Probabilistic seismic risk assessment for cities: Counterfactual analysis in a Chilean case study

Jünemann, Rosita; Urrutia, Alejandro; Damian, Monserrat; Ortiz, Oscar; Zurita, Felipe; Crempien, Jorge G. F.

doi:https://doi.org/10.5194/nhess-2023-141

Preprints

https://doi.org/10.5194/nhess-2023-141

Preprints

03 Aug 2023

| 03 Aug 2023

Status: this preprint has been withdrawn by the authors.

Probabilistic seismic risk assessment for cities: Counterfactual analysis in a Chilean case study

Rosita Jünemann, Alejandro Urrutia, Monserrat Damian, Oscar Ortiz, Felipe Zurita, and Jorge G. F. Crempien

Abstract. We develop a comprehensive probabilistic seismic risk assessment model and apply it to the coastal city of San Antonio in Chile. We use this model to analyze the implications of various counterfactuals of the exposed building stock, in terms of different risk metrics. We begin by generating a synthetic earthquake catalog via Monte Carlo simulations. We then simulate spatial seismic intensities by using ground motion models, considering correlation parameters specifically estimated for the Chilean subduction zone, to evaluate ground motion intensity measures at each site. Additionally, we develop a high-resolution exposure model that characterizes the building stock and population distribution at the census block level. We use the proposed methodology to obtain risk curves for the current exposed building stock, as well as for four counterfactuals involving changes in building materials and/or design levels. Thus, we quantitatively identify the most effective alternatives for mitigation plans. These alternatives consider not only physical damage but also economic losses and casualties, showing the method's potential for its use as a valuable public policy planning tool for decision-makers. Specifically, we find that changing either the building material or the design level of the predominant building class results in significant reductions in expected annual losses of physical damage, casualties and economic losses.

This preprint has been withdrawn.

Received: 31 Jul 2023 – Discussion started: 03 Aug 2023

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 23064 KB)

Withdrawal notice
This preprint has been withdrawn.
Preprint (23064 KB)

Download & links

This preprint has been withdrawn.

Rosita Jünemann, Alejandro Urrutia, Monserrat Damian, Oscar Ortiz, Felipe Zurita, and Jorge G. F. Crempien

