Integration of Geoscience Information for Disaster Resilience in Kuala Lumpur, Malaysia

Reviewer comment

Authors’ Response

Reviewer 1

  However, the manuscript mainly focused on mapping and the novelty, new findings, and scientific contribution of this paper are not clear. Moreover, descriptions in the conclusion section are completely repeated from the results and discussion sections, and key conclusions/findings of the study are completely missing, and the manuscript needs substantial revisions and clarifications before consideration for publication.

The results reveal that 80% of the land area in Kuala Lumpur has medium to high ground constraint, and this includes around 25% of the city that is susceptible to landslides and floods. As the city is moving to the phase of rejuvenation due to land scarcity, the pace of development is more deliberate, and factors that control both these hazards are relatively more stable compared to the trigger factor i.e. rainfall, which the IPCC projects will increase in intensity and frequency in this region due to climate change. Thus, the maps can be used under current and future climate conditions. 

Please see the substantial revisions and clarifications provided below, to address the constructive comments of the reviewers, which has improved the manuscript tremendously.

1.     Introduction and Methods: What is the novel part in terms of methodology and scientific contributions of this study is not clear. In addition, some parts/results of the paper have been already published by authors in their previous paper (https://doi.org/10.3390/app13020768). Authors should clearly mention improvement in this manuscript.

This paper highlights a novel approach for integrating geoscience information to determine engineering ground conditions and susceptibility to multiple geohazards, which are then combined to demarcate suitable zones for urban development and enhance the resilience in Kuala Lumpur, Malaysia. A concise review of the literature has been added to indicate the use of geoscience information for two separate purposes, which is now combined. The scientific contribution of this study lies in the method of weighted overlaying to generate engineering ground classification map, landslide susceptibility map and flood susceptibility map to produce the development suitability zoning map, which can be used under current and future climate conditions.

This has been added to page 2 in section “1.Introduction”,  paragraph 3-5 in lines  50-89.

This manuscript, builds on previously published papers, which focuses on the validation of a single hazard i.e. landslide susceptible model [19], and flood model [47] from the research team members.

2. Lines 21-22 and also in discussion section, authors claimed that 80% of the Kuala Lumpur areas could be associated with more intense and frequent hazards due to climate change without showing any evidences. The climate can be one of the factors, but other environmental factors and development activities might be also the reasons for more intense and frequents hazards in the areas.

The authors agree that climate is one of several factors (including other physical and environmental factors and development activities) that cause more intense and frequents hazards. In this study, susceptibility modelling was conducted to delineate areas where there is potential for a hazard to occur, based on local geological, geomorphological and other physical conditions, which are mapped through field observation or remote sensing. Rainfall is considered a trigger factor for both landslides and floods, and is not used as a parameter for susceptibility modelling.

The results reveal that 80% of the land area in Kuala Lumpur has medium to high ground constraint, and this includes around 25% of the city that is susceptible to landslides and floods. As the city is moving to the phase of rejuvenation due to land scarcity, the pace of development is more deliberate, and factors that control both these hazards are relatively more stable compared to the trigger factor i.e. rainfall, which the IPCC projects will increase in intensity and frequency in this region due to climate change. Based on these assumptions, and adopting the worst case scenario in line with the precautionary principle, climate change is expected to exacerbate the occurrence of landslides and floods in these susceptible areas.  Hence, further development of such areas should be minimised, and appropriate measures must be taken to reduce the exposure and vulnerability of communities and urban assets to future hazards within these susceptible areas.

The paragraph has been revised for clarity in page 13, Section 4 on Discussion, paragraph 1 and 2, lines 416-436. A brief explanation on susceptibility modelling has also been added in Section 1 on Introduction,  paragraph 4, lines 62-67.

3. Lines 130-135, authors used LiDAR DTM for landslide susceptibility mapping, however different DTM from Civil Engineering and Urban Department was used for flood susceptibility mapping. I think in terms of vertical accuracy, LiDAR DTM might be better. Please clarify the reasons for using different DTM data for landslide susceptibility and flood susceptibility mappings. Why same DTM was not used for both susceptibility mapping?

