Flood Vulnerability Assessment of Urban Traditional Buildings in Kuala Lumpur, Malaysia

Flood hazard is increasing in frequency and magnitude in Southeast Asia major metropolitan areas due to the effects of fast urban development and changes in climate, threatening people’s properties and life. Typically, flood management actions are mostly focused on large scale 15 defenses, such as river embankments or discharge channels or tunnels. However, these are difficult to implement in historic centres without disturbing their heritage value, and might not provide sufficient mitigation in these areas. Therefore urban heritage buildings may be particularly exposed to flood events, even when they were originally designed and built with intrinsic resilient measures, based on the local knowledge of the natural environment and its threats at the time. Their attractiveness, cultural 20 and economic values, means that they can represent a proportionally high contribution to losses of any event. Hence it is worth to pursue more localised, tailored, mitigation measures. Vulnerability assessment studies are essential to inform the feasibility and development of such strategies. In the present paper we propose a multi-level methodology to assess the flood vulnerability of residential buildings in an area of Kuala Lumpur Malaysia characterised by traditional timber housing. 25 The multi-scale flood vulnerability model is based on a wide range of parameters, covering building specific parameters, neighbourhood conditions and catchment area condition. Parameters for 163 buildings were measured in detail by a field surveys integrated with Google Street View. The vulnerability model is combined with high resolution fluvial and pluvial flood maps providing likely water depths for a range of different flood return periods. The obtained vulnerability index shows ability 30 to reflect different exposure by different building types and their relative locations. The study provides evidence that results obtained for a small district can be scaled up at city level, to inform both generic and specific protection strategies. The paper discusses these in relation to a scenario event of 0.1% Annual Exceedance Probability (AEP), based on hydrological and hydraulic models developed for the Disaster Resilient Cities Project. 35 https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c © Author(s) 2020. CC BY 4.0 License.