Interactive discussion

Status: closed

RC1:
'Comment on nhess-2023-141', Anonymous Referee #1, 12 Nov 2023

The study presents a seismic risk assessment for San Antonio, Chile. The model is used to quantify the potential impact of four different exposure/vulnerability scenarios - involving changes in building materials or design standards - on seismic risk estimates in terms of physical damage, economic losses, and casualties. The study considers a synthetic earthquake catalogue using Monte Carlo simulations together with ground motion models, taking into account specific correlation parameters for the Chilean subduction zone.
The manuscript is generally well-written, with some minor faults. The quality of the illustration is excellent.
Nevertheless, I found it challenging to grasp from the manuscript any actual novelties and advances to existing methodologies for seismic risk assessment of building portfolios or for counterfactual analysis. The authors failed to consider at least a decade of research on the topics discussed in the manuscript; see, for instance, the work by the Global Earthquake Model (GEM) Foundation (only a limited number of studies/references from GEM are considered in this manuscript), the work by Cremen et al. (e.g., https://doi.org/10.1029/2021EF002388 or https://doi.org/10.1177/87552930221134950, among others), the work by the Disaster Analytics for Society Lab at Nanyang Technological University (e.g., https://doi.org/10.3389/feart.2022.847196; https://doi.org/10.3389/feart.2020.575048, among others).
Now, while using existing, state-of-the-art (or at least state-of-practice) tools on a different problem/case study on its own does not provide novelty, learning new insights (even if the use is a simple adoption) does, of course, and this would be acceptable.
Yet, the authors do not make any convincing arguments for new insights in using existing tools on a new case study – the central, trivial conclusion of the study is that “changing either the building material or the design level of the predominant building class results in significant reductions in expected annual losses of physical damage, casualties and economic losses.”
Based on these remarks, I cannot recommend this manuscript for publication in a highly-regarded journal as NHESS.
Specific comments:
Line 42-54: surprisingly, no discussion on the work by the GEM Foundation is included here – and similarly at Line 70 (when discussing existing earthquake risk models for Chile).
Section 2 contains textbook material and could be significantly shortened.
Section 2.1: the use of HAZUS building types and fragility models is not well justified nor validated for the specific case study and context. HAZUS is very US-centric, and, for instance, its fragility models are derived through outdated approaches, not state-of-the-art of state of practice. This is a significant limitation of this study.
Line 127: It is unclear how the building height is mapped to the number of apartments in a building.
Line 139: it is unclear how the authors obtained the built area ak,j for each building class k within each census block j.
Lines 190-200: The use of the ground-motion model of Candia et al. is not well justified. There are other available models for Chile; why has Candia et al. been selected? Did the author apply any selection/scoring procedure using empirical data (e.g., ground motions for Chile) to choose the best available ground-motion model for their case-study region? Is the Heresi and Miranda (2019) spatial correlation model compatible with the selected ground-motion model (e.g., calibrated based on the same data or at least for a similar tectonic context)? What about ground-motion IM cross-correlation needed for regional seismic risk analysis (e.g., buildings in the building stock can have different fundamental periods and hence different IMs used in their fragility models)?
Section 2.4: the capacity spectrum method, especially in the version of Porter (2009), is obsolete, and it has been shown to produce biased damage (and hence fragility) estimates. Record-to-record variability, for instance, is not adequately captured in the methods, and we know this is a critical source of uncertainty in seismic risk assessment exercises.
Page 9 contains a lot of unnecessary textbook material.
Line 242, “The building repair and replacement ratio, which represents the ratio of the total cost of repairing a building to the total cost of replacing it with a new one and is obtained from Hazus (FEMA, 2020) for each building class, building component and damage level.”: is this consequence model appropriate for Chile? How did the author reconcile these loss ratios with the depreciated reconstruction value in their exposure model (e.g., lines 140-145).
Section 3.1: the practical feasibility of the proposed contrafactual strategies is not discussed nor their cost implications. In other words, “modifying the seismic design level of the predominant building type RM1L from LC to HC, resulting in a significant enhancement of its structural performance” may not be technically feasible in practice (or similarly “changing the design level but also the construction material from masonry to reinforced concrete”) and would require high costs (including indirect ones due to demolishing an existing building, environmental impacts, etc.).
Figure 8: are fragility curves in terms of Sd rather than IMs? I guess this is because of the capacity spectrum method approach used in the study, but this is inconsistent with the current state of practice in regional seismic risk assessment.
The conclusions do not offer any new insights from the case-study application. The results are all expected and well-established.

Citation: https://doi.org/10.5194/nhess-2023-141-RC1
- AC1:
  'Reply on RC1', Rosita Junemann, 19 Dec 2023
  
  Dear Reviewer,
  Thank you for your feedback. Please find attached the response to your main observations in the file.
  
  Citation: https://doi.org/10.5194/nhess-2023-141-AC1
  - EC1: 'Reply on AC1', Jörn Lauterjung, 10 Jan 2024
    
    Dear Rosita,
    thank you for the detailed response to the reviewers' comments and sorry for the late reply. I would like to ask you to submit a revised version of the article that takes the reviewers' comments into account.
    Best regards
    Jörn
    
