Enhancing Risk Analysis Capacities for Flood, Tropical Cyclone Severe Wind and Earthquake for the Greater Metro Manila Area

FOR OFFICIAL USE ONLY

Component 4 – Tropical Cyclone Severe Wind Risk Analysis

PHILIPPINE ATMOSPHERIC, GEOPHYSICAL AND ASTRONOMICAL SERVICES ADMINISTRATION GEOSCIENCE AUSTRALIA

M.A. Cecilia Monteverde1, Thelma. A. Cinco1, Flavia D. Hilario1, Cynthia P. Celebre1, Analiza Tuddao1, Emma Ares1 and W. Craig Arthur2

1. Philippine Atmospheric, Geophysical and Astronomical Services Administration 2. Geoscience Australia

FOR OFFICIAL USE ONLY

Contents

Executive Summary ...... v 1 Introduction ...... 1 1.1 Background ...... 1 1.2 Project objectives ...... 2 1.3 Study area ...... 2 2 Literature Review ...... 4 2.1 Historical Tropical Cyclone (TC) Data ...... 4 2.2 LiDAR and Elevation Data ...... 5 2.3 Land Cover Data ...... 5 2.4 Exposure Data ...... 6 2.5 Vulnerability ...... 6 3 Methods ...... 8 3.1 Regional tropical cyclone severe wind hazard for the Philippines and GMMA ...... 9 3.2 Development of site-exposure multipliers ...... 10

3.2.1 Hillshape multiplier (Mh) ...... 12 3.2.2 Land cover classification ...... 15

3.2.3 Terrain multiplier (Mz) ...... 15

3.2.4 Shielding multiplier (Ms) ...... 17 3.3 Combined multipliers ...... 18 3.4 Exposure data development ...... 21 3.5 Vulnerability models ...... 22 3.6 Severe wind risk estimation for GMMA ...... 25 3.6.1 Measures of damage ...... 25 3.6.2 Risk calculation ...... 25 3.6.3 Aggregation to barangay and municipality/city level ...... 27 4 Discussion ...... 28 4.1 Severe Wind Hazard ...... 28 4.1.1 Regional severe wind hazard ...... 30 4.1.2 Local severe wind hazard ...... 32 4.2 Estimation of damage from wind hazard ...... 42 4.2.1 Damaged Floor Area Equivalent for the 0.2% AEP (1/500) event ...... 43 4.2.2 Barangay Building Damage Cost (PHP million/km²) ...... 48 5 Conclusion ...... 55 References ...... 56 Appendix A - Taguig City Severe Wind Hazard Maps ...... 57 Appendix B - GMMA Severe Wind Hazard Maps ...... 63 Appendix C - Taguig City Damaged Floor Area Equivalent Maps ...... 69 Appendix D - GMMA Damage Floor Area Equivalent Maps ...... 75 Appendix E - Taguig City Damage Building Cost Maps ...... 80

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the iii Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix F - GMMA Damage Building Cost Maps ...... 85 Appendix G - Sample Case: Typhoon Milenyo (International Name: Xangsane - 0615) ...... 90 Appendix H - Risk calculation ...... 93 Measures of damage ...... 94 Risk calculation process ...... 94

iv Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Executive Summary

In this study, a tropical cyclone severe wind hazard modeling and impact assessment for Greater Metro Manila Area (GMMA) is presented. The Tropical Cyclone Risk Model (TCRM) developed by Geoscience Australia (GA) is used in the study to generate the regional level wind speed data across the entire Philippine Area of Responsibility (PAR) based on historical tropical cyclone (TC) record from 1951-2011. It uses statistical and parametric models to simulate tropical cyclone behavior and effects, and a statistical model to generate several thousand years of storms to determine an average recurrence interval or annual exceedance probability for given wind speeds. Statistical and probabilistic approaches using the General Extreme Value (GEV) and the Generalized Pareto Distribution (GPD) analyses are employed to model the gust wind speeds. The Regional maximum wind hazard maps for various return periods were generated for the whole Philippines using the Powell Method. The model output is combined with the analysis of the observed maximum wind speed (3-second peak gust) to adjust the various return periods (RPs) of maximum wind gust. The regional wind speed is translated into a local or site specific wind speed to account for the effects of terrain, shielding due to the building structures and topographic or hill shape factors. These factors were evaluated by using digital elevation model (DEM) and land cover classification derived from LIDAR data for GMMA using GIS and ENVI software. The wind risk assessment is a function of the interaction of the wind hazard, building exposure and the vulnerability of the building structures that will be damaged by the wind hazard. This wind risk assessment can be used to determine what might be the expected losses in terms of property damage. This will provide planners and emergency managers the estimates of the expected damage.

Results show that the eastern to north-eastern part of GMMA are considered to experience the highest wind hazard. In terms of risk, the western and the central sections of GMMA are subject to severe wind impact and have a higher risk than the other half of GMMA. Said areas are densely built-up with high proportion of vulnerable building types (makeshift, wood-type) and high proportion of old structures. The expected cost of wind damage depends on the proportion of wind damaged buildings and the cost of the building. The method can be replicated to other areas in the Philippines.

The specific objectives are the following:

1. To compile necessary tropical cyclone datasets for the project, including datasets on tropical cyclone severe wind impacts from previous cyclones; 2. To undertake modelling to determine the frequency and severity of tropical cyclones affecting the Greater Metro Manila area; 3. To develop severe wind multipliers for the GMMA; 4. To undertake tropical cyclone severe wind impact modelling in the pilot area of Taguig City; 5. To undertake severe wind impact modelling for GMMA; and 6. To generate tropical cyclone severe wind risk information for GMMA and produce maps and educational materials.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the iii Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

vi Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

1 Introduction

1.1 Background

Natural hazard events in the Philippines have caused adverse economic impact, damage to properties and loss of life. The most dominant among these events are due to hydro-meteorological hazards, such as tropical cyclones, causing extreme winds, heavy rainfall, flood/flash flood, landslides, storm surges and tornado. Stronger tropical cyclones are becoming more prevalent. The real damage brought by tropical cyclone is complicated by the vulnerability of the people who lie along its path.

The Philippines is situated in a geographical location often visited by tropical cyclones – the most frequently occurring natural hazard. Annually, about 19 to 20 tropical cyclones entered the Philippine Area of Responsibility (PAR) and about five to seven tropical cyclones are destructive and crossed the islands. According to the Office of the Civil Defense (OCD) of the National Disaster Risk Reduction and Management Council (NDRRMC), ninety-one (91) of the one hundred seventy five (175) destructive tropical cyclones that hit the country from 2004 to 2011 brought billions of pesos in damages and a number of casualties. Some tropical cyclones are very strong with maximum winds of more than 185 km/h near the center, which could damage houses and buildings and topple down power lines over a wide area. A mature cyclone may have a diameter of about 1000 km which can cover the whole archipelago of the Philippines.

The most notable tropical cyclone events in the Philippines include the occurrence of a series of tropical cyclones that caused devastating impacts due to strong winds, namely Typhoon Rosing (1995), Typhoon Milenyo (2006) and Typhoon Reming (2006), Typhoon Juan (2011), Typhoon Pablo (2012) and just recently Typhoon Labuyo, Typhoon Santi an Typhoon Yolanda (2013). Milenyo was considered as the worst tropical cyclone to impact Metro Manila since Typhoon Rosing in 2005 (in terms of severe wind). These typhoons caused severe damage to properties and loss of lives and highlighted the vulnerability of the country to hydrometeorological hazards.

Tropical cyclones are classified according to their strength and location, and are determined by the speed of the maximum sustained winds near the center. A tropical cyclone, once formed, progressively intensifies when the environmental conditions are favourable. Table 1.1 shows the classifications of tropical cyclones.

Table 1.1. Tropical cyclone classifications.

Type Range of wind speeds Tropical Depression (TD) 45 to 63 kilometers per hour (km/h). Tropical Storm (TS) 64 km/h to 118 km/h Typhoon (Ty) more than 118 km/h

Typhoons of similar intensity may become more frequent because of climate change. Studies have shown that there are substantial increases in the frequency of most intense cyclones and increases are likely to continue with rising global temperatures (Knutson et al., 2010).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 1 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

1.2 Project objectives

This report describes the results of the project titled “Enhancing Greater Metro Manila’s Institutional Capacities for Effective Disaster/Climate Risk Management towards Sustainable Development or the Risk Analysis Project”. This project was undertaken by the Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA), in collaboration with Geoscience Australia (GA) and CSCAND Agencies (comprising the Philippine Institute of Volcanology and Seismology (PHIVOLCS), Mines and GeoScience Bureau (MGB), the National Mapping and Resource information Authority (NAMRIA) and the Office of the Civil Defense (OCD)) with financial support from AusAID. The main objective of the project is to analyse the risk from severe wind due to tropical cyclone in Greater Metro Manila (GMMA) through the development of fundamental datasets and information on hazard, exposure and vulnerability. Ultimately, the outcomes of this project will lead towards strengthening the resilience of communities to the impacts of tropical cyclones.

