Development of Building Inventory Data in Ulaanbaatar, Mongolia for Seismic Loss Estimation

Round 1

Reviewer 1 Report

This article presents a very interesting way to measure the construction costs that could be derived from the possibility of an event of momento magnitude of 7.4 occurring in the city of Ulaanbaatar in Mongolia. The text is very well written, the objective of the article, the presentation of the problem, as well as the methods used to carry out this study. The authors give enough information about the constructions and the vulnerability of some of them. They put in front of the reader the detail to be able to have an estimate of the construction costs in the studied city. They locate the fault and estimate the earth movement based on the stochastic Green's function method (SGFM).

I have a couple of questions, first, although they do a geographic contextualization, I don't really understand if the area has some kind of cycle of occurrence of seismic events of moment magnitudes over 6.0, is this so? Could you add in the introduction a little of the seismic history that has affected this area? I understand that this work is complete as it is, but could the effect of the different waves present in an earthquake be considered in another analysis? that is, what about the surface and rotational waves?

The construction costs that you give as a conclusion of this work are quite high. Could you identify a range of costs, or consider a margin of error? 

Author Response

Authors’ Responses to Reviewer #1

 

General comment:

This article presents a very interesting way to measure the construction costs that could be derived from the possibility of an event of moment magnitude of 7.4 occurring in the city of Ulaanbaatar in Mongolia. The text is very well written, the objective of the article, the presentation of the problem, as well as the methods used to carry out this study. The authors give enough information about the constructions and the vulnerability of some of them. They put in front of the reader the detail to be able to have an estimate of the construction costs in the studied city. They locate the fault and estimate the earth movement based on the stochastic Green's function method (SGFM).

Authors’ response:

First of all, we appreciate your kind comments. We would like to reply to your questions as shown below.

 

Comment 1:

I have a couple of questions, first, although they do a geographic contextualization, I don't really understand if the area has some kind of cycle of occurrence of seismic events of moment magnitudes over 6.0, is this so? Could you add in the introduction a little of the seismic history that has affected this area?

Authors’ response:

 Thank you for your helpful comment. In the introduction part, we added some descriptions for the seismic history in Ulaanbaatar city, as shown in Line 37-43 of Page 1.

 

Comment 2:

I understand that this work is complete as it is, but could the effect of the different waves present in an earthquake be considered in another analysis? that is, what about the surface and rotational waves?

Authors’ response:

 Thank you for your comment. If the three-dimensional shear-wave velocity structure model is completed in the target city, we can apply detailed numerical ground motion simulation techniques such as the 3D finite difference method that can predict not only body-wave but also basin-induced surface waves. Since our current Vs-model did not support the 3D model, we applied Stochastic Green’s function method to consider the propagation of S-wave. We added future insight for detailed ground motion simulation in the Conclusion part, as shown in Line 505-510 of Page 22.

 

Comment 3:

The construction costs that you give as a conclusion of this work are quite high. Could you identify a range of costs, or consider a margin of error?

Authors’ response:

  The construction costs were estimated based on the procedure developed by the Ministry of the Mongolian government. Although the coefficients would contain range or error, they did not evaluate the variance. So, we just applied the coefficients introduced in the procedure. Since we do not have available data to validate the coefficients, we would like to evaluate the coefficients for future work. Thank you very much for your comment.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors present an interesting study providing an attempt to establish building inventory database for seismic loss estimation of Ulaanbaatar, Mongolia. In general, the study has potential and practical value. However, there are a good many issues (see below) that need to be addressed before accepting final decision on relevance of the manuscript for publication in ISPRS International Journal of Geo-Information.