    Citation: https://doi.org/10.5194/nhess-2023-141-EC1
    
    AC3: 'Reply on EC1', Rosita Junemann, 05 Mar 2024
    
    Dear Editor:
    After carefully considering the reviewers' comments and discussing them among all authors, we reached the conclusion that most of the suggested changes are better suited to an article that approaches seismic risk conditioned to a specific event, rather than to ours, with a fully probabilistic approach. We know our work to be novel and a contribution to the literature on seismic risk assessment at a regional scale. Although we are not providing a particular new methodology to estimate seismic risk, we do integrate different state-of-the-art pieces involved in a fully probabilistic risk assessment, at a city level, along with a very high-resolution exposure model. This makes our results on the entire probability distribution for any timeframe of various consequence variables (physical damage, economic losses and casualties), unique. Finally, the tool referred to in the document corresponds to a code developed by the researchers to address the questions posed in this article, and it is not intended to be used or patented as software. Neither the data nor the code constitutes a standalone contribution. Based on these arguments, we decided to conclude the review process at NHESS. We are confident that our work will be attractive to journals with a more closely aligned technical scope and appreciative of the risk assessment approach we adopt. We are nevertheless very grateful to you and the referees for your valuable feedback and insight.
    
    Citation: https://doi.org/10.5194/nhess-2023-141-AC3
RC2:
'Comment on nhess-2023-141', Anonymous Referee #2, 19 Nov 2023

Dear Dear Rosita, Alejandro, Monserrat, Oscar, Felipe and Jorge,
Dear Editor Jörn Lauterjung,
I believe that significant revisions are necessary to move forward with the next steps in publishing this work.
I have provided comprehensive comments that can hopefully be addressed in the next version of the manuscript. Please find attached the general, specific, and technical comments.

Citation: https://doi.org/10.5194/nhess-2023-141-RC2
- AC2: 'Reply on RC2', Rosita Junemann, 19 Dec 2023
  
  Dear reviewer,
  Thank you for your feedback. Please find attached the response to your main observations in the file.
  
  Citation: https://doi.org/10.5194/nhess-2023-141-AC2

Interactive discussion

Status: closed

RC1:
'Comment on nhess-2023-141', Anonymous Referee #1, 12 Nov 2023

The study presents a seismic risk assessment for San Antonio, Chile. The model is used to quantify the potential impact of four different exposure/vulnerability scenarios - involving changes in building materials or design standards - on seismic risk estimates in terms of physical damage, economic losses, and casualties. The study considers a synthetic earthquake catalogue using Monte Carlo simulations together with ground motion models, taking into account specific correlation parameters for the Chilean subduction zone.
The manuscript is generally well-written, with some minor faults. The quality of the illustration is excellent.
Nevertheless, I found it challenging to grasp from the manuscript any actual novelties and advances to existing methodologies for seismic risk assessment of building portfolios or for counterfactual analysis. The authors failed to consider at least a decade of research on the topics discussed in the manuscript; see, for instance, the work by the Global Earthquake Model (GEM) Foundation (only a limited number of studies/references from GEM are considered in this manuscript), the work by Cremen et al. (e.g., https://doi.org/10.1029/2021EF002388 or https://doi.org/10.1177/87552930221134950, among others), the work by the Disaster Analytics for Society Lab at Nanyang Technological University (e.g., https://doi.org/10.3389/feart.2022.847196; https://doi.org/10.3389/feart.2020.575048, among others).
Now, while using existing, state-of-the-art (or at least state-of-practice) tools on a different problem/case study on its own does not provide novelty, learning new insights (even if the use is a simple adoption) does, of course, and this would be acceptable.
Yet, the authors do not make any convincing arguments for new insights in using existing tools on a new case study – the central, trivial conclusion of the study is that “changing either the building material or the design level of the predominant building class results in significant reductions in expected annual losses of physical damage, casualties and economic losses.”
Based on these remarks, I cannot recommend this manuscript for publication in a highly-regarded journal as NHESS.
Specific comments:
Line 42-54: surprisingly, no discussion on the work by the GEM Foundation is included here – and similarly at Line 70 (when discussing existing earthquake risk models for Chile).
Section 2 contains textbook material and could be significantly shortened.
Section 2.1: the use of HAZUS building types and fragility models is not well justified nor validated for the specific case study and context. HAZUS is very US-centric, and, for instance, its fragility models are derived through outdated approaches, not state-of-the-art of state of practice. This is a significant limitation of this study.
Line 127: It is unclear how the building height is mapped to the number of apartments in a building.
Line 139: it is unclear how the authors obtained the built area ak,j for each building class k within each census block j.
Lines 190-200: The use of the ground-motion model of Candia et al. is not well justified. There are other available models for Chile; why has Candia et al. been selected? Did the author apply any selection/scoring procedure using empirical data (e.g., ground motions for Chile) to choose the best available ground-motion model for their case-study region? Is the Heresi and Miranda (2019) spatial correlation model compatible with the selected ground-motion model (e.g., calibrated based on the same data or at least for a similar tectonic context)? What about ground-motion IM cross-correlation needed for regional seismic risk analysis (e.g., buildings in the building stock can have different fundamental periods and hence different IMs used in their fragility models)?
Section 2.4: the capacity spectrum method, especially in the version of Porter (2009), is obsolete, and it has been shown to produce biased damage (and hence fragility) estimates. Record-to-record variability, for instance, is not adequately captured in the methods, and we know this is a critical source of uncertainty in seismic risk assessment exercises.
Page 9 contains a lot of unnecessary textbook material.
Line 242, “The building repair and replacement ratio, which represents the ratio of the total cost of repairing a building to the total cost of replacing it with a new one and is obtained from Hazus (FEMA, 2020) for each building class, building component and damage level.”: is this consequence model appropriate for Chile? How did the author reconcile these loss ratios with the depreciated reconstruction value in their exposure model (e.g., lines 140-145).
Section 3.1: the practical feasibility of the proposed contrafactual strategies is not discussed nor their cost implications. In other words, “modifying the seismic design level of the predominant building type RM1L from LC to HC, resulting in a significant enhancement of its structural performance” may not be technically feasible in practice (or similarly “changing the design level but also the construction material from masonry to reinforced concrete”) and would require high costs (including indirect ones due to demolishing an existing building, environmental impacts, etc.).
Figure 8: are fragility curves in terms of Sd rather than IMs? I guess this is because of the capacity spectrum method approach used in the study, but this is inconsistent with the current state of practice in regional seismic risk assessment.
The conclusions do not offer any new insights from the case-study application. The results are all expected and well-established.