The specific objectives of the project were the following:

1.3 Study area

The GMMA is the major center of economic activities in the country and the most densely populated regions with a land area of about 875.6 square kilometers (GMMA total land area) and a population of 4863.11 persons in 2010 (NSO) in Metro Manila. It showed an increase of 3,105 persons per square kilometer (19.137 percent) from 16,032 persons per square kilometer in 2000. But common to any urban area is the problem on rapid urbanization, a phenomenon that is affecting adversely the quality of life in GMMA.

Metro Manila lies along the flat alluvial and deltaic land extending from the mouth of the Pasig River in the west and the high rugged lands of the Marikina valley and the Sierra Madre Mountains in the east (Figure 1.1). Due to its geographical location and urban setting, Metro Manila suffers from the impacts of both natural and man-made disasters such as hydrometeorological (e.g. tropical cyclones, floods, storm surge) and geological hazards.

2 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 1.1. Map of the study area – Greater Metro Manila Area.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 3 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

2 Literature Review

In 1984, Amadore, et al. proposed the Typhoon Damage Scale II (TDS) Model for the Philippines which is a crude relationship between the surface wind speed of a typhoon and the corresponding wind damage to structures and vegetation at a certain locality.

In 1986, Kintanar et al. first attempted to synthesize climatological information on wind data into a form which would be useful to engineers and designers of building and/or low cost housing of the Philippines. The return periods of maximum wind speed at selected communities were calculated using the standard extreme value analysis.

Amadore (2011) developed an Idealized Typhoon Damaged Model (ITDM) to simulate the location, maximum winds, direction/speed of movement, radius and size/shape of a tropical cyclone and come up with a typhoon wind profile which became the basis for the degree of hazard. A vulnerability map was generated showing areas at the municipal level with the most number of nipa houses. The typhoon risk mapping model was limited to wind damage risk, areas on flat/ocean surface and residential structures only.

Arthur et al. (2011; Arthur, in prep) developed the Tropical Cyclone Risk Model, a statistical-parametric model to simulate tropical cyclone behavior and effects, to generate several thousand years of storms to determine an average recurrence interval or annual exceedance probability for given wind speeds. It can also determine the individual wind swath of tropical cyclone wind events.

2.1 Historical Tropical Cyclone (TC) Data

Historical datasets on tropical cyclones that formed along the western north Pacific regions and dataset that entered the Philippine Area of Responsibility (PAR) covering the period 1951 to 2011 were considered. The TC track data used in the study were taken from Japan Meteorological Agency (JMA1), which include the 6-hourly position (latitude, longitude) of tropical cyclones, intensity data, central pressure, maximum wind, and other relevant information.

There are a total of 109 tropical cyclones that crossed Metro Manila for the last 63 years (1948-2011) (Figure 2.1). The impacts of previous tropical cyclones were also gathered that served as an information source for verifying the results of the risk analysis.

1. 1 http://www.jma.go.jp/jma/jma-eng/jma-center/rsmc-hp-pub-eg/besttrack.html

4 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 2.1. Tropical cyclone tracks that passed within 100 km of the Greater Metro Manila Area for the period 1948 to 2011. Data from JMA Best Track Archive.

2.2 LiDAR and Elevation Data

The LiDAR Digital Elevation Model (DEM) for Metro Manila: This is a LiDAR-derived ‘bare-earth’ elevation dataset, with a pixel size of 1 m2, which was produced by FUGRO from the raw LiDAR point data. The latter was flown in March 2011, and has a vertical accuracy of ± 15cm (1 standard deviation) in bare earth areas, and a point density of approximately 4 points per square metre. Spatially the LiDAR DEM covers all areas of the Pasig Marikina Catchment included in the hydraulic model in this study. Although considered accurate in terrestrial areas, the LiDAR DEM is not accurate in ‘water areas’ such as rivers, lakes and ponds, which were underwater at the time of data capture. It was used as the main source of elevation data in this study. The 1-m horizontal resolution DEM was resampled to a 20-m resolution DEM for the purpose of deriving the land cover (see below) and the topographic multipliers (see Section 3.2.1).

2.3 Land Cover Data

The land cover was classified for GMMA using a combination of the LiDAR-derived DEM, DSM and multispectral aerial photography. Vegetation areas were identified using NDVI (normalized difference

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 5 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

vegetation index), derived from the multispectral aerial photography. The 20-m resolution DEM and DSM were used to determine the height of features (both vegetation and built environment) over the landscape. This resulted in a classification that distinguished the following features: i) high-rise city centers; ii) forests; iii) high density metropolitan and/or industrial areas; iv) residential subdivisions; v) informal settlements; vi) few trees with long grass; vii) crops; viii) rough open water surfaces, uncut grass, airfields; and ix) cut grass, rice paddies and enclosed waters.

2.4 Exposure Data

The exposure data refers to exposed ‘elements at risk’ and describes the statistical properties of the building stock (people, building type, age and construction types). The building inventory database was developed by the Philippine Institute of Volcanology and Seismology (PHIVOLCS) and Geoscience Australia. The dataset was developed by combining spatial datasets representing the settlement pattern of the Greater Metro Manila Area with statistical information on buildings and population. The statistical information was derived from data provided byte Assessor’s Office from LGUs, from data developed in projects such as the Metro Manila Earthquake Impact Reduction Study (MMEIRS) and the structural information collection in the Census of Population and Housing, which is conducted by the National Statistical Office (NSO). Digital and analog maps (50K and 250K scale maps) and aerial photographs are available at the National Mapping and Resource Information Agency (NAMRIA) and the LIDAR data were also used to derive the exposure information.

Each exposure polygon contains data describing a number of physical attributes of the land within the polygon. These attributes include the land-use category (e.g. residential, commercial); fraction of the land area covered by the buildings; total floor area of the buildings; total footprint area (i.e. ground floor area) of the buildings; total floor area of different building types (W1 for wooden1, W2 for wooden 2,N for makeshift\informal, CHB for concrete hollow block, CWS for concrete with steel, C1 for concrete 1, C2 for concrete 2, S1 for steel 1 and S2 for steel 2); total floor area inside buildings with different number of storeys ( 1-storey, 2-storey, 3-8 storey, 9-16 storey, 16-25 storey); and information on a replacement costs of various building types. The May 2013 version of the exposure dataset was used for this study.

2.5 Vulnerability

Vulnerability models are used to estimate how much damage occurs to a given building type as a result of wind impacts. They take the form of wind speed-damage curves. Several curves have been developed by engineering experts, each of which is suitable for a different building type, building age and storey category (Figure 2.2).

6 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

1 0.9 W1-L W3-L 0.8 N-L 0.7

MWS-L-W 0.6 MWS-L-S 0.5 CHB-L-W 0.4 CWS-L-W

0.3 CWS-L-S Damage Damage fraction C1-L-W 0.2 C1-L-S 0.1 C1-M 0 S1-L 0 50 100 150 200 250 300 350 S1-M Wind speed (km/h)

Figure 2.2. Wind speed-damage curves for different building types developed by UPD-ICE for the present study.

The wind speed-damage curves describe the damage fraction (i.e. ratio of damage to building construction cost) as a function of the peak wind speed, for a range of different building types. For example, a damage fraction of 0.5 means that the cost of repairing the building is equal to 50% of its construction cost. This defines the first measure of damage used for the wind risk in this study.

A complication is that the building-type categories for these vulnerability curves do not exactly match the building-type categories in the exposure polygons. To overcome this issue, a simple mapping between the building types was developed (see Table 5 at the end of this document).

For the development of basic vulnerability curves for the key building types with respect to tropical cyclone severe wind, the University of the Philippines (UP) through the Institute of Civil Engineering (ICE) develops a basic suite of vulnerability curves of key building types in the GMMA which are used in risk assessment of impacts of tropical cyclone severe wind. Vulnerability curves for the thirteen (13) building types were developed for the study.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 7 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

3 Methods

The wind risk assessment is a function of the interaction of the wind hazard, building exposure and the vulnerability of the building structures that will be impacted by the wind hazard. The wind risk assessment can be used to determine what might be the expected losses in terms of property damage and the corresponding damage cost due to wind hazard.

To evaluate the risk associated with typhoons, a four-stage process was set out, with information progressively refined through the process. The first stage is to evaluate the regional scale hazard from typhoons and tropical storms, based on the intensity, frequency and tracks of typhoons that have been observed to impact GMMA. This stage entails an analysis of historical typhoons, and then simulation of a catalogue of a large number of typhoons that have similar characteristics. This process leads to the development of so-called regional wind hazard estimates. The wind hazard estimated through this process are considered regional, as the modelling is performed are at a resolution of approximately 2 km, and represent wind speeds over a flat, featureless landscape.