 

1. Citing of relatively old references – almost all cited refs. are older than 5 years. Besides, the most important source [20] considered in the study is in Monogolian language and not broadly available, what makes a study difficult to revise or repeat.
2. Research material: the main shortage of the study seems to be the lack of the building information in the applied existing inventory data (p. 6, line 174): questionable inventory data reliability, incomplete database (e.g. more than 50 % of the building structural types are not registered – p. 3, line 118; cca. 55 % of the inventory with “unknown” No. of storeys (Tab. 2), cca. 63 % with unknown construction year!? (Tab. 3)), data of poor quality (inaccurate or even erroneous data, misclassifications in the inventory etc.). The background of the data collection is missing – when was the building census performed, who were the registrars, what was the applied methodology?

Although the authors tried to “correct” the original data, this reviewer cannot rid of the well-known undesired fact in the research approach i.e. “garbage in – garbage out”. This absolutely reduces scientific soundness and reliability of the presented research.

3. The selection of the building inventory data extent in relation to structural seismic resistance seems unusual and needs additional explanation. Among common (simplified) structural parameters (structural type, location, height, year of construction) a heating system was also considered. Why only this system? What about other building elements (e.g. non-structural elements (facades), cooling systems, thermal insulation envelope, solar energy systems etc.) which might also substantially affect construction costs and increase seismic loss?

In addition, the selection of the required heating system and the thermal performance of the building are strongly related to the climate zone (p. 6, line 168-169), and of the built thermal insulation envelope. The latter fact is missing in the presented study.

4. The authors’ criteria for the estimation and correction (related to comment #2) of the existing census data (section 2.3 – 2.5) needs additional justification and thorough revision. The applied approach is unclear and does not convince this reviewer as well as potential readership. The methodology and criteria (Tables 4-6) for determination of the assumed building characteristics do not follow common scientific techniques and methods, use non-engineering, illogical, or arbitrary choice approach for selection of some investigated parameters (e.g. building’s shape VS. structural type & material (p. 7), the criterion of google street photos and the authors’ subjective deductions (p.8, Fig. 4), etc.). Most of the applied criteria are non-convincing and open plenty of questions. Applying mostly hypothetical approach, without quotation of real and provable background data makes the study incomplete and reduce its scientific soundness.
5. Among the quoted structural types, some common and frequent structures (e.g. bare reinforced concrete structures) seem to be forgotten in the analysis. Why? It's hard to believe that there are no structures of this type in Mongolia. Explain, please.
6. Table 2: incorrect number in the last row of the 6th column – 32,550 instead of 3,2550!
7. Fig. 3: the figure’s caption should be revised. Current caption is very similar/equal to caption of Fig. 2. A distinction between figures is desired – use a diction as it is stated on page 3, line110-111.
8. Figure presentation quality needs to be improved – e.g. poor readability of Fig. 2, Fig. 5, Fig. 8, Fig. 14, and Fig. 15.
9. Section 2.5: more explanation on the notion “construction cost” is needed – what is included in the considered cost values? Only the cost of material, also the costs of execution, execution to which construction phase? Etc.

Additional comment on the construction costs relates to the illogical (in EU or USA not valid) fact that steel structures are cheaper than RC structures. Additional argumentation is needed.

For easier comparison, instead of the quotation of the total (absolute) costs (Figs. 5-7) the consistent use of cost per unit (USD/m2 – as in Tab. 7) is recommended.

10. Observing the results in Fig. 7 this reviewer does not agree with the statement on p. 10, line 280 (for BGD district the max. discrepancy between statistics and estimation is larger than factor 2!?
11. Section 2.6: The explanation on the quoted (Table 8) “typical period” is missing. How were the values (precision 0,01 s!?) calculated (“estimated”)? Was the same period assumed for the buildings with different number of storeys? As an expert in structural dynamics and earthquake engineering this reviewer cannot agree with this rough approach. Please, justify.

For the applied vulnerability functions adopted from GAR-13 (Fig. 9) more data would be appreciated (background for their calculation, determination of the design level etc.). In addition, the label of X axis of the graph for building class C1M needs correction (erase “1”).