Citation: https://doi.org/10.5194/nhess-2023-141-RC1
- AC1:
  'Reply on RC1', Rosita Junemann, 19 Dec 2023
  
  Dear Reviewer,
  Thank you for your feedback. Please find attached the response to your main observations in the file.
  
  Citation: https://doi.org/10.5194/nhess-2023-141-AC1
  - EC1: 'Reply on AC1', Jörn Lauterjung, 10 Jan 2024
    
    Dear Rosita,
    thank you for the detailed response to the reviewers' comments and sorry for the late reply. I would like to ask you to submit a revised version of the article that takes the reviewers' comments into account.
    Best regards
    Jörn
    
    Citation: https://doi.org/10.5194/nhess-2023-141-EC1
    
    AC3: 'Reply on EC1', Rosita Junemann, 05 Mar 2024
    
    Dear Editor:
    After carefully considering the reviewers' comments and discussing them among all authors, we reached the conclusion that most of the suggested changes are better suited to an article that approaches seismic risk conditioned to a specific event, rather than to ours, with a fully probabilistic approach. We know our work to be novel and a contribution to the literature on seismic risk assessment at a regional scale. Although we are not providing a particular new methodology to estimate seismic risk, we do integrate different state-of-the-art pieces involved in a fully probabilistic risk assessment, at a city level, along with a very high-resolution exposure model. This makes our results on the entire probability distribution for any timeframe of various consequence variables (physical damage, economic losses and casualties), unique. Finally, the tool referred to in the document corresponds to a code developed by the researchers to address the questions posed in this article, and it is not intended to be used or patented as software. Neither the data nor the code constitutes a standalone contribution. Based on these arguments, we decided to conclude the review process at NHESS. We are confident that our work will be attractive to journals with a more closely aligned technical scope and appreciative of the risk assessment approach we adopt. We are nevertheless very grateful to you and the referees for your valuable feedback and insight.
    