Wind hazard is represented as a return period wind speed – the likely wind speed to be exceeded, on average, once within a given period in time. For example, the wind hazard is described as a 1-in-100 year wind speed. This does not mean that the corresponding wind speed will be exceeded only once in any 100-year period. There is about a 63% chance of the 100-year wind speed being exceeded once or more, and a 37% chance of that wind speed will not be exceeded over a 100-year period.

The regional wind hazard estimates are compared to return period wind speeds calculated from observations at weather stations across GMMA and surrounding areas, to ensure the modelled wind hazard displays sub-regional variations (e.g. proximity to open waters). Site-specific effects (such as the effect of hills and ridges, vegetation and built-up areas) are incorporated through use of site- exposure multipliers. These site-exposure multipliers are developed using the high-resolution digital elevation and digital surface models in conjunction with multispectral aerial photography captured as part of the project.

The local wind hazard estimates are used to inform damage calculations for GMMA, where the level of hazard in each land parcel defines the expected level of damage to different building types. The level of damage within a land parcel depends on the level of hazard, the proportion of different building types in the land parcel, and the susceptibility of those building types to damage from severe winds. Within GMMA, land parcels contain a range of building types, so a weighted sum of damage, proportional to the floor area of each building type, is used to represent the damage for each land parcel.

The losses caused by severe wind events can be reported in a number of ways, each of which highlights different aspects of the vulnerability of the community. For example, reporting losses in financial terms (Philippine Peso per square kilometre (PHP/km2)) highlights areas where the building stock has greater monetary value, while the metric of damaged floor area per square kilometre (ha/km2) generally draws attention to those areas with more vulnerable building stock.

8 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

3.1 Regional tropical cyclone severe wind hazard for the Philippines and GMMA

The first stage in assessing the risk due to tropical cyclones is to determine the level of hazard. Severe winds are described in terms of likelihood and the strength of the wind. Wind hazard is the 3-second peak wind gust measured at 10-meter height over open terrain. Historical records are used to determine the probability of occurrence for the different wind speeds. Most frequent maximum wind speeds are most likely in places where tropical cyclones are most frequent. The Tropical Cyclone Risk Model (TCRM) was used to assess the wind hazard. This was developed by Geoscience Australia (GA) for calculating the wind hazard from tropical cyclone (Arthur et al., 2008a, Arthur et al., 2011; Arthur; in prep, Summons, 2012).

Severe wind hazard describes the likelihood of extreme wind speeds occurring over a long period of time. For tropical cyclones, the record of observed typhoon events in the Philippines contains only 60 years of events. During this time, it is unlikely that all areas of the Philippines have experienced extreme winds, and certainly not the most extreme possible in each area. For assessing risk, it is necessary to estimate the wind speeds associated with much rarer events (e.g. events that may occur only once in 100 or 250 years). To overcome the lack of observations of extreme winds, the Tropical Cyclone Risk Model2 (TCRM) developed by Geoscience Australia (GA), is used in the study to generate the regional level wind speed across the entire Philippine Area of Responsibility (PAR) based on historical TC record from 1951-2011.

Using the historical TC record, TCRM builds statistics on the formation location, track, intensity and frequency of TCs across the Philippines. These statistics are in turn used to develop a synthetic catalogue of TCs representing thousands of years of TC activity, including credible, but as yet unobserved, very intense TCs. This overcomes the comparatively short record of observed TCs and allows a better estimation of wind speeds with return periods greater than the length of the observed record.

TCRM uses an autoregressive statistical model to generate several thousand years of synthetic TC tracks to determine an average recurrence interval or annual exceedance probability for given wind speeds. The forward motion (speed and direction) and intensity of the TC is related to the previous behaviour of the TC. For this study, 5,000 simulations were produced which generated 97,500 synthetic tracks across the Philippines.

A parametric wind field is applied to each track, providing a spatial representation of the wind speed from that track. The parametric wind field used the Powell et al. (2005) radial profile and the Kepert (2001) boundary layer model to account for surface friction and wind field asymmetry (i.e. stronger winds on the right side of the cyclone).

As a final step, the wind speeds from all events are fitted to the Generalized Extreme Value (GEV) distribution to determine an average recurrence interval or annual exceedence probability for a given wind speed. By randomly sampling from the synthetic events, an estimate of the statistical confidence of the GEV fitting was also generated, providing an indication of the potential range of return period wind speeds.

2 http://github.com/GeoscienceAustralia/tcrm

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 9 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.1. Process flow for information in developing regional wind hazard.

The model output is combined with the analysis of the observed maximum wind speed (3-second peak gust) to adjust the various return periods (RPs) of maximum wind gust, which are determined using a Generalized Pareto Distribution (GPD).

Figure 3.1 shows the schematic diagram in assessing the regional hazard levels. The wind hazard was derived using the historical TC track dataset that entered the Philippine Area of Responsibility (PAR) from 1951 to 2011 period. Return period wind fields are calculated by taking an average over a very long period of time and extrapolated to estimate the likely distribution of wind speed across the Philippines. The model derives the Regional wind speed and utilizes distribution of TCs properties (speed, bearing, intensity) developed on a grid covering the Philippine Area of Responsibility.

The regional severe wind hazard maps can be used to update the wind zoning map of the Philippines and can be considered for building design and as a guide for emergency managers and planners for evacuation planning. The wind hazard estimated by TCRM represents the regional (3-second) gust wind speed over flat terrain. This does not account local effects such as terrain roughness, topography and shielding from neighbouring structures. To assess the wind hazard at the local, the effect of various factors like topography, shielding of the surrounding buildings and terrain roughness must be considered.

3.2 Development of site-exposure multipliers

The impact of severe wind varies considerably between structures at various locations, due to the geographic terrain, the height of the structure concerned, the surrounding structures and topographic

10 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

factors. In order to accurately estimate damage to buildings from severe winds, an understanding of wind speeds at the site of the building is required. The wind speeds described in the previous section are still indicative at a regional resolution only. There are a number of factors that need to be considered to determine the local wind speed within the area. The regional wind speed needs to be modified to reflect the effects of local land cover (e.g. forests, high-rise buildings or water bodies), the shielding effect due to upwind structures and topographic effects. This is done using so-called site-exposure multipliers (Lin, X. G. and K. Nadimpalli, 2005, Cechet et al. 2009).

Site-exposure multipliers relate the regional wind speed to local site conditions. It allows for the increase in wind speed with height on the various terrain (land-use) types, decrease in wind speed as influenced by the surrounding buildings at the site and the increase in wind speed for the shape and slope of the ground contours in undulating terrain. There are four multipliers, the terrain (roughness) multiplier (Mz), the shielding multiplier (Ms) and the topographic (hillshape) multiplier (Mh). Site- exposure multipliers are calculated for each of eight cardinal directions (north, northeast, east, southeast, south, southwest, west and northwest).

Figure 3.2 provides an overview of how the site-exposure multipliers are generated. The land-cover classification was developed from remotely sensed data using LiDAR and aerial photography datasets for GMMA. The other derived datasets are prepared from DEM (slope and aspect raster layers). The land cover values are reclassified into base values to generate the terrain (Mz) and the shielding (Ms) multipliers. The directional convolution filters were applied to the base values of Mz and Ms. The Ms directional data are then combined with the slope and aspect data to account for the decrease in shielding provided by the downslope structures. To calculate the local site wind multipliers (Vsite), the following equation is used:

Local Site ind Speed ( ) where:

Vsite = Maximum local (site-specific) wind field

VR = Regional wind speed in open terrain at 10 m height

Mz = Terrain multiplier (effects of land-use)

Ms = Shielding multiplier (effects of shielding from surrounding building)

Mh = Topographic multiplier (effects of hill shape, topography shape)

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 11 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.2. Schematic diagram for the generation of site exposure multipliers.

3.2.1 Hillshape multiplier (Mh)

The hillshape multiplier (Mh) represents the effects within a local topographic zone of accelerations caused by wind flowing over topography such as hills, rides and escarpments (Figure 3.3). The topography of Metro Manila is generally flat in the urban central district where there is no increase in the wind speed while the elevated areas in the north-eastern part of GMMA particularly in Montalban, San Mateo, Antipolo, Tanay and Angono experienced a 20 to 40% increase in wind speed. There is an increase in the wind speed as elevation goes higher up the mountains or at higher elevated areas and no change in the wind speed is experienced over flat areas. For the GMMA, Mh was derived from a 20-meter resolution resampling of the LiDAR-derived DEM. (Figure 3.4).

For slopes of less than 5%, Mh is set to a value of 1; for slopes greater than 45% Mh is set to 1.71.

Thus, across much of the relatively flat topography of GMMA, Mh is close to or equal to 1. Only in the elevated areas in the north-eastern part of the GMMA are there high values of Mh.

12 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.3. Development of a vertical wind profile over a hill (developed from Figure 4.2 of AS/NZS 1170.2). Mh gives the magnitude of the difference U.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 13 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.4. 20 meter resampled digital elevation model for GMMA. Shuttle Radar Topography Mission (SRTM) data shown in areas outside of GMMA Study region (from http://www2.jpl.nasa.gov/srtm/).