Introduction
The Sendai Framework 2015-2030 identifies clearly both climate change and rapid urbanisation as disaster risk drivers (UNISDR, 2015). Temperature rise and global warming are strictly correlated to increased rainfall (Wang et al., 2017;Min et al., 2011) and in turn with the increased frequency and extent of droughts and floods (IPCC, 2013(IPCC, , 2014Mysiak et al., 2016;Pall et al., 2011). Flood risk 40 however is compounded not only by intensified hazard, but very importantly also by increased exposure due to increased urbanisation along coastlines, river basins and flood plains (Kundzewicz et al., 2014;Neumann et al., 2015). Such flood risk become even more challenging in the South and Southeast Asia, as observed (Najibi and Devineni, 2018) and projected (Hirabayashi et al., 2013) flood frequency show dramatic increasing trends. 45 Following studies on the increased flood risk caused by the increasing rate of impervious surface to drainage capacity in urban areas, (e.g. Ashley et al., 2005;Jacobson, 2011;Jha et al., 2012;Liao, 2012), the shift from control to adaptation in urban flood resilience is increasingly advocated by governmental agencies, experts and developers alike. Mitigating structural measures have the objective of reducing the hazard, i.e. the runoff, by diverting it and channel it. However, structural measures are mostly 50 planned at the large scale, require large investments, long implementation periods, extensive sociopolitical negotiation. As a consequence of the long timeframe, they might turn out to be inadequate, postponed or irreversible (Aerts et al., 2014), and in many cases they prove to be unsuitable for developing countries on economic and financial grounds (Inaoka et al., 2019). Non-structural measures, however, can provide faster flood risk mitigation, yielding improved adaptability, more distributed 55 benefits and, as a result, better governance (Tullos, 2018;Roslan et al., 2019). Studies specific to Malaysia have shown that rapidly increasing flood events in recent decades are due to unrestrained occupation of rivers by human activities, destruction of forest and extreme weather events caused by climate change (Aliagha et al., 2015). Statistics show an average of 143 floods per year since 2001, of which more than 90% are flash floods (Mohd Anip and Osman, 2017). Such 60 frequently occurring floods cause a high level of threat to Malaysian citizens' personal safety and property, thereby, inflicting considerable damage to the country's infrastructure (Nasiri and Shahmohammadi-Kalalagh, 2013). Data from the UNDRR's Country Disaster and Risk Profile (PreventionWeb, 2019) shows for Malaysia that floods account for 98% of average annual loss in the period 1990 to 2014. A report from the Malaysian Department of Irrigation and Drainage, identified 65 an average of 29,000 sq.km or 9% of the country's total land area and more than 4.82 million people (22% of the population) as affected by flooding every year. The annual losses were evaluated at RM915 million (CFE-DMHA, 2019;Nasiri and Shahmohammadi-Kalalagh, 2013). At the beginning of the millennium an integrated flood management strategy was launched, whereby the Malaysian government invested in some major structural measures, along with non-structural measures and community 70 https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. participation (KTA Tenaga Sdn. Bhd., 2003). In terms of urban flood mitigation, among the structural measures, the most conspicuous intervention was certainly the SMART project, Stormwater Management and Road Tunnel, aim at alleviating the flooding problem in the City Centre of Kuala Lumpur caused by the Klang River, as well as reducing traffic congestion (Abdullah, 2004). Among the non-structural measures the government invested in flood detection and warning systems, awareness 75 campaigns and flood proofing guidelines for buildings with basement (DID, 2006(DID, , 2008. Notwithstanding this proactive approach, the "Malaysia Disaster Management Reference Handbook 2019", (CFE-DMHA, 2019) states that: "Annually, floods account for the most frequent and significant damage, with 38 damaging events in the last 20 years, and are responsible for a significant number of human lives lost, disease epidemics, property and crop damage, and other losses". The Handbook also 80 points out that risk of floods has increased due to climate change and that "Malaysia had the highest percentage of the population (67%) exposed to floods among ASEAN (Association of Southeast Asian Nations) member states between July 2012 and January 2019". With six major events in the last five years, flooding remains a major source of risk and losses in Malaysia, with a dramatic three-fold increase of population exposure in two decades. While the Malaysian government has officially adopted 85 a holistic approach to flood risk reduction from preparedness to post event relief, its implementation has received critical reviews by several researchers (Shafiai and Khalid, 2016).
Flood vulnerability, refers to the susceptibility of goods and people in any region to suffer damage and losses. An accurate assessment of such vulnerabilities is essential to devise effective flood risk management (Rehman et al., 2019). Vulnerability assessment studies, focusing on different scales 90 (Kundzewicz et al., 2019) and different dimensions (Rehman et al., 2019), have demonstrated the capability of predicting socio-economical damage and risk by floods. In an urban context, flood vulnerability assessment of individual buildings, and the management of the associated risk, has also proven to be an effective way to increase the flood resilience of the whole city (Aerts et al., 2014). Two approaches are common in flood vulnerability assessment, the physical approach and empirical 95 approach (Balica et al., 2013). Physical approaches use hydrological models to estimate the flood hazard and compute economic consequences for a particular event or area on the basis of a damage index relating a measure of intensity of the flood to the associated economic loss. Parametric approaches use a set of quantitative or qualitative indicators to rate the vulnerability of a building or area, with no particular reference to the hazard intensity. 100 In the present study, the flood risk to residential buildings in Kuala Lumpur, identified by several studies as one of the major contributors to disaster losses in Malaysia (Bhuiyan et al., 2018), is studied by adopting a hybrid approach using a hydrological model to determine the flood hazard and a set of indicators to determine the vulnerability of individual buildings. However, the present model does not compute the mechanical response of the envelop to the water pressure (Custer and Nishijima, 2015). 105 https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License.
To determine the actual risk of different types of flooding, fluvial and pluvial, the present study uses a multi-scale approach to assess the vulnerability of traditional houses in Kampung Baru (Figure 1), thus providing evidence to suggest appropriate mitigation strategies at individual building, local compound and catchment scale. The empirical vulnerability model used is particularly suitable for studies at the micro to meso scale levels, aiming at identifying effective non-structural mitigation measures. It relies 110 on a number of quantifiable and qualitative parameters which allow to identify a number of construction typologies typical of the district, with diverse vulnerability level. The local elevation around the building site and its position with respect to any river courses are also surveyed. By conducting on site and virtual surveys the parameters that influence vulnerability can be determined and quantified, and the economic losses due to flood hazards under different return period can be estimated, allowing to 115 produce mappings which identify a ranking of risk at the building and district scale, for a given hazard magnitude. The hazard magnitudes used are water depths, calculated by developing 2D hydrodynamic models to simulate the behaviour of water conveyed by overland flow and river systems in response to rainfall events of different frequencies and intensities. Finally, possible mitigation strategies to the flood risk will be proposed to increase the resilience of residential buildings in the study area, and other Asian 120 regions with the similar climate.