    Citation: https://doi.org/10.5194/nhess-2023-141-AC3
RC2:
'Comment on nhess-2023-141', Anonymous Referee #2, 19 Nov 2023

Dear Dear Rosita, Alejandro, Monserrat, Oscar, Felipe and Jorge,
Dear Editor Jörn Lauterjung,
I believe that significant revisions are necessary to move forward with the next steps in publishing this work.
I have provided comprehensive comments that can hopefully be addressed in the next version of the manuscript. Please find attached the general, specific, and technical comments.

Citation: https://doi.org/10.5194/nhess-2023-141-RC2
- AC2: 'Reply on RC2', Rosita Junemann, 19 Dec 2023
  
  Dear reviewer,
  Thank you for your feedback. Please find attached the response to your main observations in the file.
  
  Citation: https://doi.org/10.5194/nhess-2023-141-AC2

Rosita Jünemann, Alejandro Urrutia, Monserrat Damian, Oscar Ortiz, Felipe Zurita, and Jorge G. F. Crempien

Viewed

Total article views: 908 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
559	299	50	908	59	73

HTML: 559
PDF: 299
XML: 50
Total: 908
BibTeX: 59
EndNote: 73

Views and downloads (calculated since 03 Aug 2023)

Month	HTML	PDF	XML	Total
Aug 2023	146	36	7	189
Sep 2023	40	15	0	55
Oct 2023	29	10	4	43
Nov 2023	29	14	0	43
Dec 2023	36	12	4	52
Jan 2024	43	10	2	55
Feb 2024	23	14	2	39
Mar 2024	29	14	4	47
Apr 2024	15	13	8	36
May 2024	12	5	2	19
Jun 2024	19	14	3	36
Jul 2024	17	10	3	30
Aug 2024	10	5	2	17
Sep 2024	7	2	0	9
Oct 2024	15	2	2	19
Nov 2024	11	5	0	16
Dec 2024	4	9	0	13
Jan 2025	15	6	0	21
Feb 2025	7	5	0	12
Mar 2025	17	14	3	34
Apr 2025	15	17	1	33
May 2025	9	13	3	25
Jun 2025	10	53	0	63
Jul 2025	1	1	0	2

Cumulative views and downloads (calculated since 03 Aug 2023)

Month	HTML	PDF	XML	Total
Aug 2023	146	36	7	189
Sep 2023	40	15	0	55
Oct 2023	29	10	4	43
Nov 2023	29	14	0	43
Dec 2023	36	12	4	52
Jan 2024	43	10	2	55
Feb 2024	23	14	2	39
Mar 2024	29	14	4	47
Apr 2024	15	13	8	36
May 2024	12	5	2	19
Jun 2024	19	14	3	36
Jul 2024	17	10	3	30
Aug 2024	10	5	2	17
Sep 2024	7	2	0	9
Oct 2024	15	2	2	19
Nov 2024	11	5	0	16
Dec 2024	4	9	0	13
Jan 2025	15	6	0	21
Feb 2025	7	5	0	12
Mar 2025	17	14	3	34
Apr 2025	15	17	1	33
May 2025	9	13	3	25
Jun 2025	10	53	0	63
Jul 2025	1	1	0	2

Viewed (geographical distribution)

Total article views: 867 (including HTML, PDF, and XML) Thereof 867 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 04 Jul 2025

Download

This preprint has been withdrawn.

Preprint (23064 KB)
Metadata XML

Short summary

We developed a probabilistic seismic risk assessment model and applied it to a coastal city in Chile, incorporating high-resolution exposure model and Monte Carlo simulations for seismic hazard. Evaluating various counterfactual scenarios of building stock, we quantitatively identified the most effective mitigation options, considering not only physical damage but also economic losses and casualties. The method has potential use as a valuable public policy planning tool for decision-makers.


Total:	0
HTML:	0
PDF:	0
XML:	0