14 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

3.2.2 Land cover classification

Land cover classification was performed using the LiDAR-derived DEM, DSM and accompanying multi-spectral aerial photography. These three datasets were manipulated to identify areas of the landscape falling into one of nine different classes (Table 3.1).

Firstly, vegetation and non-vegetation areas were identified using NDVI values determined from the multispectral aerial imagery. The availability of a near-infra-red channel in the aerial imagery was crucial in permitting a high-resolution NDVI image, allowing the 20-meter resolution of the resulting land cover map.

Following identification of vegetated and non-vegetated regions, the height of land cover above the ‘bare earth’ surface (i.e. ground level) was determined using the difference between the DSM and DEM. This ‘land cover height model’ was then used to classify areas into one of 9 classes (Table 3.1). The land cover types are based on classifications detailed in AS/NZS 1170.2 (2011), adapted to Philippine conditions (e.g. inclusion of informal settlements).

Table 3.1. Terrain (Mz) and shielding (Ms) multiplier values for each land cover category.

Class Description Terrain Roughness Terrain Interpolated Shielding Category Length (m) Multiplier Mz Multiplier 1 (Mz) (Ms)

1 High-rise city centers 4.2 4 0.8 0.66 0.7 3 Forests 3.77 1 0.8 0.70 1 4 High density metropolitan, 3.65 0.8 0.8 0.71 0.7 industrial areas 6 Residential subdivisions 3 0.2 0.8 0.77 0.85 7 Informal settlements 2.75 0.12 0.8 0.79 0.9 8 Few trees with long grass 2.5 0.06 0.8 0.82 1 9 Crops 2.25 0.04 0.875 0.84 1 10 Rough open water surfaces, 2 0.02 0.9 0.89 1 uncut grass, airfields 11 Cut grass, rice paddies, 1.6 0.008 0.9 0.96 1 enclosed waters 1. (AS/NZS 1170.2)

3.2.3 Terrain multiplier (Mz)

In a complex landscape such as that of Metro Manila, the site-specific wind speed is affected by the aerodynamic roughness of the landscape due to the presence of different land cover elements – vegetation, water bodies, residential buildings or high-rise buildings all influence the surface wind speeds to greater or lesser extents. As the regional wind speeds are calculated with an assumed constant surface roughness, we must use the terrain roughness (Mz) multiplier to relate that constant roughness to the actual surface roughness across the landscape. The terrain multiplier values are based on the land cover classification map described in Section 3.2.2, with the values derived from those specified in AS/NZS 1170.2 (2011). Figure 3.5 presents the workflow for determination of Mz values for eight directions.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 15 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.5. Workflow for determination of the terrain multiplier.

As the values described in AS/NZS 1170.2 are for use in design specifications, there is a level of conservatism included in the listed values. Physically, increasing roughness length would continue to reduce nominal wind speeds at 10-meter elevation, but design standards specify a minimum value for the reduction of nominal wind speeds (note the Mz values of 0.8 for all categories with roughness length > 0.05 m). For risk assessment purposes, this would result in an overestimate of the local wind speed, leading to an overestimate of the damage level. For the GMMA, the relationship between roughness length and Mz values was mapped and a modified value was estimated by fitting a logarithmic function to values of Mz greater than 0.8. This fitted function was then extrapolated to higher roughness lengths to estimate an Mz value for these lengths (Figure 3.6).

1.2 For T< 2.5: 1.1 Mz = -0.2827Ln(T) + 1.0762 R2 = 0.9455

0.9z M 0.8 0.7 0.6 0.5 1 1.5 2 2.5 3 3.5 4 Terrain category

Figure 3.6 Relationship between terrain category (based on roughness length) and terrain multiplier (Mz). Blue points are the values specified in AS/NZS 1170.2 (2011). Red line is a fitted curve to terrain categories 2.5 or lower and represents values used in this study.

16 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.7 Land cover classification for GMMA, derived from aerial photography and LiDAR-derived DEM and DSM.

3.2.4 Shielding multiplier (Ms)

Shielding represents the protection afforded to a downwind building by another building. Shielding only affects the built environment; forests do not produce any shielding. It depends on the number of buildings upwind in a shielding zone with at least the same height as the structure concerned. The built-up areas have higher shielding factor and thus a reduction in wind speed. The taller building shields the lower building which tends to lower the wind speed towards the lower building.

To generate the shielding multipliers, three components are required, the calculation of aspect, slope percent and the shielding factors for the 8 cardinal directions and considering only the built-up area (city buildings). The land-cover values were reclassified into base values for the terrain multipliers and shielding (Ms) multipliers. The directional convolution filters was applied to the base values for Ms. The

Ms directional data with the slope and aspect data are combined to account for the reduced shielding provided by downslope structures for each of the eight directions. Figure 3.9 shows the datasets used for determining Ms.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 17 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.8 Workflow for determination of the shielding multiplier.

Figure 3.9. The LiDAR-derived aspect and slope data within GMMA.

3.3 Combined multipliers

To assess the local site wind speed, the regional wind hazard was combined with the local site multipliers by multiplying the three wind multipliers for the eight (8) cardinal directions (20 m x 20 m resolution) (Figure 3.10). Consider the highest/maximum values of all the multipliers (topographic, terrain, shielding and directional) as the final multiplier.

18 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Legend

Multiplier Values < 0.8 0.8 - 0.99 0.99 - 1.01 1.01 - 1.2 1.2 - 1.4 1.4 - 1.6 > 1.6

Figure 3.10. Combined multiplier value. Presented here are the maximum values from all eight directions considered.

The multipliers relate regional wind speed to local site conditions: a decrease in wind speed with height for various terrain (land cover) types; decrease as influenced by the building upwind and an

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 19 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

increase for the shape and slope of the ground in undulating topography. The local wind speed is presented at various annual exceedance probabilities (AEPs: for 5%, 2%, 1%, 0.5% and 0.2%) for a given wind speed or at various return periods (20, 50, 100, 200 and 500-yr RP) (Figure 3.11).

Figure 3.11. Local Site Severe Wind Hazard Map.

20 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

3.4 Exposure data development

The exposure is specified as the statistical properties of the building stock (people, building type, age and construction types) in each ‘exposure polygon’. The exposure polygon boundaries are defined manually, based on land-use considerations (Figure 3.12).

Each exposure polygon contains data describing a number of physical attributes of the land within the polygon, including:

1. The land use category (e.g. Residential, Major Commercial, etc.) 2. The total land area of the polygon (m²); 3. The total floor area of the buildings in the polygon (m²); 4. The total footprint area (i.e. ground floor area) of the buildings (m²); 5. The total floor area of different building types (m²). The building categories are developed by engineering experts. They include: W1 (Wooden 1), W2 (Wooden 2), N (Makeshift/Informal), CHB (Concrete Hollow Block), CWS (Concrete with Steel), C1 (Concrete 1), C2 (Concrete 2), S1 (Steel 1), S2 (Steel 2) and others; 6. The total floor area inside buildings with different numbers of storeys (m²). The numbers-of- storeys are split into categories including: 1-storey, 2-storey, 3-8 storey, 9-16 storey, 16-25 storey; 7. The total floor area in each combination of building type and storey category (m²). For example, the floor area in 1-storey wooden buildings, in 1 storey reinforced concrete buildings, in 2 storey reinforced concrete buildings, etc. 8. The Era of Construction (e.g. Pre-1972, 1972-1992 or Post-1992). This helps to define the construction type for certain building types.

There are several floor area attributes that are provided in the exposure database, in the risk calculation methods/procedures. Information on the construction cost (PHP/m²) of various building types was estimated from taxation data, as well as engineering guides on building construction costs in Manila. It varies depending on the land-use of the exposure polygon. Further details on exposure data can be found in the report for Component 2 - Exposure Information Development.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 21 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 3.12. Sample exposure polygon boundaries based on land-use considerations.

3.5 Vulnerability models

Vulnerability models are used to estimate how much damage occurs to a given building type as a result of wind impacts. They take the form of wind speed-damage curves. Several curves have been developed by UPD-ICE, each of which is suitable for a different building type, building age and storey category (Figure 3.13) and Table 3.2 lists the key building types identified for Severe Wind.

For the project, there are 17 building types or sub-types are identified for wind; the sub-types are grouped into wood, masonry, concrete, steel, and special structures (Pacheco, et al., 2013 - Table 3.2).

Vulnerability is defined as the damage ratio in a building type at a specified intensity or measure of the hazard (3-second gust wind speed). Damage ratio basically compares the cost of repairs to the total cost of the building. Historical damage data from PAGASA were useful for deriving vulnerability curves using the empirical method.

The other two approaches used are: computational method, and heuristic method. In the former, computational fluid dynamics was used for simulating wind flow and wind pressure effects, In the heuristic method, surveys of experts and specialists were conducted in writing preceded by an explanation of the project objectives to the respondents.