Study Area
The Kampung Baru district is located in the central area of Kuala Lumpur enclosed between the Klang River on the south east and the Sungai Bunus on the north-west (Figure 2 Although these houses might have been altered in time, in terms of materials and form, they still maintain two substantial characteristics related to the local environmental conditions: steep sloping roof and floor raised on stilts (Figure 2(b)). These two iconic design features protect the space within from 145 high intensity precipitation and frequent flooding, rendering these houses intrinsically resilient to Malay climate.
Examples of building on stilts in the area of study are shown in Figure 3. Earlier constructions are characterised by buildings on short timber stilts (3a). In some cases, the space below is enclosed by timber grids (3b). In wealthier construction, the stilts might have been made of stone (3c) and in modern  The maps were developed by analysing time series data from a selection of rain and river gauges across the Klang Basin to calculate design rainfall hyetographs and river hydrographs for return periods of 20, 165 50, 100 and 200-years. The design rainfall and river flows were used as input for 2D hydraulic modelling using JBA's proprietary JFlow® software (Lamb et al., 2009) to give estimated depths of inundation. The methods used to calculate the rainfall hyetographs and river hydrographs are described in section 2.2.1. An important input to the flood mapping process is a digital terrain model (DTM). For this study, a 0.5m resolution bare-earth DTM was provided by the Civil Engineering and Urban 170 Transportation Department, KL City Hall and City Planning Department, resampled to 5m resolution. JFlow can be run in different configurations for different purposes. For large rivers, a fluvial model configuration is used to apply hydrographs to the model at regularly spaced inflow points along the https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. drainage network. The volume of water that can be held within the river channel is estimated and removed from the flood simulation. A JFlow simulation is run for each return period using a solver 175 based upon the two-dimensional Shallow Water Equations. For the SMART scenario a dischargelimited directional culvert is constructed in the JFlow model, to represent the diversion and storage of flood water between Kampung Berembang and the Desa Lake at Salak South, and is adjusted for each of the four SMART operational modes as explained in Table 1. For small rivers and pluvial flooding, a direct-rainfall configuration is used. This approach applies the relevant hyetographs to each cell of the DTM. Different runoff and drainage rates are applied to reflect spatial variations in soil type and land cover. Urban drainage systems are accounted for by removing a proportion of the total rainfall volume prior to running the JFlow simulation. Water depth in metres is 185 calculated for each flood type (pluvial, fluvial, and fluvial with SMART defence) and return period (20, 50, 100-year) and recorded in a set of GeoTIFF rasters for use in Geographical Information Systems (GIS). In this study, flood maps of three flood types under 100-year return period are used in the estimation and flood hazard and risk, as is widely used in the communication and decision in flood risk prevention and management. 190

Calculation of rainfall hyetographs and river hydrographs
Rainfall totals (in mm) were calculated at 11 rain gauge stations within a 6km radius of the centre of Kuala Lumpur. This was done by extracting peak-over-threshold values from the hourly rainfall record at each gauge and fitting them to a Generalised Pareto Distribution, to enable return period rainfall totals https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. to be estimated for each gauge. This was done separately for the 1-hour, 3-hour and 24-hour storm 195 durations. Spatial interpolation was then used to convert the estimates at the gauge stations into a set of continuous rainfall surface rasters across the entire study area, providing a rainfall total (mm) for each return period and storm duration on a 110m x 110m grid. Each gridded rainfall total was converted into a hyetograph to describe the temporal distribution of the rainfall for each of the three storm durations.
Normalised rainfall profiles were developed by analysing hourly rainfall data for 20 events between 200 1997 and 2016 and calculating a mean 3-hr storm profile and a mean 24-hour storm profile across all stations. Due to the lack of sub-hourly rainfall data, the 1-hour storm profile was assumed to be a simple triangular shape. The storm profiles are illustrated in Figure 4(a) below. hydrograph shape River hydrographs were calculated at 2km intervals along river network of the study area. Each hydrograph was constructed using a linear function, defined by peak flow and time to peak estimates.
More advanced methods for deriving the shape of hydrographs are available, but in all but exceptionally flat topographies peak flow can be considered the key variable in hydrograph shape, so for this study a 210 generalised triangular profile was considered appropriate. Firstly, peak flow was calculated at 10 streamflow gauges within the Klang River basin, using non-stationary flood frequency analysis. These values were then regionalised using a linear regression equation for each return period, enabling peak flow to be estimated at all ungauged locations within the study area, based on their catchment area (in km 2 ). 215 The time to peak at each gauge was calculated by extracting the median time to peak from all discrete flood events recorded at the 7 streamflow gauges with hourly flow records available. A linear regression equation was used to estimate time to peak at all ungauged locations within the study area, which correlated time to peak (hours) to catchment area (km 2 ). Figure 4(b) shows a schematic diagram of the river hydrograph shape. 220 https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License.