The authors would like to apologize for the confusion.

The same LiDAR dataset with 1 m resolution was used for landslides and flood modelling. The data was downscaled to 5 m to reduce the processing time. 

This information has been modified in page 4, section “2.2 Acquisition of geoscience information”, in paragraph 1, line 149

4.  Lines 138-141, authors mentioned that aerial photographs and topographic maps as well as satellite images were collected, but it is not clear the purposes of collecting such maps. Authors should clearly mention that how such information from the maps and satellite images were incorporated in the study.

Aerial photographs, topographic maps and satellite images were used to assess temporal and spatial land use changes. This is critical to identify the areas affected by past tin-mining activities as well as reclamation works.

This has been added to page 4, Section “2.2 Acquisition of geoscience information”, paragraph 1, lines 156-157.

5. Lines 175-188, I suggest to elaborate more on determining weight for each landslide-controlling factor. In addition, it is also not clear that how the surface materials and roughness were determined.

The authors thank the reviewer for the valuable feedback.

The model adopts a bivariate statistical approach where the determining weight of each class factor is derived from the correlation value between each class of the landslide-controlling factor and the landslide population. The weight of the factor class is determined by calculation of the natural logarithm of landslide density of each class within each factor divided by the overall landslide density and expressed in Equation 1 [45].

The surface material in Kuala Lumpur is determined by referring to the geological map, aerial photographs, topographic map and field mapping. Surface roughness indicate the degree of variation of surface elevation derived from the average standard deviation calculated from a 10 m by 10 m moving window. A small value represents a smooth surface while a large value represents a rough surface.

This has been added to page 5, Section “2.3 Assessment of geoscience information” (Assessment of geohazard susceptibility), paragraph 1, lines 201-215.

6.  Lines 190-196, could you please clarify that why JFlow model was chosen for study? I also suggest to elaborate more on flood scenarios simulation using JFlow model, including how the model results were validated, and also flood simulation with a defended scenario for different return periods etc.

The flood susceptibility model was developed using JFlow®, a hydraulic modelling software developed by JBA Risk Management (JBA), United Kingdom. Benchmarking exercises conducted in 2012 revealed a high-performance model, producing fast and accurate representations of flow routing across flood plains. The 5 m grid model created using JFlow® and GIS was derived by using 1 m DTM to define the appropriate rainfall and river flow input with various flood return periods in the pilot study area of Kuala Lumpur [47].

The model showcased a defended scenario where it considered the Stormwater Management and Road Tunnel (SMART) tunnel system and other flood mitigation projects emplaced within the city showing a 20-, 50-, 100, and 200-year return periods. Flood susceptible areas were demarcated using two-dimensional flow paths to capture both river and pluvial floods. Retrospective validation using open source flood incidents revealed a reasonable level of confidence in the susceptible areas that were demarcated.

This explanation has been added with two relevant references in page 6, Section “2.3 Assessment of geoscience information” (Assessment of geohazard susceptibility), paragraph 2, lines 241-253.

7. Lines 238-240, please clarify that how the threshold criteria for each landslide susceptibility class were determined.

The threshold criteria for each landslide susceptibility class were obtained by extracting the landslide susceptibility value at each landslide point. A graph of cumulative percentage of the landslide value was plotted against the susceptibility value.

The threshold value was extracted from the 50th, 25th, 12.5th and 6.25th percentile (X-axis) to get the susceptibility value on the Y-axis.

This has been added to page 6, Section “2.3 Assessment of geoscience information” (Assessment of geohazard susceptibility), paragraph 1, lines 219-223.

8.      Lines 263-264, please elaborate more on "highly detailed", and how were the results validated?

The highly detailed model refers to the scale of resolution of 5 m to estimate the flood extent. Validation of the flood model was done by observing culvert data that writes out maximum and average flow rates through a culvert. In addition, monitoring lines were added to the JFlow® models based on locations where flow data and maximum water depths could be validated. Current flood events were also compared with the flood model to assess its accuracy.