22 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

1 0.9 W1-L W3-L 0.8 N-L 0.7

MWS-L-W 0.6 MWS-L-S 0.5 CHB-L-W 0.4 CWS-L-W

0.3 CWS-L-S Damage Damage fraction C1-L-W 0.2 C1-L-S 0.1 C1-M 0 S1-L 0 50 100 150 200 250 300 350 S1-M Wind speed (km/h)

Figure 3.13. Wind speed-damage curves for the different building types developed by UPD-ICE for the study.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 23 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table 3.2. Key Building Types Identified for Severe Wind, Flood and Earthquake.

Building Building Structural type or Building type or Building type or Building type or group type description sub-type code sub-type code sub-type code (for wind) (for flood) (for earthquake) WOOD W1 Woo, light frame W1-L W1-L-1 W1-L

W1-L-2

W3 Bamboo W3-L W3-L W3-L

N Makeshift N-L N-L-1 N-L

N-L-2

MASONRY MWS Concrete hollow blocks MWS-L-W MWS-L MWS-L with wood or light material MWS-L-S

CHB Concrete hollow blocks CHB-L-W CHB-L-1 CHB-L

CHB-L-S CHB-L-2

URA Adobe - - URA-L URM Unreinforced masonry - - URM-L bearing walls CONCRETE CWS Reinforced concrete CWS-L-W CWS-L CWS-L moment frames with wood or light metal CWS-L-S

C1 Reinforced concrete C1-L-W C1-L-1 C1-L moment frames C1-L-S C1-L-2

C1-M C1-M

C4 Concrete shear walls and - - C4-M frames C4-H

PC2 Pre-cast frame - - PC2-L PC2-M STEEL S1 Steel moment frame S1-L S1-L-1 S1-L S1-L-2 S1-M S1-M S1-M

S3 Light metal frame S3-L - S3-L S4 Steel frame with cast-in- - - S4-M place concrete shear walls SPECIAL BB Billboards BB STRUCTURES PT Power transmission towers PT

Total types or sub-types for hazard component 17 15 18

24 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

3.6 Severe wind risk estimation for GMMA

The risk calculations allow us to estimate the damages that tropical cyclone could cause to buildings. The calculation methods require 3 key inputs: the wind hazard model, the exposure data, and a vulnerability model.

A large number of events were simulated for the project to calculate the probability of being exceeded in any one year, termed the ‘Annual Exceedance Probability’ (AEP). The project calculated AEP wind speeds ranging from 5% (on average every 25 years) through to 0.2% (occurring on average about once in 500 years).

In practice, a damage calculation is done for a range of AEPs that allows one to assess the damages from both relatively common typhoon impacts and from more extreme events. Figure 3.14 shows the step-by-step process for calculating the measures of damage.

3.6.1 Measures of damage

Several measures of damage are used for the study:

1. Damage fraction: the ratio of the cost of repairs to the total construction cost. This can apply to individual building types, a group of buildings (such as in one exposure polygon) or across an area such as a barangay or city 2. Damaged floor area equivalent: This is equal to the total floor area (m2) of a given building type multiplied by the damage fraction for that building type, divided by the total land area of the polygon. Intuitively, this measures the intensity of physical damage to buildings, irrespective of their monetary value. 3. Building damage cost: This is equal to the damaged floor area equivalent multiplied by the associated building construction cost. Intuitively, this measures the physical damage to buildings in terms of the cost of construction. For this study, building damage cost is reported in units of millions of Philippine Peso per square kilometre (million PHP/km2).

Providing several measures of damage enables different users to understand and interpret the results of the risk analysis for their specific needs.

3.6.2 Risk calculation

Following is an outline of the process to calculate the loss and resulting severe wind risk to a single land use polygon. This process was applied to all land use polygons containing structures in GMMA. The results were then aggregated to barangay and LGU level for reporting (see Section 3.6.3 for the aggregation method).

1. Calculate a representative wind speed for the polygon, usually the mean wind speed across the polygon; 2. Define the combinations of building type and storey categories from those available in the exposure databases. In the exposure database, these are populated with an estimated floor area for each combination. 3. Using the building type, land use category (as defined in the exposure database) and the building cost database, calculate the value of each building type/storey category class within

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 25 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

the polygon. The building values for each combination are added to obtain the total value of built assets in the polygon. For each building type/storey category combination: (Building value (PHP)) = (Floor area (m2)) x (Construction cost (PHP/m2)). 4. For each building type/storey category combination, use the representative wind speed and the corresponding vulnerability model to calculate the damage fraction for each combination. 5. For each building type/storey category combination, calculate the cost of damage:

(Cost of damage (PHP)) = (damage fraction) x (Building construction cost (PHP/m2)) x

(Building floor area (m2))

6. The cost of damage from all building type/storey category classes is summed to obtain the total cost of damage in the polygon. 7. The damage fraction for the polygon is calculated as the total cost of damage (step 6) divided by the total value of built assets in the polygon (step 3). 8. For each building type/storey category class: (Damaged floor area equivalent (ha/km2)) = (Damage fraction) x (floor area (m2)) x 100 / (Total land area of polygon (m2)). This gives the damaged floor area equivalent in units of hectares/km2 (the factor of 100 converts from m2/m2 to ha/km2). 9. Add the damaged floor area equivalent values for all building type/storey category classes to obtain the total damaged floor area equivalent for the polygon.

A worked example is given in Appendix H.

Figure 3.14. Schematic diagram of risk calculation.

26 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

3.6.3 Aggregation to barangay and municipality/city level

The loss calculations above are performed on each individual land-use polygon across GMMA. However, it is more appropriate to report the losses at some aggregated level, such as barangay, municipality or city. For aggregating the measures of damage to barangay, municipality or city level (regions), a number of methods are used.

1. In the case of building damage costs, the cost from all exposure polygons within the region is summed to obtain the aggregated damage costs. 2. For damage the total building damage cost for the region is calculated (as above), then divided by the total building value in the region:

(Damage fraction for region) = (Total building damage costs (PHP)) /

(Total value of buildings in region (PHP))

3. The damaged floor area equivalent for the region is based on the damage fraction for the region (calculated above), and the ratio of building floor area to land area for the selected region:

(Damaged floor area equivalent (ha/km2)) = (Damage fraction for region) x

(Floor area of buildings within region (m2)) x 100 /

(Total land area of region (m2))

These aggregations are independent of the region chosen.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 27 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

4 Discussion

4.1 Severe ind Hazard

This chapter discusses the results of the various components of the wind risk assessment, described in Chapter 4.

For this project, the Japanese Meteorological Agency’s best track archive was used as input to TCRM.

Genesis of Tropical Cyclone Formation

Tropical cyclone formation is mostly developed in the eastern part of the Philippines (7.5°N to 15°N and 128°E to 138°E) or in the Western North Pacific Ocean. About 50.2% of the TC developed inside the Philippine Area of Responsibility (PAR), 49.8% outside the PAR and 7.2% developed in the West Philippine Sea (Cinco et al., 2006). Tropical Cyclone Risk Model generated the genesis of TC formation and it shows in good agreement with the actual TC formation in the PAR (Figure 4.1).

28 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.1. Formation location of tropical cyclones that entered the Philippine Area of the Philippines (PAR) from 1948 to 2006. Contours are the probability of genesis based on all events in the JMA best track archive, as evaluated by TCRM.

Mostly, the tropical cyclones originate from the western north Pacific and normally approach from the east and recurve to north northwest (NNW) to northwest (NW) direction as shown in Figure 4.2. The southern most parts of the country, especially Mindanao and Palawan, are mostly affected by Intertropical Convergence Zone (ITZC). It is rare that tropical cyclones cross the southern part of the country. However, the result of the study by Brunt (1969) indicated that it was not unusual for the tropical cyclone to originate near the Equator in the western Pacific although very rare reached typhoon intensity. Since 2011, a series of disastrous typhoons crossed the Mindanao region in succession, Typhoon Sendong (December 2011), Typhoon Pablo and Tropical Storm Quinta in December 2012.

The left hand image in Figure 4.2 depicted the actual tracks of tropical cyclones formed in the Western North Pacific (WNP) and have impacted during the period 1948-2006 where the observed annual frequency is compared to the synthetic technique. A total of 1728 tropical cyclone developed in the Basin in which 1,148 entered the Philippine Area of Responsibility (PAR) and about 66% reached typhoon intensity. In the right hand image in Figure 4.2, the model was able to simulate and replicate the tracks when compared with the actual TC tracks that entered the PAR.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 29 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.2. Spatial distribution of 100 years synthetic tropical cyclone tracks (Left) and the actual tropical cyclones tracks in the Philippine Area of Responsibility for the period 1948 to 2011 (PAR) (Right). The color scale represents increasing intensity from blue to red.

