Habitat Analysis of Endangered Korean Long-Tailed Goral (Naemorhedus caudatus raddeanus) with Weather Forecasting Model

Sec.1

A deep review is not acutely described in the introduction section. Please provide more

up-to-date references if available in the literature.

Many references (until 2019) are added and the introduction was significantly updated.

In some reference citations I would suggest to include the author name. (Line 39: For example, Thomas et al. [2] reported that…)

Corrected the reference starting with reference number with name and reference number. (Thomas et al. [2] and Thuiller [10])

Sec.3

L80: Please introduce the Appendix I at the end of the manuscript.

Instead of adding appendix, we added the website link for the Appendix I.

The Korean long-tailed goral population within the park (100 individuals) is taken from a reference of 2002. Is there any more updated data of the current situation?

The data was updated to the most recent census data of 2015.

L86: ‘several studies have examined natural factors affecting the goral habitat’. Which ones? Please provide some references of these studies.

They are deleted during reorganizing and updating the introduction section.

Sec.4

The fecal samples with which the study was carried out are from 13-14 years ago. Is not there the possibility of having a more updated database? How has the goral population been influenced in these last years?

Unfortunately, we do not have any updated dataset on hand. Instead, we found a reference about the influence of the population from a journal article and it is included. The population is mainly influenced by human. Under appropriate control such as poaching regulations and restoration projects, the goral seems to be increased for a decade.

Please provide some information about the three climate scenarios used (B1, A1B and A2) to assess the future climate for the goral habitat. What are their differences? Why have been chosen these 3 scenarios in particular?

The modeling and scenarios were all updated with the latest WRF v.3.9 with the latest climate scenario (RCP8.5) in AR5 in 2014.

Sec.5

Wontong weather station (Line 177) and Inje weather station (Line 204) is the same?

It was intended to be the Wontong AWS station in Inje. However, during simulation for update, we found that there are missing data for four days in the data from Wontong. So, we changed the station to another station called Inje, which is not AWS.

L207: ‘The bias was much greater than the result of the sensitivity test but it is still acceptable’ Why or how was that conclusion reached?

We deleted the sentence because we did not compared it to the wontong data but used a reference for the general accuracy of the algorithm we used.

The area of the predicted future mean temperatures presented in Figure 11 corresponds to a resolution of 1 km?

Yes, it corresponds to the resolution of 1 km. Added more comments on that and now changed to Figure 7.

‘Suitable habitat (colored dark purple) declined and almost disappeared in the 2050s (for winter) or 2070s (for summer)’ What characteristics does a suitable habitat have? Why is it limited to the dark purple area?

It is delineated based on the current temperatures (18.7°C for summer and -8.3°C for winter) of the goral habitat including the fecal sample locations so the blue area includes the fecal locations.

Sec.6

L238: ‘Based on the statistical results of our fecal samples and other studies, the Korean long-tailed goral in Seoraksan National Park tend to prefer middle elevation (501-900m) habitats (Table 1).’

Table 1 is not presented in the manuscript. Please cite some of the referred studies and

compare the results of the present study with them.

A typical suitable habitat for goral is peaky rock mountains with the elevation of 600-700m and we added the comment in the Section 5 with a reference. The information is presented at the introduction section with a reference.

Minor1

For multiple citations, please use the style [1-4] instead of [1,2,3,4].

Corrected the format.

Minor2

Figure 2 is described before Figure 1.

Corrected the order of figures.

Minor3

Please cite the Figures within the manuscript as ‘Figure 1’ instead of ‘Fig. 1’.

Corrected the format.

Minor4

Line 136: Please delete the bracket at the end of the sentence.

Corrected.

Minor5

Please complete the ‘Author Contributions’ and ‘Conflicts of Interest’ sections.

Deleted. Those are for communication with authors.

Line

Reviewer’s comment

Author’s answer

General

In general, this study uses two numerical models to simulate main meteorological elements (called climate factors in this manuscript) in the area of Seoraksan National Park. For the WRF model, two 14-day simulations are made for winter 2005 and summer 2006, and the simulated temperature evolution in summer 2006 is shown to be similar to the observation at the Wontong weather station. Then the evolutions of temperature, relative humidity, wind direction, and wind speed of both simulations at different altitudes are also shown, but I wonder about their purpose for no related discussions.

