear correlation only makes sense if both of the variables are normally distributed, which I don't think is the case here. What does the term “prom” mean? (table 4)

Answer:

Is correct. The Pearson linear correlation makes sense if both of the variables are normally distributed. The only reason to use the linear correlation was establishing one indicator of spatial variation. The spatial correlation was carried out between the three established raingauges: JORGE BASADRE raingauge, Virtual raingauge 01 (lower part of the Creek), and Virtual raingauge 2 (upper part of the Creek).

The word corrected is Mean.

• Figure All fluxes should be explained in the figure description. Does ETP realy mean potential evapotranspiration? The direction of the arrow indicates that this is the ETR, and the ETP is some water supply, what?

Answer:

All fluxes in the Figure 11, were explained. ETP means potential evapotranspiration and the arrow direction  ETP was corrected.

On figures 12, 14, 16 and 18 blue dashed line is described as a red line. Why the ensemble value  is sometimes equal to the highest value among models?

Answer:

It was corrected and changed to Blue Line as correspond. There was a scale issue, however, the Figures have been updated, now.

• On figures 13, 15, 17 and 19 The linear extensions represent the highest and lowest values; the upper, middle and lower limits of the box represent the percentiles of 75%, 50%, and 25% respectively. Why some values are higher than the highest value? How the outliers are defined?

Answer:

An outlier is a single observation "far away" from the rest of the data that can lead to unrealistic interpretations of the hydrological system dynamics, especially when evaluating extreme events. In our study, the whiskers are not more than 1.5 times longer than the box, consequently each boxplot can contain one or more possible identified outliers. This is very useful, especially in precipitation data under climate change scenarios that have a high degree of uncertainty.

• If I understand correctly, the model was calibrated on the day the dam broke. Then a large amount of water stored above the dam entered the system below the dam. Is this a good case to calibrate? There is no justification as to why this particular day was chosen.

Answer:

Is correct, the model was calibrated on the day the dam broke. The reason is that the Devil’s Creek does not have a gauging station. Two criterion were use: (1) The maximum flow of 10. 72 m³/s was contrasted with the water footprints left by the event in the riverbed of the Devil’s Creek, (2) The maximum flow of 2550.8 m³/s, produced the debris flow and floods that left loss of three human lives and great economic losses in the Tacna city.

• In case of temperature changes, Figures 22 and 23 show a clear increase in the annual maximum in both emission scenarios. Figures 24 and 25 show that both the minimum and maximum temperature predicted in both scenarios for both forecast periods is a few degrees higher than the cumulative temperature for the historical period 1981-2005. In this context, the negative anomalies of the minimum temperature shown in Figure 26 and the information about the projected minimum temperature drop in winter (June, July, August) are completely incomprehensible and require a thorough explanation.

Answer:

Figures 24 and 25 show the monthly average for each period, and Figures 26 and 27 show the ranges of anomalies using boxtplot of all monthly data. For example, 360 months for the period 2021-2050, in which it is possible to detect negative anomalies, but on the average a positive value is obtained, as evidenced in the June boxtplot in Figure 26 for the tasmin RCP4.5, period 2021-2050, where quartile 75 indicates positive anomalies of approximately 2 ^oC. However, to be consistent with Figures 24 and 25, the anomaly Figures have been changed taking into account the monthly average chanhes for the period only.