4.1.1 Regional severe wind hazard

The regional severe wind hazard maps were developed using TCRM, and represent a 3-second gust wind speed at 10-meter height above open, flat terrain (Class 10 in Table 3.1). Wind speeds were calculated at 0.02° horizontal resolution (approximately 2 kilometers), for return periods ranging from 2 years to 10,000 years. Figure 4.3 presents the 1% annual exceedance probability (AEP) wind speed for the entire Philippines. As expected, the highest wind speeds are in the northern and eastern sections of the country, corresponding to those regions most often impacted by tropical cyclones which initially developed in the Western North Pacific (WNP) area and moved westerly to northwest direction. 1% AEP gust wind speeds exceed 250 km/h over Catanduanes. Gust wind speeds are lowest over southern and western parts of Mindanao, where tropical cyclones are infrequent and less intense.

For the 1% AEP (100-year return period) for GMMA (Figure 4.4), it shows that the higher wind speed values are located in higher elevation in the north-eastern and eastern part of GMMA and these are most exposed to severe winds (e.g. mountainous areas in Montalban, San Mateo, Cainta, Antipolo, Tanay, Taytay, and Angono). The maximum gusts ranged from 160.1 to 180 km/h which is of typhoon intensity. In terms of wind hazards, elevated areas are more vulnerable to severe winds than those located on the shielded side of the mountain.

30 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.3. 1% annual exceedance probability (100-year return period) regional wind speeds for the Philippines.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 31 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.4. Regional severe wind hazard maps over open terrain in Greater Metro Manila.

4.1.2 Local severe wind hazard

The local severe wind hazard map identifies the specific locations susceptible to severe wind. The local severe wind hazard maps show critical areas where potentially extreme wind speeds could cause damage to buildings, incorporating local variations due to topography, land cover and shielding effects. The local wind hazard is a combination of the impact of the local effects, captured by the multipliers.

4.1.2.1 Hillshape multiplier (Mh)

For the majority of GMMA, Mh is very close to 1, due to the relatively flat topography (Figure 4.5).

Highest values of Mh occur in the mountainous areas to the northeast in Rodriguez, San Mateo, Antipolo, Taytay and Angono – values in these areas exceed 1.6. Running almost north-south across GMMA, the escarpment along the West Valley fault (along the western side of the Marikina River) also stands out, though to a lesser extent, with Mh values between 1.2 and 1.4.

32 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.5. Maximum value (from eight directions considered) of Mh across GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 33 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

4.1.2.2 Terrain multiplier (Mz)

The terrain multiplier map was based on the land cover classification described in Section 3.2.3. The majority of GMMA is heavily urbanised, composed of high-density residential, industrial and commercial areas and high-rise city centers. Most of the inner parts of GMMA have Mz values of less than 0.8 (Figure 4.6) - southern parts of Quezon City, Caloocan, Makati, Manila, San Juan and Mandaluyong. The presence of some open spaces (e.g. the Pasig River mouth in Manila) within the urban area can be seen as areas of higher Mz value. Further to the north, east and south, urbanization is reduced and higher Mz values are indicated.

34 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Legend

Multiplier Values < 0.8 0.8 - 0.99 0.99 - 1.01 1.01 - 1.2 1.2 - 1.4 1.4 - 1.6 > 1.6

Figure 4.6. Maximum value of Mz (from eight directions considered) for GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 35 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

4.1.2.3 Shielding Multiplier (Ms)

The heavily urbanized nature of GMMA also impacts the Ms values (Figure 4.7). Closely spaced buildings throughout the area provide high levels of shielding to neighbouring structures, so across much of the built-up area values are 0.8 or lower. Again, open areas are clearly identifiable in the Ms values, more so than in the Mz values. This is due to the short range of the shielding effects. For example, NAIA and golf courses stand out due to the absence of buildings in these areas. Ms values in the rural outskirts of GMMA (Rodriguez, San Mateo and Antipolo) are generally 1.

36 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.7. Maximum value of Ms (from eight directions considered) for GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 37 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

4.1.2.4 Combined site-exposure multiplier (M3)

For convenience, the three multipliers (Mh, Mz and Ms) are combined into a single multiplier. This allows us to see the overall effect of the landscape on local wind speeds – combining the effects of topography, land cover and shielding. As shown in Figure 4.8, generally there is a decrease in gust wind speed by over 20% in central and southern part and an increase in the mountainous area in the north-eastern side by 10 to 50% per increase in height as a result of the effects of the wind multipliers.

38 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Legend

Multiplier Values < 0.8 0.8 - 0.99 0.99 - 1.01 1.01 - 1.2 1.2 - 1.4 1.4 - 1.6 > 1.6

Figure 4.8. Combined wind multiplier generated for GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 39 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

4.1.2.5 Local severe wind hazard

Figure 4.9 presents the local wind speed hazard (0.2% AEP) for GMMA. There are isolated pockets in Manila LGU where wind speeds are in the range 60-100 km/h. Angono, Antipolo, Rodriguez, San Mateo and Taytay are potentially threatened with 161 to 202 km/h mean wind speed for 0.2% AEP (500-year return period) because these are mountainous areas and are located in higher elevated areas in the north-eastern part of GMMA. Most of the urban areas of GMMA are in the range of 100- 140 km/h gust wind speed. Slightly higher wind hazard is present along the shoreline of Laguna de Bay in Muntinlupa, Taguig and Angono, due to the low roughness of the water body to the south and east of these areas. Areas along the Pasig and Marikina Rivers, and the Manggahan Floodway experience higher wind hazard, as do areas neighbouring the Ninoy Aquino International Airport.

Table 4.1 details the average gust wind speed across each LGU within GMMA, for a range of AEPs. San Mateo, Rodriguez and Antipolo experience the highest wind speeds at all AEPs, due to the mountainous terrain and lack of shielding effects. The lowest AEP wind speeds are experienced by San Juan, Manila and Makati LGUs – this is because of the heavily urbanized landscape and relatively flat topography in those areas.

40 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.9. 0.2% AEP local gust wind speed for GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 41 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table 4.1 Maximum gust wind speed (km/h) in GMMA for different AEPs.

LGU 5% AEP 2% AEP 1% AEP 0.5 % AEP 0.2% AEP

Angono 131 144 153 161 171 Antipolo 147 162 172 181 192

Cainta 107 117 125 131 138

Caloocan 104 114 121 127 134

Las Piñas 104 114 121 127 135

Makati 93 102 109 114 121

Malabon 105 116 122 129 136

Mandaluyong 96 106 112 118 125

Manila 92 101 107 112 119

Marikina 102 112 119 125 132

Muntinlupa 114 125 133 140 148

Navotas 114 125 132 139 146

Paranaque 103 113 121 117 124

Pasay 106 116 123 130 137

Pasig 100 110 117 123 130

Pateros 97 106 113 119 125

Quezon City 104 115 122 128 135 Rodriguez 151 167 177 186 196

San Juan 91 100 106 111 118

San Mateo 156 172 182 191 202

Taguig 114 126 133 140 148

Taytay 124 136 145 152 161

Valenzuela 106 116 123 129 136

4.2 Estimation of damage from wind hazard

To be able to assess the wind damage likely to occur in Metro Manila, severe wind risk assessment was conducted. Wind Risk assessment is a function of the interaction of the wind hazard, building exposure and the vulnerability of the building structures that will be impacted by the wind hazard and can be used to determine what might be the expected losses in terms of property damage.

In this study, the wind risk assessment is presented using the different measures of property damages, which are aggregated by barangay level and are computed based on the following:

42 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

1. Physical damage in terms of building damaged floor area equivalent values for each barangay/municipality (in hectares/km²) for the different AEPs. The risk calculations allow estimation of the damages that hypothetical typhoons could cause to buildings. The estimated damage is expressed in terms of direct building damaged economic losses on the building structures aggregated by barangay level. 2. Economic loss in terms of Damage/Repair cost is used to estimate the economic cost in PHP million/km² associated with tropical cyclone damage to building.

4.2.1 Damaged Floor Area Equivalent for the 0.2% AEP (1/500) event

Building damaged intensity measures the intensity of physical damage to buildings due to tropical cyclones, regardless of monetary value. The potential damage to building is calculated by integrating wind speed, building exposure and the vulnerability. To determine the impact of severe wind, the buildings that are exposed to damage was assessed and only the direct impact to buildings was considered. The physical impact data can be used as an input for the estimation of economic loss to determine how the community can be impacted by tropical cyclone severe wind. 4.2.1.1 Taguig City

For the 0.2% AEP (1/500) wind event, South Daang Hari, San Miguel and Bagong Tanyag have the highest physical damage followed by North Daang Hari, Fort Bonifacio and Lower Bicutan with a total damaged floor area equivalent of 3.96, 2.03, 1.85, 1.77, and 1.69 ha/km² respectively (Figure 4.10). These locations have been identified as having moderate to extreme wind risk as compared to other barangays in Taguig City. These are densely built-up area with multi-type of buildings are located. More intensely damaged by severe wind was found in South Daang Hari due to high distribution of wooden, hollow-blocks and makeshift-type of buildings, (W1, C1, N, CHB, and CWS) and has a high hazard level.