These factors simply show the model reliability for our application. The model showed reliable sensitivity and bias with the factor estimations.

Table 1 is missing.

Table 1 is inserted.

All the figure captions are oversimplified.

More comments are added.

The purpose of the entire WRF experiment seems to be not clear. Sections 4.2 and 4.3 are titled as “Brief description of WRF climate model” and “Climate simulation for current habitat climate analysis”, respectively.

We need to briefly explain what WRF and its characteristics are to the reader. Section 4.2 covers this. Section 4.3 explains the actual steps we made for retrieving climatic parameters in depth but it looks complicated to be followed. We modified Section 4.2 and 4.3 in a more understandable way.  

In fact, WRF is intrinsically a regional weather prediction model rather than a climate model,

We changed the terminology.

I think the 14-day simulations are too short to represent climate simulations and analyses.

In remodeling, we used 30 days. 30 days are enough to understand the current weather condition of the habitat, for example, weather forecasting does not provide the weather for a long time line a few months or years.

The similarity between the simulated and observed temperature evolutions does not mean much. Therefore, it would be better to have further discussion for the purpose of the WRF experiment in this study.

The observed data simply used for the sensitivity of WRF after remodeling so the similarity comments were deleted. Instead, a reference was provided for that.

What kind of the statistical downscaling method is used?

We change the method to the dynamic downscaling method of Advance Research WRF algorithm in remodeling. Indicated in Section 3.3.

When comparing the difference between the present and future climate, the method of downscaling should be consistent. In this study, scenarios of IPCC AR4 is used. Nowadays, the IPCC Sixth Assessment Report (IPCC AR6) is going to be released. The latest scenarios for projection climate would be the RCP series made by IPCC AR5.

The design of scenarios between AR4 and AR5 is quite different. On the other hand, the version 2.2 (released in 2006) of WRF is way too old for studies owing to huge dynamical and physical modifications during the past 13 years. Based on these points, please elaborate more about the reason of using AR4 scenarios and the necessity of using early version of WRF model.

We used the stable version of WRF (V3.9) and RCP 8.5 in AR5. All of our results are renewed based on the results.

L16

As aforementioned, WRF is not a climate model. The predictability of WRF dissipates rapidly after weeks of integration.

Modified to “future weather conditions.”

L19-20

Add scenario names B1, A1B, and A2, respectively.

Changed the scenarios to RCP8.5 in AR5.

L35

… adapt “to”, and…

Modified to “adapt to”

L36

 … might “be” a…

Modified to “might be a shift”

L46

 … there “are” insufficient…

Modified to “are”

L49

 … species “have” become…

The sentence was deleted during update.

L59

I think detail (1) is about the WRF experiment. Rewrite it if necessary.

The sentence was rewritten (L102).

L70

Snow before the middle of October? Should be after?

The weather in the area is very unique and the first snow generally falls in the middle of October.

L76

Add a description of the red dots. Add map coordinates.

Modified to “Figure 1. The location of Seoraksan National Park in Korea and its typical topographic feature. The red marks indicate the sampling sites of fresh fecal deposits of the Korean long-tailed goral in the park.”

L99

Replot this conceptual figure (for the WRF part) if necessary.

Eliminated the first figure and reploted the second figure.

Sec.4.1

Elaborate whether there are representative errors for collecting goral fecal deposits on the 4 routes since the sample locations gather in small areas of the entire park.

We have no idea how to measure the representative error in collecting fecal samples. We believe we do not proper understand the intend of the reviewer’s comment. Need more explanation on the comment.

L111

Again, WRF is not a climate model.

Changed to “the WRF model”

L114

… to “analyze” historic…

Changed to “analyze”

L122-123

Is any data assimilation performed in the WRF experiment?

No.

L136

… globe “(“. Using…

Corrected to “… globe. Using …”

L137

… days “initialized at” two…

Changed to “initialized at”

L140-141

What are the horizontal resolutions of 70 m, 70 m, 70 m, and 130 m? Are they not 27, 9, 3, and 1 km?

It meant the size of grid like 70x70 and 130x130. Corrected the mistakes.