This has been added to page 8, Section “3.2 Geohazard susceptibility”, paragraph 2, lines 323-329.

9. In Fig. 4, the flood susceptibility map presented in this figure is not clear or confusing. It seems that extent areas of flood susceptibility for 20 year return period is comparatively higher than that for 50, 100, and 200 year return periods.

The areal extent of flood susceptibility is the cumulative of the lower return periods, i.e., the extent of the 200-year return period actually includes the cumulative extent of 20-, 50-, 100- and 200-year return periods.

The caption for the figure has been revised for better clarity.

10. Conclusions: The major findings of this study are completely missing.

In this study, geoscience information is integrated to determine engineering ground conditions and susceptibility to multiple geohazards, which are then combined to demarcate suitable zones for urban development. The combination of both engineering geology conditions and susceptibility to multiple geohazards contributes to better understanding, for building resilience to disasters influenced by climate change. The major findings using this novel approach in Kuala Lumpur is presented, with suggested actions for building disaster resilience.

This information is clarified in the Conclusion section (page 15), lines 481-499.

Reviewer's 2 Comment

Authors' response

1. The authors should frankly state the limitations of this method. For example, during a river flood event, the high river flow can erode the river bed, which increase the risk of bank collapse, flood inundation, and even induce serious seawater intrusion (Xie et al., 2017; Xie et al., 2022). These potential problems cannot be treated by integration of geoscience information, but they does exist, especially for a city that has many rivers and is next to the Malacca strait.

The authors would like to thank the reviewer for this input.

We agree that during a river flood event, high river flow can erode the river bed, and increase the risk of bank collapse. While Kuala Lumpur is located relatively far from the coastline (about 45 km away), more research is required to determine areas that are susceptible to the cascading impacts of high river flow in the city. The requirements for further work is e.g. to determine the rainfall threshold values that would actually trigger a landslide or flood incident, etc. is now mentioned in an additional paragraph.

This has been added to page 14, Section “4 Discussion”, paragraph 4, lines 449-463.

2. In the abstract, please add one or two sentences about the scientific value of the present study for other similar systems or cities.

The abstract and conclusion have been revised to include the scientific value of the study and its novelty to the research community.

3.  It seems there are too many texts in Figures 1, 3, 5, 6. Generally, it's better to keep a figure concise. The authors can move or explain them in the context.

The authors would like to thank the reviewer for the helpful suggestion.

The figures and their legends have been modified to ensure clarity and conciseness.

4. The text in figure 5 is a repetition of the text in table 1, right?

Thank you for pointing out the repetition in our paper. The authors have revised and simplified the legend in Figure 5 so the information in Table 1 is not repeated.

Reviewer's comment

Authors’ Response

Reviewer 3

1. I agree and appreciate: “While geoscience data holds significant value, the information on ground characteristics and hazards must be transformed into an easily understood and relevant output for better understanding of non-specialists.” Information of this sort should be available beforehand and easy to understand for non-specialists who want to acquire and develop a terrain. They need to understand the implications, specifically the potential danger of not hiring experts and using cheaper solutions.

If there is a possibility to produce a guide for future developments from your research, and thus avoid associated risks and find suitable technical solutions, it seems to be a commendable idea.

The authors would like to thank the reviewer for the valuable suggestions that improve the manuscript. The suggestion on producing a guide has been incorporated in the last paragraph of Discussion section, lines 501-505, (page 16) on future work.

2. for “Areas with high suitability (UDS Class I)” do you take into account other factors that would advice against an urban development (i.e. reserve of green spaces)

The current assessment of urban development suitability primarily focuses on evaluating the ground constraints and susceptibility to geohazards. The approach does not take into account land use restrictions in the city. Although an area is classified as highly suitable for development, all future projects need to comply with local guidelines including gazetted areas for green and open spaces.

This is mentioned in the revised version of the first paragraph in Results, Section 3.3, lines 360-362 (page 11).