The mean maximum wind speed for 0.2% AEP wind event in Taguig City ranged from 122 to 163 km/h. The south-western side and central part of Taguig City are most prone to wind impact. The impact (damage) of severe wind varies significantly between locations, those that have significantly more wind risk depends on the combination of the hazard, era of construction, high proportion of building stocks, and the types of building. Table 4.2 summarizes the building damaged intensity due to severe wind across Taguig City.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 43 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.10. Building Damage Intensity for 0.2% AEP (1/500) event in Taguig City and spatial distribution of buildings in Fort Bonifacio.

44 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table 4.2. Barangay-level damaged floor area equivalent (ha/km2) for different AEPs in Taguig City.

BARANGAY 5%AEP 2%AEP 1%AEP 0.5%AEP 0.2%AEP (1/20) (1/50) (1/100) (1/200) (1/500)

TAGUIG - BAGUMBAYAN 0.26 0.44 0.60 0.78 1.03 TAGUIG - TANYAG (BAGONG 0.36 0.71 1.07 1.47 2.03 TANYAG) TAGUIG - NORTH DAANG HARI 0.36 0.68 1.00 1.36 1.85

TAGUIG - CENTRAL BICUTAN 0.25 0.43 0.61 0.81 1.08

TAGUIG - UPPER BICUTAN 0.25 0.43 0.60 0.79 1.06

TAGUIG - LOWER BICUTAN 0.47 0.77 1.04 1.32 1.69

TAGUIG - NEW LOWER BICUTAN 0.17 0.28 0.39 0.51 0.68

TAGUIG - MAHARLIKA VILLAGE 0.14 0.25 0.35 0.46 0.62

TAGUIG - KATUPARAN 6 0.20 0.36 0.51 0.67 0.89

TAGUIG - WESTERN BICUTAN 0.20 0.34 0.46 125 0.77

TAGUIG - PINAGSAMA 0.28 0.51 0.72 0.59 1.26

TAGUIG - NORTH SIGNAL VILLAGE 0.16 0.28 0.40 0.95 0.71

TAGUIG - CENTRAL SIGNAL 0.19 0.33 0.46 0.61 0.81 VILLAGE (SIGNAL VILLAGE) TAGUIG - SOUTH SIGNAL VILLAGE 0.18 0.31 0.43 0.57 0.76

TAGUIG - SAN MIGUEL 0.12 0.21 0.29 0.36 0.47

TAGUIG - HAGONOY 0.25 0.41 0.56 0.71 0.93

TAGUIG - WAWA 0.14 0.23 0.32 0.41 0.57 TAGUIG - BAMBANG 0.35 0.61 0.84 1.08 1.41

TAGUIG - USUSAN 0.25 0.46 0.66 0.88 1.18

TAGUIG - TUKTUKAN 0.26 0.45 0.62 0.80 1.05

TAGUIG - SANTA ANA 0.15 0.25 0.33 0.42 0.53

TAGUIG - CALZADA 0.16 0.26 0.34 0.45 0.62

TAGUIG - PALINGON 0.26 0.44 0.58 0.74 0.97

TAGUIG - LIGID-TIPAS 0.16 0.27 0.37 0.48 0.65

TAGUIG - IBAYO-TIPAS 0.21 0.36 0.49 0.63 0.83

TAGUIG - NAPINDAN 0.16 0.26 0.36 0.48 0.67

TAGUIG - FORT BONIFACIO 0.35 0.67 0.97 1.31 1.77

TAGUIG - SOUTH DAANG HARI 0.67 1.43 2.17 2.95 3.96

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 45 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

4.2.1.2 GMMA

Figure 4.11. Building Floor Area Equivalent for the 0.2% AEP (1/500) event in GMMA.Figure 4.11 shows the expected building damage for the 0.2% AEP (1/500) wind event in GMMA at barangay level. The western central sections of GMMA have a higher risk than the other half of GMMA. These areas are highly influenced by a high proportion of building stocks and where the proportions of old structures are also high. The City of Mandaluyong shows the highest risk (in terms of damaged floor area equivalent - 1.07 ha/km²) due to high concentration of “makeshift-informal” type of settlements with large areas of makeshift (N), concrete hollow blocks (CHB) and wood, light frame (W-1)-types of buildings and its exposure to high winds. Payatas in Quezon City is highly vulnerable to severe wind, it is dominated by makeshift (N) and wooden types of building.

Table 4.3 shows the estimated property damage (buildings only) expressed in building floor area equivalent for 0.2% AEP wind event for all the municipalities/cities in GMMA. As expected the probability of losses to buildings are either due to high hazard levels, older types of structures, building types and building density or the combination of all the factors.

46 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.11. Building Floor Area Equivalent for the 0.2% AEP (1/500) event in GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 47 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table 4.3. Summary of the damaged floor area equivalent by municipality/city for the different AEPs in GMMA.

Municipality 5% AEP 2% AEP 1% AEP 0.5% AEP 0.2% AEP (1/20) ha (1/50) ha (1/100) ha (1/200) ha (1/500) ha per km² per km² per km² per km² per km²

ANGONO 0.22 0.35 0.47 0.60 0.79

ANTIPOLO 0.16 0.26 0.35 0.44 0.56

CAINTA 0.21 0.36 0.49 0.63 0.83

CALOOCAN 0.18 0.32 0.44 0.58 0.75

LAS PINAS 0.15 0.26 0.36 0.47 0.63

MAKATI 0.15 0.28 0.41 0.57 0.79

MALABON 0.23 0.39 0.54 0.70 0.92

MANDALUYONG 0.21 0.39 0.57 0.78 1.07

MANILA 0.17 0.31 0.45 0.61 0.85

MARIKINA 0.20 0.36 0.49 0.64 0.84

MUNTINLUPA 0.17 0.30 0.42 0.55 0.74

NAVOTAS 0.17 0.29 0.40 0.51 0.66

PARANAQUE 0.19 0.34 0.47 0.62 0.83

PASAY 0.20 0.35 0.49 0.64 0.85

PASIG 0.19 0.34 0.48 0.64 0.87

PATEROS 0.24 0.41 0.58 0.76 1.00

QUEZON CITY 0.15 0.25 0.35 0.47 0.62

RODRIGUEZ 0.07 0.12 0.15 0.20 0.25

SAN JUAN 0.18 0.31 0.45 0.61 0.83

SAN MATEO 0.07 0.12 0.15 0.19 0.25

4.2.2 Barangay Building Damage Cost (PHP million/km²)

Barangay building damage cost measures the physical damage to buildings in terms of the construction/replacement cost and reflects the overall impact to buildings. It can cause damage to residential, commercial and critical facilities. 4.2.2.1 Taguig City

For the 0.2% AEP of wind event, it is estimated that the expected cost of damage to buildings to be PHP 130.2 million/km² for the whole of Taguig City (Figure 4.12, Table 4.4). Fort Bonifacio showed a high cost of damage followed by North and South Daanghari and Barangay Lower Bicutan showed the highest expected loss due to dense settlements and combination of high hazard, dominated with makeshift and wooden-type of building and older types of building.

48 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.12. Wind Risk Map Damage Cost (Replacement Value) for the 0.2% AEP (1/500) event in Taguig City.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 49 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table 4.4. Taguig Barangay Damage Cost in PHP million /km² for the different AEPs.

MUNICIPALITY 5% AEP 2% AEP 1% AEP 0.5 % AEP 0.2% AEP (1/20) (1/50) (1/100) (1/200) (1/500)

TAGUIG - BAGUMBAYAN 31.91 54.48 74.96 97.30 128.58

TAGUIG – TANYAG (BAGONG 24.22 48.55 72.98 100.50 138.61 TANYAG) TAGUIG – NORTH DAANG HARI 41.79 78.39 114.54 155.29 211.78

TAGUIG – CENTRAL BICUTAN 36.29 63.91 89.98 118.79 158.42

TAGUIG – UPPER BICUTAN 25.62 44.40 62.01 81.51 108.45

TAGUIG – LOWER BICUTAN 52.06 85.77 115.30 146.32 187.52

TAGUIG – NEW LOWER BICUTAN 15.78 2671 36.76 48.05 64.78

TAGUIG – MAHARLIKA VILLAGE 13.12 23.02 32.46 43.02 57.84

TAGUIG – KATUPARAN 6 18.80 33.24 46.80 61.73 82.16 TAGUIG WESTERN BICUTAN 24.19 40.52 55.03 70.50 91.49

TAGUIG - PINAGSAMA 40.07 73.77 105.06 138.71 183.32

TAGUIG – NORTH SIGNAL 14.52 25.92 36.75 48.78 65.43 VILLAGE TAGUIG – CENTRAL SIGNAL 19.19 33.51 47.01 62.01 82.81 VILLAGE TAGUIG – SOUTH SIGNAL 16.58 28.73 40.18 52.94 70.73 VILLAGE TAGUIG – SAN MIGUEL 14.30 25.32 34.64 44.23 57.13 TAGUIG – HAGONOY 17.93 29.83 40.19 51.38 67.04