12. 13: the legends need additional information on the meaning of the colours – the units for the values 100 – 3000 are missing.
13. The current section Conclusions needs revision and upgrade in terms of:

• providing the limitations of the study and of the proposed methodology;
• discussion on the availability and quality of the applied input data;
• additional argumentation for the currently non-convincing statement on p. 20, lines 441-444 is needed: HOW did the authors confirm the agreement “estimated VS. actual”?;
• maybe addition of wider-practical aspect and value of the performed study – e.g. the applicability of the obtained data for the earthquake insurance market for building stock in UB, Mongolia.

Author Response

Authors’ Responses to Reviewer #2

 

General comment:

The authors present an interesting study providing an attempt to establish building inventory database for seismic loss estimation of Ulaanbaatar, Mongolia. In general, the study has potential and practical value. However, there are a good many issues (see below) that need to be addressed before accepting final decision on relevance of the manuscript for publication in ISPRS International Journal of Geo-Information.

Authors’ response:

  First of all, we appreciate your kind comments. Your review is indispensable for this paper. We would like to reply to your questions as shown below.

 

Comment 1:

Citing of relatively old references – almost all cited refs. are older than 5 years. Besides, the most important source [23] considered in the study is in Mongolian language and not broadly available, what makes a study difficult to revise or repeat.

Authors’ response:

Thank you for your suggestions. We added recent studies related to building inventory development and earthquake loss estimations in other countries, as shown in Line 77-81 of Page 2 of the revised manuscript.

   We are also sorry that the reference [20è23] was not translated into English. In order to assess building inventory in a country, original documents written in the local language are indispensable, especially in developing countries such as Mongolia. Since the reference [23] is available on the website, the revised manuscript adds the URL.

 

Comment 2:

Research material: the main shortage of the study seems to be the lack of the building information in the applied existing inventory data (p. 6, line 174): questionable inventory data reliability, incomplete database (e.g. more than 50 % of the building structural types are not registered – p. 3, line 118; cca. 55 % of the inventory with “unknown” No. of storeys (Tab. 2), cca. 63 % with unknown construction year!? (Tab. 3)), data of poor quality (inaccurate or even erroneous data, misclassifications in the inventory etc.). The background of the data collection is missing – when was the building census performed, who were the registrars, what was the applied methodology?

Although the authors tried to “correct” the original data, this reviewer cannot rid of the well-known undesired fact in the research approach i.e. “garbage in – garbage out”. This absolutely reduces scientific soundness and reliability of the presented research.

Authors response:

Thank you for your comments. As you pointed out, the original building inventory data was not perfect and included a lot of information lacks. Besides, our corrected building inventory is not also perfect, and the updated information would include not a few errors. However, our corrected inventory data is certainly better than the original one since the unknown data is updated based on the criteria developed from the available data, such as the urban expansion map, the land use map, and the known data in the original inventory data. Although the number of validations is limited, the reliability of the proposed approach is discussed, as shown in Fig. 4. Future studies would be required to complete the inventory data by cooperation with the national government and/or local municipalities. The descriptions for future studies have added the explanation in Line 510-514 of Page 22 in the revised manuscript.

 

Comment 3:

The selection of the building inventory data extent in relation to structural seismic resistance seems unusual and needs additional explanation. Among common (simplified) structural parameters (structural type, location, height, year of construction) a heating system was also considered. Why only this system? What about other building elements (e.g. non-structural elements (facades), cooling systems, thermal insulation envelope, solar energy systems etc.) which might also substantially affect construction costs and increase seismic loss?

In addition, the selection of the required heating system and the thermal performance of the building are strongly related to the climate zone (p. 6, line 168-169), and of the built thermal insulation envelope. The latter fact is missing in the presented study.

Authors response:

As you suggested, all other elements need to be considered to evaluate construction costs for a specific building. However, collecting all necessary information for urban scale building inventory would be very difficult. According to the procedure in Ref. [23], construction cost was evaluated from structural type, location, height, and heating system. That is why we used the current parameters for construction cost estimation. As described in the manuscript, the heating system is one of the important factors in constructing a building in Mongolia because Mongolia is located in a cold region. UB has a central heating system, the main infrastructure for buildings (supply heating energy). The parameters for construction cost estimation applied in this study were based on the procedure introduced by the Ministry of construction and urban development in Mongolia [Ref. 23]. Other parameters can be considered in order for more detailed cost estimation in future studies, but the current study did not consider other parameters.