3. in many parts of the world the rapid urban development has triggered an inverse response, as in the urban areas are not to expand horizontally anymore and instead, the existing areas are to be densified. 

The authors agree with the reviewer’s comment. The circumstances are evident in Kuala Lumpur where 76% of the area is urbanized and land that is available for development is limited. Major infrastructure projects are already directed towards the sub-surface to cope with the scarcity of space. Redevelopment projects with higher densities are increasingly common in the city.

Section 4.Discussion, Paragraph 1, lines 431-435 (page 14), has been revised to incorporate this aspect.

4. Have you checked if the orientation for the available land in zone B is not one of the major reasons why the land is underdeveloped?

We have checked the orientation for the available land in zone B. The underdeveloped land is minimal where the largest parcel is Sg. Penchala, which is classified as Malay reserve where development is strictly regulated. The paragraph in Results, (section 3.3, Para 4, lines 400-404, page 13) has been revised to reflect this situation.

Reviewer's comments

Authors' response

1. Introduction

[lines 52-54] What is the definition of geoscience information? Three articles are referenced for this, but it would be necessary to have more peer reviewed articles to support this. Also, please look at the term "geospatial data".

The authors would like to thank the reviewer for the valuable suggestions that improve the manuscript.

Geoscience information is data and knowledge about the Earth and its systems, which is used to manage natural resources, minimize risks, and develop plans for sustainable growth, among others. Geospatial data is information with location on the Earth's surface, which allows for visualization and study of complex data using GIS to create maps and models.  Geospatial data is a sub-set of geoscience information. This is briefly mentioned in the Introduction Section (Para 3, page 2, lines 50-56), with two additional references.

[lines 62-66] Regarding the susceptibility model, it would be good to have more references other than one journal article. Another question would be: Are there any research to use susceptibility model regarding resilience study?

Three additional references have been added related to susceptibility modelling for resilience.

The Introduction section, Para 4, lines 67-76, (page 2), has been revised to include the additional references.

2. 

2.1 Study area

It would be necessary to provide the context regarding resilience instead of providing generic information about case study area. You might provide some context by asking the following questions: 1) What are the challenges of the city in perspective of resilience? 2) (If any) If the previous efforts were not sufficient to achieve resilience goals of city, what should be added or modified from methodological perspective?

Some of the challenges faced by Kuala Lumpur today in the context of resilience include:

a.     Rapid urbanization which puts a strain on the city’s infrastructure and resources, making it more vulnerable to climate-induced disasters.

b.     Kuala Lumpur is located in a region that is prone to landslides and floods. The city's geological conditions make it more vulnerable to these disasters, which can cause significant damage to property and infrastructure.

c.      Climate change is making extreme weather events more common and more severe. This is increasing the risk of climate-induced disasters in Kuala Lumpur, such as floods, heat waves, and droughts.

d.     Kuala Lumpur is not well-prepared for climate-induced disasters. The city lacks the necessary infrastructure and resources to respond to these disasters effectively.

Recent work has delineated areas susceptible to landslides and floods, but more work is required to link this to the planning process. A paragraph on section 2.1 Study Area (Para 4, lines 138-146, page 3) has been added to reflect this.

2.4 Integration of geoscience models and zonation

[lines 254-273] It seems the final results indicate that you actually utilized suitability modelling process. The susceptibility model has been used for creating one of the suitability layer. However, I don't see any other literature regarding suitability modelling and resilience.

Yes, you are right. Two maps were produced using the suitability modelling process, i.e. the Engineering ground classification (EGC) map, and the Urban development suitability (UDS) map.

While zoning of an area using suitability modelling is not new [22-24,33], many studies do not use information on hazard susceptibility, and if they do so, the focus is on a single geohazard such as seismicity, flooding and groundwater pollution [28,34-36]. 

Introduction Section (Para 4, lines 69-76, page 2) has been revised to update this information.