Data Collection
Given the multiscale approach adopted for the assessment of the flood risk in Kampung Baru, data is obtained from multiple sources. A 3D building dataset and 0.5-meter resolution DEM dataset were buildings' typologies present and the identification of critical parameter to best target the field survey.
The field survey of Kampung Baru, was conducted in July 2018, to gather specific data relative to individual buildings. Critical parameters, difficult to identify from the GSV, such as the location and dimensions of the drainage system, were typologically classified and measured on site, along with other geometric parameters. A thorough photographic survey was also conducted at this stage, taking shots 235 for all visible and accessible elevations of sample buildings, as well as larger overview shots of the whole study area. Specific features aimed at mitigating flood damage were also observed and recorded during the field survey.
After detailed data was taken on a small sample of buildings during the field survey which also allowed for identification of buildings' typologies, a further survey based on Google Street View (GSV) was 240 undertaken to gather additional data and cover a sample of buildings in excess of 163. This procedure was successfully used by one of the authors to survey buildings to determine vulnerability and damage in post-earthquake reconnaissance (Stone et al., 2017;Stone et al., 2018), and it is increasingly used to produce exposure databases in an expedient and economic manner (Pittore et al., 2018). In GSV, a continuous series of 360-degree panoramas, created by sewing multiple overlapping photos together to 245 display the real portrayal of a specific location, were observed according to the location and the time of when the photos were captured. In Kampung Baru images were collated in three different years of survey, 2013, 2015 and 2017. In this study the latest version was chosen, and a full front sight of a target building could be accessed online through the observation points allocated on each street. During the survey, the qualitative parameters were collected visually as for the field survey. For quantifying the 250 other parameters, such as the height of door threshold and window sills, measured samples from the field survey were used as a reference to apply a measure of scale.

Vulnerability Model
Research on flood vulnerability and risk assessment encompasses a wide range of methods and focuses (Rehman et al., 2019). In an urban context a substantial component of losses is ascribable to physical 255 https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. damage to vulnerable buildings and their contents (Chen et al., 2016). Current flood risk assessment study and damage models use either an empirical approach, relying on post event damage data collection to determine vulnerability functions, or synthetic approaches, whereby the vulnerability functions are based on expert opinion. Empirical methods are basin or catchment specific (Merz et al., 2010), hence of limited transferability and applicability to other locations without substantial calibration. adjustment studies, and damage surveys carried out for past flood events. Historic data on flood damage and insured losses is not available for Kuala Lumpur or Kampung Baru. It is increasingly recognised that models need to account for multiscale, from single asset to full catchment area, and be able to consider many variables, in terms of both hazard intensity and asset response (Amadio et al., 2019).
Such models may rely on sophisticated physical modelling of the flood event, while hazard-damage 270 correlations are then determined using artificial neural networks or random forests analysis of past damage data (e.g. Carisi et al., 2018;Merz et al., 2013), or Bayesian networks (Vogel et al., 2014). For the majority of these models, however, while hazard and exposure are treated to a high level of resolution, the individual building's vulnerability descriptors are limited in number and often of a qualitative nature. Although Custer and Nishijima (2015), Herbert et al. (2018), Kelman and Spence 275 (2003) and Milanesi et al. (2018) have used mechanical approaches to determine the structural capacity of individual masonry walls to water pressure and derive vulnerability functions which correlate physical damage to depth of water, such physical models have not so far found direct application at urban scale.
In the present study, a vulnerability index approach is applied to determine the relative vulnerability of 280 individual buildings. The vulnerability index is obtained by identifying a number of parameters which are considered specifically representative of the local setting, ranging from building characteristics to surrounding conditions. The parameters used in the present study for characterising the building vulnerability are adapted from studies conducted by one of the authors on historic buildings in UK (Stephenson and D'Ayala, 2014) and the Philippines (D' Ayala et al., 2016). To these a classification of 285 drainage conditions and relative position of the building to local topography and relative position to waterways are added. The full list of parameters is illustrated in Figure 5 and Table 2. The attributes for each parameter and the rating scheme adopted are further described in the next section. https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License.