 

Comment 4:

The authors’ criteria for the estimation and correction (related to comment #2) of the existing census data (section 2.3 – 2.5) needs additional justification and thorough revision. The applied approach is unclear and does not convince this reviewer as well as potential readership. The methodology and criteria (Tables 4-6) for determination of the assumed building characteristics do not follow common scientific techniques and methods, use non-engineering, illogical, or arbitrary choice approach for selection of some investigated parameters (e.g. building’s shape VS. structural type & material (p. 7), the criterion of google street photos and the authors’ subjective deductions (p.8, Fig. 4), etc.). Most of the applied criteria are non-convincing and open plenty of questions. Applying mostly hypothetical approach, without quotation of real and provable background data makes the study incomplete and reduce its scientific soundness.

Authors’ response:

Thank you for your comment. As you pointed out, our criteria for the justification of census data need a detailed explanation. That’s why we added Fig. 4 for the revised manuscript and explanation in Lines 227-252 of Page 7-8. There are other possible methods to find out structural types, but this is the one of criteria for our existing data. We have tried more complex ways to define it, but in the end, we prefer a simpler way to find out.

 

 

Comment 5:

Among the quoted structural types, some common and frequent structures (e.g. bare reinforced concrete structures) seem to be forgotten in the analysis. Why? It's hard to believe that there are no structures of this type in Mongolia. Explain, please.

Authors’ response:

Thank you for your interesting comments. Of course, bare reinforced concrete structures are common in the world. But Mongolia is a nomadic country, historically we have used yurt traditional small moveable houses (ger house), 16th-century timber and masonry temples, and modern masonry apartment house have been built since 1940s. Until now people are living in ger houses in the rural district of Ulaanbaatar, because of the cheap price and ease of construction. That’s why bare reinforced concrete structures are not common in Mongolia due to harsh climate (dry and cold) and poor construction practices. The most ordinary buildings are simple wooden bulk with masonry facade or masonry wall with a wooden slab in the Mongolian building practice. Therefore, we are focused on timber structure, masonry, and reinforced concrete structures.

 

Comment 6:

Table 2: incorrect number in the last row of the 6th column – 32,550 instead of 3,2550!

Authors’ response:

Thank you for your correction. It is corrected in Table 2 of the revised manuscript.

 

Comment 7:

Fig. 3: the figure’s caption should be revised. Current caption is very similar/equal to caption of Fig. 2. A distinction between figures is desired – use a diction as it is stated on page 3, line110-111.

Authors’ response:

   Thank you for your correction.  We have changed the caption of the figures in the revised manuscript.

 

Comment 8:

Figure presentation quality needs to be improved – e.g. poor readability of Fig. 2, Fig. 5, Fig. 8, Fig. 14, and Fig. 15.

Authors’ response:

Thank you for your comments. As per your suggestions, the quality of those figures is updated in the revised manuscript.

 

Comment 9:

Section 2.5: more explanation on the notion “construction cost” is needed – what is included in the considered cost values? Only the cost of material, also the costs of execution, execution to which construction phase? Etc.

Additional comment on the construction costs relates to the illogical (in EU or USA not valid) fact that steel structures are cheaper than RC structures. Additional argumentation is needed.

For easier comparison, instead of the quotation of the total (absolute) costs (Figs. 5-7) the consistent use of cost per unit (USD/m2 – as in Tab. 7) is recommended.

Authors’ response:

   Thank you for your comments. According to the procedure in Ref. [23], construction cost includes the direct cost of the entire construction, including materials, supplies, labor, tools, equipment, machine, and transportation. It excludes the cost of land acquisition for the construction of the facility, the cost of relocating the utilities, and the cost of the plant's construction technology equipment, as well as the cost of the land and sales. As your suggestions, we added an explanation in Line 161-167 of Page 6 in the revised manuscript.