L145-148

What is the distance between the sample locations and the Wontong weather station? Please show the location of the Wontong weather station in Fig. 2.

In remodeling, we replaced it with the Inje station and the location was marked with a red dot.

L149

Add the unit of meter to the color bar.

The unit was added.

L152

… Numerical “modeling” for…

Corrected to “modeling”

L162

Is the Inje weather station the same as the aforementioned Wontong weather station? If so, use the same name.

“The Inje weather station” meant “the Wontong weather station” in the Inje county. However, there are two different weather stations called “Wontong” and “Inje” around the park and during remodeling, we used the Inje weather data. Replaced “Wontong” to “Inje.”

L163

… 2000) “were” also…

Corrected to “were”

Sec.5.1

Show Table 1.

Table 1 was included.

L176-177

Did not differ? From my point of view, Fig. 4 only shows that the diurnal cycles are captured but the oscillating amplitude of the simulation is excessive.

The sentence was modified. Figure 4 shows hourly-based temperatures and the coefficient and bias is for daily mean temperatures.

L179-180

“which means that the result was good enough to be used for the future climate simulation” does not make sense because the future climate simulation is not carried out with the same WRF model but the ECHO-G model.

This sentence is not valid any more because we used different model. Refer to the modified Section 3.4.

Sec.5.2-5.3

In Figs. 4-10, are the simulated values averaged for the entire domain of Fig. 2? Or averaged for the 4th domain of Fig. 3? Or averaged for the sample locations? Clarify this in the text or figure captions.

Figure 5-10 simply showed the result of WRF and do not have more meanings. So we just reduced them to two examples with smaller sizes of figures for temperature and wind speed. For Figure 4, the temperature only includes the habitat area where the fecal samples were collected like in Figure 7. Added more comments in the figure caption.

L233

… research “report” [28] …

Corrected to “report”

L239

… Park “tends” to…

Corrected to “tends”

L251-252

“successfully produced the current and future climate conditions with acceptable reliability” is questionable when only temperature is verified.

Discussion section was rewritten and the sentence was deleted.

L254-255

“the results of the current and future climate simulation were all clearly acceptable based on the sensitivity test for our time period” is questionable when only temperature is verified.

Discussion section was rewritten and the sentence was deleted.

L272-273

I have no clues what this sentence means.

Discussion section was rewritten and the sentence was deleted.

Sec.7

Modify the conclusion after the major revision.

Conclusion wes revised.

I would suggest to include the website link of Line 63 as a reference instead of in the main text linked to an appendix, in order to avoid misunderstandings looking for an appendix in the manuscript.

It was added as a reference.

The text within some figures is a little bit blurry (for instance, the y-axis label of Figures 4, 5 and 6). Please, try to improve the resolution of these figures, if possible.

Blurry texts are rewritten.

Is the sentence between lines 212-214, which discusses the wind speed values and directions, related with Figure 6?

Figure 6 is mislabeled. It was corrected.

Is the scale of the color bar in Figure 7 correct? Summer [-1, -14]; Winter [27, 14]. In the original version of the manuscript the scale was different. Please indicate the temperature units in the Figure (ºC).

They are corrected like winter (-14, -1) and summer (14, 27). Also, ^oC was inserted in the figure.

The terms of "prediction" and "predicted" should be changed into "projection" and "projected" in the article, including figure captions. Is the result of Figure 6 done by WRF or MK-PRISM? 

“predict” was changed to “project.” Figure 6 is done by MK-PRISM and for better understanding the beginning of the methodology section was modified including Figure 2.

The scale of the color bar in figure 6 between summer and winter is different. Which one is correct?

It is wrong. Summer corresponds to 14-17 and winter to -14 - -1. They are all corrected.

Line 212: The wind speed is not found in Figure 6.

“Figure 6” in the text is removed since it is not in Figure 6. It is in Figure 5.

What kind of role does WRF model play for the future climate analysis in the article?

Because climatic conditions in deep forest and hummocky mountains like Seoraksan National Park cannot not be identified. Note that the closest weather station is inje, 20 km far from the park. Therefore, WRF simulated weather parameters in the park instead. Now, we can understand the climate conditions (more specifically, temperature) of the goral habitat in the park and set the conditions as the goral suitable habitat climate.