Footprint
Building Footprint area at ground floor m 2

Height of plinth
Height of the plinth relative to the road m

Height of Stilts
Stilt height over building plinth m

Height of door
Height of door threshold to the plinth m

Height of window
Height of window sill to the plinth m 7. Building fabric Structure and cladding material -

Building condition
The level of maintenance and building quality -

Drainage system
The level of drainage system around the building -

Surface condition
Type of surface around the building, surface cover, inclination and permeability -

Prevention features
The measures of flood prevention for the target building -

Vulnerability Ratings 295
For each parameter a range of attributes varying between 3 and 5 was determined through logical derivation of the maximum possible number of responses and these were assigned a vulnerability rating (VR) on a scale from 10 to 100. The scale is divided into equal, unweighted parts according to the number of attributes, with the attribute indicating lowest vulnerability assigned the value 10, and the one indicating the highest assigned the value 100, as shown in Table 2, following the PARNASSUS 300 https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License.
V.1 procedure (Stephenson and D'Ayala, 2014). For instance, the parameter 'drainage system' has three possible outcomes: 'good', 'poor' and 'no', so that the numerical rating among these three outcomes can be assigned as 10, 55 and 100, to represent the increase in vulnerability. Table 3 summarise each parameter range of attributes and its conversion into vulnerability rating. The surface condition consists of three sub-parameters and the building fabric consists of two sub-parameters. In both cases, the 305 vulnerability rating is calculated as the average ratings of the sub-parameters. 10 yes * factor that used in equation (6) Hence for each building and for each parameter a vulnerability rating , can be defined, whereby i , ranging from 1 to 163, denotes the building id, and j, ranging from 1 to 11, denotes the parameter under 310 consideration. The vulnerability index VRi for each building is therefore computed by summation of the vulnerability attribute for each parameter:
The vulnerability rating for each building can range from a minimum of 110 for lowest vulnerability to a maximum of 1100 for the highest vulnerability. To compare the cumulative frequency of each 315 parameter, a normalised vulnerability rating of each parameter nv and the total vulnerability index nVR was calculated based on Eq (2) and (3).
To further analyse the data, buildings are grouped in four classes by dividing the vulnerability range in 320 4 equal parts: Low vulnerability (0.0 to 0.25), Low-medium vulnerability (0.25 to 0.50), medium high (0.50 to 0.75) and high (0.75 to 1.00).
To determine the relative contribution of each parameter to the high and low vulnerability classes rVR was calculated based on Eq(4): where j denotes the parameter considered, k denotes the number of buildings in a given vulnerability class and i is the total number of buildings surveyed.

Economic loss
The computation of the economic losses caused by flood events includes different components, that can 330 be classified as tangible costs, including the physical damage to the building and contents, interruption of work etc.., and other intangible costs, such as loss or damage to objects with sentimental or cultural value, difficult to quantify (Kreibich et al., 2014). The economic loss model proposed in this study considers the physical damage to each building and its content as it can be estimated on the basis of its specific vulnerability (see section 2.5) and a normalised damage factor D(ℎ ) expressed as a function 335 of the flood depth. Two different damage factors Db(hi) and Dc(hi), for the building and contents, respectively, are used in the present study.
The physical damage to individual buildings can be calculated as the total replacement cost Ei where i indicates the building identifier, , , and AT are the construction cost per unit area of 340 building, the Damage factor, the Vulnerability factor and the surface area of the building directly affected by the flood, respectively. They are derived as follows.
Building cost: The replacement cost of buildings C(i) includes two parts, the replacement cost of the building ( ) and the replacement cost of contents ( ). 345 https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License.
where C 0 ( ) is the estimated construction cost in the study area depending on building type and materials, FB(i) is a value factor depending on the perceived value of the building, FH(i) is a value depending on the historic and cultural status of the building. The value factor FB can be used to account for the depreciated cost, i.e. the current remaining value, rather than the replacement value (Huizinga 350 et al., 2017). However, as several of the buildings in the study area are either historic or traditionally built, neither the depreciated cost or replacement cost might be appropriate to account for their cultural value. ARCADIS (2019) uses a range from 2415 to 4105 RM (525 to 890 €) per square meter to compute the basic construction cost C 0 ( ) of a detached house in Kuala Lumpur. This value includes the construction and services (electrics, hydraulics and mechanical) costs. In this study the building fabric 355 material (timber, masonry, concrete) is used to determine the low, medium and high cost range, while the building condition (poor, good and excellent) is used to determine the values of the adjustment factor FB = (0.4, 0.7, 1), respectively. If the building is among the ones identified as of traditional construction by Ju et al. (2012), or listed as of historic value, a factor of FH(i) = 1.3 is applied to account for the additional cultural value as a touristic attraction. 360 Replacement cost for damage suffered by contents is also a non-negligible component of the total loss suffered by building affected by floods. Huizinga et al. (2017) and FEMA (2013) assume that the replacement cost of content is typically ranging between 40 and 60% for residential properties. However, studies at the microscale (Appelbaum, 1985;Oliveri and Santoro, 2000) show that the proportion of 365 content cost to structure cost also depends on type and quality of construction, level of household income, etc. with a range from 15 to 60 %. Therefore, the content cost can be expressed as: where assumes values in the range (0.15 -0.60), which is also determined according to the building condition in this study. The Flood depth-damage ratio function D(hi), is a function of the water depth ℎ , which in this study is computed as the differential at each building site between the inundation depth computed by the flood hazard model at the road elevation and the elevation of the first floor above ground. The latter is 375 computed as the height of plinth to road + height of stilts + height of door threshold, where the height of plinth to road can be positive or negative.
Depth-damage ratio functions specific for Malaysia or Kuala Lumpur do not exist in literature, as data on losses from past events has not been systematically collected and analysed to date, notwithstanding the frequency of these, even in the last decade (Romali et al., 2018). The derivation of synthetic depth-380 damage functions relies on appropriate exposure databases, ad-hoc surveys, or heuristic information on losses. When conducting studies at micro scale, as the present one, it is important that the depth-damage ratio function used reflects the damage to single buildings, rather than aggregation at grid cell level or https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. larger, and also reflect the actual response of the construction to flood. A systematic review of several depth-damage ratio functions produced in literature (Appelbaum, 1985;Dutta et al., 2003;Englhardt et 385 al., 2019;Huizinga et al., 2017;MLIT, 2005) show the relevance of parameters such as construction material and quality, number of storeys, conditions, etc, in determining the depth-damage function, leading to a non-negligible variance among the available functions. However, as the proposed vulnerability model discussed in section 2 accounts for these characteristics explicitly in the computation of the vulnerability index VRi for each building, it is appropriate to derive a mean damage 390 ratio function, only dependent on water depth, while the variance due to the building characteristics are accounted by the Vulnerability Factor FVR (VRi) in equation (5). Figure 6 shows the damage ratio function obtained as regression from the mean values of several damage functions in literature, and the associated variance for each point in the series. The damage function obtained by regression has a coefficient of determination R 2 = 0.846 (significant at 0.01 level). 395