4. Discussion

In addition to summarizing the results, it would be necessary to answer the following questions:

What did you achieve from this research from methodological perspective? As a novel approach mentioned at an abstract, what makes this methodology more advanced approach? How about the comparison from previous research?

The resilience and prosperity of an urban area could be attained by reducing the risk of geohazards. This study has resulted in integrating geoscience information for susceptibility modelling and development suitability to support land use planning in Kuala Lumpur. The focus on two hazards i.e. landslide and flood susceptibility as well as engineering geological constraints has resulted in a new approach of delineating zones that are relatively more suitable for development. In addition, zones that are least suitable have also been demarcated. This enables the planning of relevant measures to increase the resilience of the city.

The first paragraph of the Discussion section has been improved to mention the novelty of the methodology that was used in this study.

Reviewer 5

Reviewer's comments

Authors' response

1.  Adding limitations assists future researchers be prepared if they chose to replicate the study and readers to discern our acknowledgements that our science always has limitations.

The authors are really grateful for this positive feedback and acknowledge the comments.

2. The revised conclusions provide, I believe, the potential for increased sustainability of this article's scholarship impact and its potential use to train the next generation of change agents in local communities.

The authors would like to express their utmost gratitude to the reviewer for the favorable commentary provided.

Reviewer comments

Authors’ Response

Reviewer 4

(1) Thank you for addressing my comments and revising the manuscript. One comment regarding the discussion chapter: there is still missing answer "How about the comparison from previous research?". To compare, the author should have other references, but I don't see it.

Thank you for pointing out our oversight. We have added two sentences and relevant references from the existing list. (4.0 Discussion in page 15, Paragraph 1, Lines 461-466) This has now further improved the discussion in this paragraph, as follows:

This enables the planning of targeted and hazard specific measures to increase resilience. In comparison, previous researchers have focused on single hazards [34-37]. Studies on multiple hazards are limited [24, 27, 28, 61], where conditioning parameters for the hazards are analysed collectively to produce one final suitability map, without generating susceptibility maps for each hazard; hindering targeted hazard specific mitigating measures.

Reviewer comments

Authors’ Response

Reviewer 6

(1) Describe the policy related to resilience at the current target site, and explain how the research results relate to the policy and can be linked. If relevant policies have not been established, how can policymakers utilize the findings? In addition, the authors described this guideline as non-professional, and it is questionable whether they can understand the map smoothly.

Thank you for the insightful questions that have improved the discussion.

The policy related to resilience in the current target site is the Kuala Lumpur Structure Plan (KLSP 2040). The KLSP 2040 outlines the aspiration for a well-integrated, sustainable, and resilient metropolitan Kuala Lumpur that serves everyone as the city addresses urbanisation and other challenges, including climate change and associated hazards. Key research results mentioned in Discussion Para 2 (i.e. areas where development should be minimised, areas where measures should be taken to protect urban assets) could be used for strategic development planning to enhance the resilience of Kuala Lumpur to climate hazards. The information on ground characteristics and hazards has been transformed into easily understood and relevant output in the form of zoning maps for better understanding of non-geoscience specialists.  This could be facilitated through the formulation of guidance and active engagement between geoscientists, planners and policy makers as well as other multidisciplinary professionals.

We do not use the word “professional” or “non-professional” in the manuscript. However, we acknowledge your point, and have replaced all mention of “non-specialists” with “non-geoscience specialists” and called for guidance and engagement for better understanding of the maps. (4.0 Discussion in page 15, Paragraph 3, Lines 486-496 and Paragraph 5, Lines 523-527)

(2) 2.3. Assessment of Geoscience Information: what criteria did you classify the areas shown on the map? Did the authors do it arbitrarily? Or is there any other research or objective basis? Also, what is the standard of UDS? Is it simply the author's calculation? I wonder by what criteria you divided it by.

There are five maps produced in this study, where areas are classified using distinct criteria as stated below. We thank the reviewer for the guiding questions, which has improved the write-up of the assessment approach.