As you pointed out, steel structure usually means large span industrial or storage buildings in Mongolia. Based on low rise and large area building cases, the unit cost is cheaper than other types [Ref. 23]. Besides, Table 7 is modified because there is no steel structures other than the office, industrial buildings, and storehouses in the inventory data.

Thank you for your recommended unit, we are tried to estimate the total cost of the damage estimation, that’s why we show the total cost of building (thousand USD) in Fig. 5-7. For that reason, we prefer to use units thousand USD instead of USD/m².

 

Comment 10:

Observing the results in Fig. 7 this reviewer does not agree with the statement on p. 10, line 280 (for BGD district the max. discrepancy between statistics and estimation is larger than factor 2!?

Authors’ response:

Thank you for your correction. We have revised the reviewer’s point and updated the manuscript.

 

Comment 11:

Section 2.6: The explanation on the quoted (Table 8) “typical period” is missing. How were the values (precision 0,01 s!?) calculated (“estimated”)? Was the same period assumed for the buildings with different number of storeys? As an expert in structural dynamics and earthquake engineering this reviewer cannot agree with this rough approach. Please, justify.

For the applied vulnerability functions adopted from GAR-13 (Fig. 9) more data would be appreciated (background for their calculation, determination of the design level etc.). In addition, the label of X axis of the graph for building class C1M needs correction (erase “1”).

Authors’ response:

Thank you for your comments. From the view of structural dynamics and earthquake engineering, the applied approach is a rough estimation for individual building case. However, in city or district scale estimation, it is almost impossible to provide the exact typical period to each building, and also to provide fragility/vulnerability functions to all the building types. For that reason, the global derived vulnerability curves suggested in the GAR-13 was applied for the building damage assessment in this study.

As you pointed out C1M is corrected in the revised manuscript.

 

Comment 12:

13: the legends need additional information on the meaning of the colours – the units for the values 100 – 3000 are missing.

Authors’ response:

Thank you for your comments. As your adjustment, units (cm/s²) are added in the figure of the updated manuscript.

 

Comment 13:

The current section Conclusions needs revision and upgrade in terms of:

providing the limitations of the study and of the proposed methodology;

discussion on the availability and quality of the applied input data;

additional argumentation for the currently non-convincing statement on p. 20, lines 441-444 is needed: HOW did the authors confirm the agreement “estimated VS. actual”?;

maybe addition of wider-practical aspect and value of the performed study – e.g. the applicability of the obtained data for the earthquake insurance market for building stock in UB, Mongolia.

Authors’ response:

 Thank you for your suggestions. We added some parts related to limitations, as shown in Line 493-514 of Page 22 of the revised manuscript.

 

 

Author Response File: Author Response.docx

Reviewer 3 Report

The paper presents the development of an inventory data in UB, including structural types, construction year, height, and construction cost to assess the buildings' vulnerability (repair cost) due to a scenario earthquake. It's a very intersting approach, well presented and structured. I suggest accepting after minor revisions.

• Figure 2a and b are not easily readable. If you have produced these maps please improve. If not, mention the reference at the figures' captions. Same for figure 5, 8.
• Figure 13: district's limits are not visible
• How could this information help stakeholder, urban planners etc?
• What are the limitations and the strengths of your approach?
• What could be the future work of this approach?
• Could this approach be generic in order to be implemented in other countries? If so, what are the criteria to implement it? If not, why?

Author Response

Authors’ Responses to Reviewer #3

 

General comment:

The paper presents the development of an inventory data in UB, including structural types, construction year, height, and construction cost to assess the buildings' vulnerability (repair cost) due to a scenario earthquake. It's a very interesting approach, well presented and structured. I suggest accepting after minor revisions.

Authors’ response:

  First of all, we appreciate your kind comments. We would like to reply to your questions as shown below.