Vulnerability Index of selected buildings
Based on the empirical model described above, the vulnerability rating VRj for each parameter were attributed to each building and the total VRi computed. Figure 7(a) shows that most of the VRj are normally distributed except the number of storeys, roof height and footprint. As the study area is a 415 relatively small neighbourhood, the type of buildings is relatively uniform, mostly are 2-storey buildings with similar footprint. Nonetheless, the total VRi shows a Lognormal distribution (Figure 7b), with a coefficient of determination 0.997 (significant at 0.01 level). https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License.

Figure 8: Distribution of normalised vulnerability index VRi
The largest VR is 852.5, and smallest is 477.5. The distribution of the values normalised with respect to the median is shown in Figure 8, together with the cumulative density function. 425 Based on the equal quartile range of the values, the Vulnerability Index range is subdivided in 4 different categories: low, medium low, medium high, and high, as shown in Table 4. Buildings with medium low and medium high vulnerability constitute the largest portion of the total 163 buildings surveyed. The low vulnerability class includes 15% of the sample, while the high vulnerability class includes 13% of the building. The spatial distribution of the vulnerability index shows a concentration of buildings in 430 class high vulnerability on the west section of the site, while the central part of the neighbour is characterised by less vulnerable structures. Nonetheless, the results show that the buildings in the eastern part of the study area, have higher vulnerability to flooding (Figure 9).

Relevance of factors contributing to vulnerability
Given the apparent random spatial distribution of buildings in the high and low vulnerability categories, it is worth examining the relevance of the different parameters contributing to the VRi of each building, 440 so that the adverse attributes can be mitigated to reduce risk to flood hazards. For buildings in the two extreme categories of vulnerability, as per eq. 4, the average scoring of each parameter in that category is divided by the average scoring of the same parameter over the whole sample, hence highlighting the parameters that most contribute to the low or high vulnerability scoring. This is graphically shown in Figure 10, where 1 is the normalised value of the mean for each parameter over the whole sample. It is 445 shown that for the high vulnerability class, poor drainage system, buildings condition and height of base, all more than 40% larger than the average score, contribute most to high values of VRi . Conversely, good drainage system, existence of stilt on the ground to elevate the plinth height, as well as good building conditions, are key parameters in reducing the vulnerability scoring. This is a relevant finding, as commonly, for studies at mesoscale, it is assumed that parameters such as drainage and building 450 conditions can be assumed as uniform over an urban block, for instance. In relation to Kampung Baru the spatial distribution of the results demonstrates that the provision for drainage might be rather fragmented, even along the same street, in parts owing to plots redevelopments at different times.
Note that the low vulnerability buildings exhibit the lower vulnerability ratings for most parameters, except number of storey, which is closely related to the total value exposed to flood hazard. As most 455 newer buildings have relatively a higher number of storeys, they have higher vulnerability for this parameter while they perform better in relation to other parameters. https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. Figure 10: Relative values of Low/High VR buildings to the average VR for each parameter