In the first map, landslide susceptibility of the areas is delineated using bivariate statistical method. The areas are classified into five classes of susceptibility based on the percentile of landslide occurrences (Section 2. Materials and Methods in page 6, 2.3 Assessment of Geohazard Susceptibility, Paragraph 1, Lines 230 - 237)

In the second map, areas susceptible to flood was determined using hydraulic modelling. The areas are classified based on flood extent for 20, 50, 100 and 200 return periods (Section 2. Materials and Methods in page 6, 2.3 Assessment of Geohazard Susceptibility, Paragraph 2, Lines 256 - 264).

In the third map, the engineering ground condition was established based on expert judgement using 5 criteria; ground stability, aggregate potential, engineered fill, foundation and excavatability of the ground materials in the study area. The classification of engineering geological characteristics based on expert-judgement has been used to determine the land-suitability for urban development [28, 53]. A value of 1, 2, 3 and 4 is given to the engineering performance for groundwork indicating ranks of very poor, poor, moderate and good, respectively. The performance has been interpreted from borehole records, site investigation reports and previous studies. Reclassification of the EGC map into five classes (Class I – V) is based on the scoring matrix, where the suitability increases with score (Figure 4). (New text added to Section 2. Materials and Methods in page 5, 2.3 Assessment of Engineering Geology Condition, Paragraph 1, Lines 194 - 202)

The fourth map, Urban Development Suitability (UDS) is based on the notion that the occurrence of landslides and floods will reduce the suitable of an area for development. The criteria are based on potential occurrences of landslides and floods, and engineering ground classification. Areas deemed to have potential landslides are those of high and very high susceptibility. Flood extent defined by 200-year return periods represent zones with potential for flooding. The weightage for landslides is relatively higher than floods because the intensity of landslide events is higher than floods in Kuala Lumpur [18, 50]. A score of 1, 3, 5, and 7 is given to negligible geohazard, flood, landslide and both geohazards (Figure 6). The five EGC classes are used directly for the overlay, and given a score from 1 to 5 for Class I to Class IV, respectively.  The sum of the scores were calculated from geohazard susceptibility and EGC maps by overlay analysis technique [28,53]. A negative relationship between the sum of score and suitability was used; the smaller the score, the higher the suitability level for development. Manual reclassification is done based on the range of values assigned to the three classes (UDS Class I- High, UDS Class II- Moderate, UDS Class III- Low).

(New text added to Section 2. Materials and Methods in page 7, 2.4 Integration of Geoscience Models and Zonation, Paragraph 1, Lines 279 -  294).

In the fifth map, development suitability zones (DSZ) is a simplified UDS map where the whole city is divided into spatially contiguous zones based on the prevalent UDS classes. (see Section 2. Materials and Methods in page 7, 2.4 Integration of Geoscience Models and Zonation, Paragraph 1, Lines 294 - 300).

(3) This study did not reveal the originality of other studies. What data or methodology do you use when evaluating resilience in other studies? It needs to be supplemented by theoretical considerations.

The authors acknowledge your point. Further work has been recommended on this aspect.

Resilience is a rapidly expanding area of study and its evaluation could cover broad aspects of economy, organization, environment and society [1], to relatively narrow areas represented by indicators of social vulnerability, hazard exposure and adaptive capacity [3,8,11-15].  The purpose of this study is to highlight the role of geoscience information in supporting strategic development planning for promoting disaster resilience in Kuala Lumpur, using information on landslide and flood susceptibility as well as engineering ground conditions. The findings provide pathways for enhancing resilience in the city through various hazard-specific measures in targeted areas; i.e. disaster risk transfer and insurance schemes for ex-posed communities, social safety nets for lower-income groups, area-based business continuity plans for industries and the commercial sector, early warning alerts and public warning systems, as well as increasing public awareness and preparedness. Further work is required for evaluating the effectiveness of the approach, or indeed any other approach of other studies.

Text has been added on the need for further work on evaluating the effectiveness of the proposed approach in the last paragraph of Discussion, page 16, Lines 522-536.