 

Comment 1:

Figure 2a and b are not easily readable. If you have produced these maps please improve. If not, mention the reference at the figures' captions. Same for figure 5, 8.

Authors’ response:

  We improved Fig. 2, 5, and 8 as shown in the revised manuscript.

 

Comment 2:

Figure 13: district's limits are not visible

Authors response:

  We improved the figure as shown in the revised manuscript.

 

Comment 3:

How could this information help stakeholder, urban planners etc?

Authors response:

  The obtained results in this work would be helpful not only to identify the vulnerable areas for future earthquakes but also to estimate necessary costs if a Magnitude-7 class earthquake happens. The information can be used by the national government and/or local municipalities to consider countermeasures for future earthquakes. The future insights for using the obtained information are added in the revised manuscript, as shown in Line 489-492 on Page 21-22.

 

Comment 4:

What are the limitations and the strengths of your approach?

Authors’ response:

  The strengths of our approach can be summarized as that monetary loss of damaged buildings can be estimated not only for scenario earthquakes but also for a real earthquake by using our developed building inventory. Generally, it would be very difficult work to estimate the total amount of loss immediately after an earthquake. Our developed approach can produce a damage distribution map and amount of building loss immediately after an earthquake if ground motion data is available.

  On the other hand, the limitations of our approach can be summarized as the applicability of the vulnerability functions and the accuracy of the seismic ground motion prediction. Although we used the global vulnerability functions proposed in GAR-13, the applicability of the functions needs to be discussed by considering the actual seismic capacity of the buildings in Ulaanbaatar city. Since detailed shear-wave velocity structure models in Ulaanbaatar is not available yet, we apply one-dimensional ground response analysis for seismic ground motion prediction. If a more detailed Vs-model would be available by future dense geophysical explorations, 3D ground motion simulation techniques such as the finite difference method can be applied to estimate not only body waves but also basin-induced surface waves.

 These strengths and limitations are added in the revised manuscript, as shown in Line 492-510 on Page 22.

 

Comment 5:

What could be the future work of this approach?

Authors’ response:

  The necessary future works are explained in the previous response. We added the description for the future works in the revised manuscript as shown in Line 501-514 of Page 22.

 

Comment 6:

Could this approach be generic in order to be implemented in other countries? If so, what are the criteria to implement it? If not, why?

Authors’ response:

  The applicability of our approach to other countries strongly depends on the availability of the data. If basic building inventory or cadastral data, coefficients of construction costs for building types, and underground shear-wave velocity structure model are available, our approach can be implemented in other areas or countries. But it is not easy to collect such data, sometimes unavailable in some countries. Therefore, the approach needs to be discussed considering the situation of each country.

 

Author Response File: Author Response.docx

Reviewer 4 Report

The paper deal with the impact, in terms of damages a losses to buildings in Ulaanbaatar city (UB), due to earthquakes with high magnitude. The consequence scenarios are estimated using available buildings' global vulnerability curves including the repair cost considering the building structural typology, construction year, height, and construction cost for possible estimated spectral accelerations.

The addressed topic is very interesting and fits well with the purpose of the Journal. Only few suggestion to the Authors:

- improve the English style;

- improve the quality of the figures (in particular Figs 2, 3, 5, 8, 14 and 15);

- well clarify that the used vulnerability curves are not referenced to the investigated building sample but are global-derived and, therefore, the results may be affected by uncertainties that can be get over if a specific vulnerability curves - for the particular considered building features - would be generated.

Author Response

Authors’ Responses to Reviewer #4

 

General comment:

The paper deal with the impact, in terms of damages a losses to buildings in Ulaanbaatar city (UB), due to earthquakes with high magnitude. The consequence scenarios are estimated using available buildings' global vulnerability curves including the repair cost considering the building structural typology, construction year, height, and construction cost for possible estimated spectral accelerations.

The addressed topic is very interesting and fits well with the purpose of the Journal. Only few suggestion to the Authors:

Authors’ response:

First of all, we appreciate your kind comments. We would like to reply to your questions as shown below.