Estimation of replacement cost due to different flood scenarios 460
To estimate the flood damage to buildings, as introduced in section 2.2, 3 different scenarios are considered: a pluvial flood, a fluvial flood without structural defences and a fluvial flood considering the effect of the SMART tunnel defence (Abdullah, 2004). For all scenarios the reference rainfall with a 100 years return period is considered here and the extent of flood water for each scenario is presented in Figure 11, together with the total losses (risk map) associated to both fluvial and pluvial events, 465 without the SMART effect. The number of buildings flooded and economic loss as a function of water depth at each building are reported in Figure 12 where the water depth is calculated as the difference between height of stilts and inundation depth, which provides a direct measure of the water depth entering the buildings.
For fluvial flood, the flooded buildings are mostly located in the west part of the study area which is 470 close to Sungai Bunus river. The maximum water depth is around 1.4 m, reducing to less than 1m with the action of SMART. For the pluvial flood, most buildings are flooded to less than 0.2 meter, and have a scattered distribution across the study area. Notwithstanding the differences in depth and spatial distribution of the three scenarios the total number of building flooded varies little, between 20% and 24% of the total building surveyed in the study area (Figure 12a). Note that buildings on the south-east 475 portion of the map, close to Klang river, are also suffering fluvial flood; however, these buildings are outside the area of the present study.
https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. the total replacement cost is calculated based on section 2.6. This amounts to around 5M RM for pluvial flood for the 163 buildings. For river floods, the total cost is considerably higher, around 15M RM without defence and 10M RM with SMART. The percentage of cost to the total replacement cost are 480 around 1.6%, 4.7%, and 3.1% for pluvial flood , river flood and river flood with SMART respectively. 11c) to estimate the total flood risk to individual building. The western part of the study area, which is located at lower topography with Sungai Bunus river passing through, was assessed to have higher risk to flood.

Discussion
Flood has become a major hazard worldwide. While major improvements in modelling flood hazard and exposure have been achieved, there is still dearth of compelling evidence on spatio-temporal 500 patterns in vulnerability of societies around the world (Jongman et al., 2015). The Southeast Asian area is more vulnerable due to the higher population density and higher frequency of rainfall. This study focusses on flood vulnerability of the buildings in a small heritage community, Kampung Baru, in the city centre of Kuala Lumpur, Malaysia. This city has experienced an increasing number of flood events due to the combined effects of observed increasing extreme rainfall referred to as Wet Wetter Dry Drier 505 pattern (Allan and Soden, 2008;Allan et al., 2010) as well as an increase of urban population, nearly doubled from 1980 to the current 1.8 million. As the trends for these two variables are not slowing or reversing, it should be expected in the future that both flood hazard and exposure in this city will continue to increase.
Buildings, being the primary shelter for people, the reduction of their vulnerability is critical in reducing 510 the risk to flood faced by population. By determining and quantifying the total value of parameters for each building, a classification of vulnerability can be generated in a spatial map to represent how vulnerable each residential building is to flood, thus providing evidence to suggest appropriate design or protection strategies specific to each building in the area. The present study has identified that higher vulnerability is related to absence or poor drainage system, poor building's conditions and poor overall 515 surrounding surface conditions. The buildings with lowest vulnerability show a combination of good drainage systems and surface condition and/or stilts at the ground floor. Thus, several possible solutions can be provided to improve the flood vulnerability of these buildings to reduce the flood risk, among which some feasible strategies are listed below: 1.
Increase the ground floor base elevation by either adding pillars or stilts at ground level in new 520 design. The raising floor on stilts is a traditional design of Malaysian vernacular buildings, common of many surveyed cases in Kampong Baru, and such design is being modernised by introduction of open car park at the bottom of high-rise building in Kuala Lumpur. This is considered as a soft measure in the Malaysian national flood prevention programme (DID, 2006). Moreover, as the maximum inundation depth due to flash flood for a 100 year return period is around 0.2m, which is less than the 525 height of most traditional stilts, the stilts are also an effective way to prevent damage from pluvial flood.
The present study shows that such strategy can effectively reduce the flood vulnerability and hence risk for individual buildings. However this solution without proper surface treatment and drainage systems may impact adversely neighbouring buildings.