 

Comment 1:

improve the English style;

Authors’ response:

 We improved the English style as shown in the revised manuscript as per your suggestion.

 

Comment 2:

improve the quality of the figures (in particular Figs 2, 3, 5, 8, 14 and 15);

Authors’ response:

 As per your suggestion, we improved Fig. 2, 3, 5, 8, 14, and 15, as shown in the revised manuscript.

 

Comment 3:

well clarify that the used vulnerability curves are not referenced to the investigated building sample but are global-derived and, therefore, the results may be affected by uncertainties that can be get over if a specific vulnerability curves - for the particular considered building features - would be generated

Authors’ response:

  Thank you for your kind comment. It is one of the limitations of our study; we hope to investigate it in future studies. The large-scale earthquake has happened in Mongolia during the last century, but the location was far from a city or concentrated area. That’s why there is no investigated building in Ulaanbaatar, Mongolia. These limitations are added in the revised manuscript, as shown in Line 501-505 of Page 22.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

After the re-review of the above-referenced paper it has to be mentioned, that the overall quality of the revised manuscript is now much improved. However, after reading the authors’ response to the reviewers’ comments and the revised paper, this reviewer is only partially pleased with the explanations given and the effort put by the authors in the revision of the manuscript. In the revised manuscript the main concerns (i.e. questionable building inventory data reliability and the authors’ criteria for the data estimation and correction) remains what makes the study incomplete and reduces its scientific soundness. There are a few additional issues that need to be addressed before accepting the manuscript for publication in ISPRS International Journal of Geo-Information:

1) In order to improve the scientific soundness and to convince the readership the authors should provide (include in the paper) more comprehensive explanations and justification on the applied approach and research material:

• The background and justification for the used corrections on the poor existing input data should be presented more broadly.
• Current arbitrary choice approach for selection of some investigated parameters base on the building’s geometry (Fig. 4) and other non-engineering approach (e.g. google street photos and the authors’ subjective deductions) etc. does not convince the critical reader and opens plenty of questions.
• The reviewer comment #11 from the 1^st round review remains unanswered. The explanation on the quoted values of “typical period” (Table 8) is missing. The periods were “estimated” with precision 0,01 s!? This is absolutely uncommon and unrealistic. Furthermore, this reviewer agrees that it is almost impossible to provide the »exact« typical period for each building in such a macro (city) scale model. However, the considered approach assuming the same period for the groups of buildings (with a large difference in the number of storeys) of the same class (building material) is too rough and unrealistic. Quotation of the periods in precision 0.01s additionally reduces the scientific confidence in the performed study.

 

2) The readability of some Figures is still poor.

 

3) English proof-reading of the revised manuscript is warmly recommended. In current paper there are still some language and style mistakes.

Besides, the authors should revise the following mistakes:

• delete space after K_natural on p. 6, line 187
• add missing full stop at the end of caption of Fig. 4
• in the sentence on p. 22 lines 499-500 a verb is missing.

 

Author Response

Authors’ Responses to Reviewer #2 (Round 2)

 

General comment:

After the re-review of the above-referenced paper it has to be mentioned, that the overall quality of the revised manuscript is now much improved. However, after reading the authors’ response to the reviewers’ comments and the revised paper, this reviewer is only partially pleased with the explanations given and the effort put by the authors in the revision of the manuscript. In the revised manuscript the main concerns (i.e. questionable building inventory data reliability and the authors’ criteria for the data estimation and correction) remains what makes the study incomplete and reduces its scientific soundness. There are a few additional issues that need to be addressed before accepting the manuscript for publication in ISPRS International Journal of Geo-Information:

Authors’ response:

  First of all, we appreciate your kind comments. We would like to reply to your questions as shown below.

 

Comment 1-1:

In order to improve the scientific soundness and to convince the readership the authors should provide (include in the paper) more comprehensive explanations and justification on the applied approach and research material:

The background and justification for the used corrections on the poor existing input data should be presented more broadly.