2.
'Improving drainage system and surface condition'. Residential buildings which have proper 530 drainage system or vegetation or permeable surrounding ground surfaces or alternatively, set on a higher ground than the road, ensuring a downward slope from the faç ade to it, were assessed to be in the low https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. vulnerability class. These conditions are also reflected in the hazard model by varying the % of run off in each grid, although this is at a 5 m resolution. Improved drainage systems are recognised as an efficient way to improve the flood resilience of residential buildings without altering their traditional 535 status. As mentioned above, good drainage is essential for the resilience to extend from the single building to the urban block to the neighbourhood.
3. Finally, the present study also highlights the difference in flood hazard and economic loss between the two fluvial flood scenarios with and without SMART, suggesting that the operation of SMART can reduce the overall flood risk to the study area, although only marginally reducing the number of affected 540 buildings.
Hence a combination of structural measures, e.g. SMART and non-structural measures, e.g. use of stilts and proper surface treatment, appears to be the most effective way to increase the flood resilience from urban scale to building scale.
Moreover, major cities in Malaysia such as Kuala Lumpur, Penang, Petaling Jaya and Shah Alam 545 among others have been established floodplains and they are increasingly prone to floods and flashfloods as they develop (Chan, 2011). The use of structural measures is currently under consideration to address the issue of flooding associated with further urban development. The findings from the present study offer decision-makers an option of increasing building scale resilience, to make structural measures more effective. This is particularly relevant in historical cities such as Penang, where 550 traditional Malay buildings are prevalent. The combination of structural and non-structural measures is also in line with the aspirations of civil society groups that seek urban resilience within ecological systems (Connolly, 2019).

Conclusions
In this study, a localised empirical model has been built to evaluate the flood vulnerability of residential 555 buildings in Kampung Baru, Kuala Lumpur. Combining a field survey, google street view and DEM information, the data of 11 different parameters composing a building level vulnerability model, have been collected and scored to rate the flood vulnerability of a sample of 163 buildings.
The assessed multi-level parameters efficiently represent the vulnerability of residential buildings in the study area. The observed higher flood vulnerability of buildings are closely related to the drainage 560 system and surface condition, as well as the height of the base and condition of the building. Conversely, good drainage system, existence of stilt on the ground to elevate the plinth height, as well as good building conditions, are key parameters in reducing the vulnerability scoring.
A new economic loss model is developed to quantify the flood risk in terms of replacement cost, taking into account both specific vulnerability and a normalised depth-damage ratio function. The function is 565 obtained as regression from the mean values of several damage functions in literatures and is https://doi.org/10.5194/nhess-2020-96 Preprint. Discussion started: 15 April 2020 c Author(s) 2020. CC BY 4.0 License. independent of the specific building typology or local exposure model, which are accounted for in the vulnerability functions. This renders the damage function of generic value and can be applied to other situations in Malaysia and worldwide. The economic loss function considers the loss from both the physical damage to each building and its content. The additional cultural value as a touristic attraction 570 was also calculated as an additional value for the identified traditional buildings. The flood damage and economic loss were then estimated based on the economic loss model under the flood hazards from 3 different scenarios. The western part of the study area, located at a lower topography crossed by the Sungai Bunus river, was assessed to have highest risk to both fluvial and pluvial flood, suggesting that elevation is a major factor at neighbourhood scale. The findings provide multi-scale flood-resistant 575 strategies for the protection of individual residential buildings.

Acknowledgments
This study was supported with funding from the Newton Ungku Omar Fund and Innovate UK for the project entitled 'Disaster Resilient Cities: Forecasting Local Level Climate Extremes and Physical 580 Hazards for Kuala Lumpur'.

Data availability
Building data were collected from a field survey and Google Street View (https://www.google.com/maps/). Primary data are strictly used within the project "Disaster Resilient Cities: Forecasting Local Level Climate Extremes and Physical Hazards for Kuala Lumpur". The data 585 of the research findings are the available from the corresponding author (DDA) on reasonable request.

Author contributions
DDA designed the research and analysed the results; KW and YY collected the data, analysed the results and visualisation; HS, AM and VP conducted the flood modelling; JJP discussed and extended the findings. All authors discussed the results and drafted the final manuscript. 590