Current arbitrary choice approach for selection of some investigated parameters base on the building’s geometry (Fig. 4) and other non-engineering approach (e.g. google street photos and the authors’ subjective deductions) etc. does not convince the critical reader and opens plenty of questions.

Authors’ response:

Thank you for your comment. We modified the explanations for structural type estimations as shown in Line 239-265 of Page 8 in the revised manuscript. In order to clarify the estimation results, the number of the classified buildings are added in Table 4, 5 and 6 as shown in the revised manuscript. As you know Figure 5 shows the examples for correctly classified buildings by our proposed criteria. Of course, misclassifications would be also found in our estimation. In order to comprehensively evaluate and/or justify our approach, numbers of answers (actual building types) need to be prepared in comparison with our estimations. However, such answers are not available in the current database, thus it is difficult to quantify the accuracies of our estimations. Since such uncertainties of our estimations still remain, the limitation of our approach needs to be clarified in Line 275-277 of Page 8-9 and Line 527-531 of Page 23 in the revised manuscript.

 

Comment 1-2:

The reviewer comment #11 from the 1st round review remains unanswered. The explanation on the quoted values of “typical period” (Table 8) is missing. The periods were “estimated” with precision 0,01 s!? This is absolutely uncommon and unrealistic. Furthermore, this reviewer agrees that it is almost impossible to provide the »exact« typical period for each building in such a macro (city) scale model. However, the considered approach assuming the same period for the groups of buildings (with a large difference in the number of storeys) of the same class (building material) is too rough and unrealistic. Quotation of the periods in precision 0.01s additionally reduces the scientific confidence in the performed study.

Authors’ response:

Thank you for your comment. We apologize for not correctly answering your previous comment. As answered in our previous response, the building class and the vulnerability functions proposed in the Global Assessment Report on Disaster Risk Reduction 2013 (GAR-13) were used in this study. As shown in the reference below, the building heights are categorized to three classes (Low-rise, Mid-rise, and High-rise) in the building classes of the GAR-13. The typical periods were also given to the building classes in GAR-13. In our understanding, the typical period of the GAR-13 were not "estimated" value but "representative" value for each building height class. Of course, the typical (representative) periods would not be accurate for individual buildings because actual building heights vary within the building category. Such building height categories, however, have been widely accepted not only in the GAR-13 but also in the HAZUS, the standardized tools for estimating risk from natural disasters by the United States. Although the accuracies of the typical periods need to be discussed considering actual seismic behaviors of buildings in each country/region, it is beyond the scope of this study. Therefore, we accepted "the typical periods" proposed in the GAR-13 in the current research. At last, we understand that it is a common procedure for urban-scale building damage estimations. As shown in Line 519-522 of Page 23, the applicability of the building categories and the vulnerability functions including the typical periods to the target area needs to be discussed in future studies.

References:

1. International Centre for Numerical Methods in Engineering and ITEC S.A.S – INGENIAR LTDA – EAI S.A. Probabilistic Modelling of Natural Risks at the Global Level: Global Risk Model; Background paper prepared for the Global Assessment Report in Disaster Risk Reduction 2013; Geneva, Switzerland, 2013. (https://www.preventionweb.net/english/hyogo/gar/2013/en/bgdocs/CIMNE%20et.al.%202013a.pdf)
2. Hazus -MH MR5, Advanced Engineering Building Module (AEBM) Technical and User’s Manual, (https://www.hsdl.org/?view&did=12756).

 

Comment 2:

The readability of some Figures is still poor.

Authors response:

Thank you for your comments. We have updated the quality of the figures.

 

Comment 3:

English proof-reading of the revised manuscript is warmly recommended. In current paper there are still some language and style mistakes.

Besides, the authors should revise the following mistakes:

• delete space after K_natural on p. 6, line 187
• add missing full stop at the end of caption of Fig. 4
• in the sentence on p. 22 lines 499-500 a verb is missing.

 Authors response:

As you suggested, we have corrected those mistakes.

 

Author Response File: Author Response.docx