TAGUIG - WAWA 11.49 18.80 25.31 33.00 45.36 TAGUIG - BAMBANG 23.46 41.16 56.70 73.15 95.48

TAGUIG - USUAN 27.71 51.34 73.64 98.16 131.65

TAGUIG – SANTA ANA 36.35 62.68 86.27 111.63 146.58

TAGUIG - CALZADA 11.60 19.70 26.27 32.88 41.63

TAGUIG - PALINGON 15.51 25.22 33.78 43.93 60.65

TAGUIG – LIGID-TIPAS 16.01 26.53 35.43 44.95 58.85

TAGUIG – IBAYO-TIPAS 21.89 36.94 50.45 65.35 86.56

TAGUIG – NAPINDAN 11.57 19.24 26.30 32.82 48.99

TAGUIG – FORT BONIFACIO 78.29 148.59 216.35 291.57 393.90

TAGUIG – SOUTH DAANG HARI 33.95 72.45 109.62 149.25 200.04

50 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

4.2.2.2 GMMA

The Greater Metro Manila Area may suffer costly wind damages due to damaged structures (residential, commercial, industrial and critical facilities and other structures) and the total cost in the Greater Metro Manila is approximately PHP 77.61 Million/km² for the 0.2% AEP (Table 4.5). The City of Mandaluyong has the highest expected economic loss amounting to PHP 163.87 Million/km², being densely built-up and due to more vulnerable building types (makeshift (N), wood one-storey (W1), and concrete (C1) and pre-1972 building stocks. There is a significant spatial variation of the risk in highly dense built-up area as a result of exposure as shown in Figure 4.13. The expected cost of damage depends on the high proportion of wind damaged buildings as well as where the building cost is high.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 51 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure 4.13. Building Damage Cost (Replacement Value) for the 0.2% AEP (1/500) event in GMMA.

52 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table 4.5. Expected Building Damage Cost (PHP million/km²) in GMMA.

MUNICIPALITY 5% AEP 2% AEP 1% AEP 0.5 % AEP 0.2% AEP (1/20) (1/50) (1/100) (1/200) (1/500) million peso million peso million peso million peso million peso per km² per km² per km² per km² per km² ANGONO 19.37 31.25 41.76 53.33 69.79 ANTIPOLO 15.18 24.84 33.15 41.70 52.48

CAINTA 19.44 33.12 45.45 58.97 77.66

CALOOCAN 19.08 33.28 46.02 59.59 77.25

LAS PINAS 15.95 27.77 38.78 50.91 68.04

MAKATI 24.34 45.84 67.62 93.10 130.15

MALABON 23.51 39.78 54.70 71.11 93.98

MANDALUYONG 32.08 33.28 46.02 59.59 163.87

MANILA 23.45 42.55 61.45 83.24 114.61

MARIKINA 22.21 38.61 53.56 69.75 91.19

MUNTINLUPA 21.71 37.66 52.43 68.83 92.32

NAVOTAS 14.35 24.27 33.05 42.37 54.91

PARANAQUE 20.68 36.27 50.89 67.02 89.47

PASAY 30.90 53.09 74.00 97.28 129.96

PASIG 26.08 46.58 66.10 87.98 118.38

PATEROS 20.59 36.04 50.34 65.94 87.19

QUEZON CITY 17.53 30.63 42.87 56.45 74.89 RODRIGUEZ 7.29 11.79 15.71 19.94 25.60

SAN JUAN 23.17 41.47 59.47 80.16 109.52

SAN MATEO 7.27 11.80 15.67 19.70 24.92

TAGUIG 29.56 52.66 74.11 97.65 130.17

TAYTAY 21.93 36.09 48.84 62.95 83.15

VALENZUELA 19.56 32.70 44.61 57.56 75.02

Total GMMA Building Damage Cost 18.55 32.12 44.70 58.56 77.61

Figure 4.14 depicts and summarizes the damaged floor area equivalent while Figure 4.15 shows the building cost estimates for the different AEPs in GMMA. It showed that the floor area equivalent and the building cost is higher with higher return period or annual exceedance probability but this type of disastrous event rarely happens and most frequent event are considered less disastrous.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 53 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

0.70

Damaged floor area equivalent

) 2 0.60

0.50

0.40

0.30

0.20

0.10 Damaged area floor equivalent (ha/km

0.00 4.9% AEP 2.0% AEP 1.0% AEP 0.5% AEP 0.2% AEP Annual Exceedance Probability

Figure 4.14. Damaged floor area equivalent (ha/km2) for each AEP.

90.0

) Cost per square kilometer (million PHP/km ) 2 80.0

70.0

60.0

50.0

40.0

30.0

20.0

10.0 Costper square kilometer (million PHP/km 0.0 4.9% AEP 2.0% AEP 1.0% AEP 0.5% AEP 0.2% AEP Annual Exceedance Probability

Figure 4.15. Damage costs (million PHP/km2) for each AEP.

54 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

5 Conclusion

This study presents the new Risk Assessment methodology for Severe Wind in the Greater Metro Manila Area (GMMA). The Tropical Cyclone Risk Model (TCRM) developed by Geoscience Australia (GA) was used in this study that provides a new set of tools to identify the critical areas most likely be impacted by extreme winds and can estimates losses.

Based from the TCRM result, the eastern and the north-eastern part of GMMA are more vulnerable to severe wind hazard as these are situated in critical locations that are exposed to severe winds. The spatial distribution of wind hazard varies significantly between structures and locations due to local site effects like geographic terrain, the height of structures, topographic factors and its location in high hazard area. At higher return periods, the wind hazards are stronger but are of less frequency and are localized while the low return periods are more frequent but less intense. The Regional Severe Wind Hazard Maps can be used to update the wind zoning map of the Philippines and can be considered in building design and as a guide for emergency managers and planners for evacuation planning. Local hazard maps can assist in site selection for evacuation centers to ensure they are in the safest location, but remain accessible to those in the community expected to utilise the centers.

In assessing the risk, the western and the central sections of GMMA are subject to severe wind impact and have a higher risk than the other areas in GMMA. These areas are densely built-up with high proportion of vulnerable building types (makeshift, wood-type), old structures or “high rise” buildings, and that are located in high hazard areas. On the other hand, the expected cost of wind damage depends on the proportion of wind damaged buildings and the cost of the building.

For those areas identified as high risk to wind damage, building codes/regulations must be strictly implemented to mitigate severe wind risks. For already developed areas, retrofitting is encouraged – the methods applied in this study can be used to set out a cost-benefit study for retrofitting older, more vulnerable building types to increase their resilience to severe winds.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 55 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

References

Amadore, L. A., J.F. Bucoy, S. D., Talib, and S.O. Yanga, 1985: Preliminary Typhoon Damage Scale in the Philippines. Joint AMS/PMS International Conference in Applied Meteorology and Climatology, 18-22, March 1985, Manila. Amadore, L. A, Development of a Conceptual and Quantitative Typhoon Damage Model, NRCP Supported Project, National Research Council of the Philippines. 2011 Arthur, W.C., A. Schofield, R. P. Cechet and L. L. A. Sanabria (2008a): Return period cyclonic wind hazard in the Australian Region. 28th AMS Conference on Hurricanes and Tropical Meteorology, 28 April 2 May 2008, Orlando, FL, USA. https://ams.confex.com/ams/pdfpapers/138556.pdf Arthur, W.C., A. Schofield, R. P. Cechet, and L. A. Sanabria, 2008b: Severe wind hazard assessment of Cyclone Tracy using a parametric tropical cyclone model. 15th AMOS National Conference. Arthur, W.C., Thomas, C.M., Sanabria, L. A. and Cechet, R.P. (2010) Reassessment of wind hazard in the current climate. Proc. Of 14th AWES Workshop, Canberra, August 5 and 6th. Cechet, R.P., A. Sanabria, T. Yang, W.C. Arthur, C.H. Wang and X. Wang, 2011.An assessment of severe wind hazard and risk for Queensland’s Sunshine Coast region.19th International Congress on Modelling and Simulation, Perth, Australia, 12–16 December 2011 http://mssanz.org.au/modsim. Cinco, T. A., F. Hilario, et al., 2011: Updating Tropical Cyclone Climatology in the Philippines. Climate Data Section, Climatology and Agrometeorology Branch. PAGASA-DOST. Hall, T. M. and Jewson, S.: Statistical modeling of North Atlantic Tropical Cyclone Tracks, Tellus A, 2007, 59, 486-498 Holland, G.J., 1980: An analytical model of the wind and pressure profiles in hurricanes, Monthly Weather Review, 108, 1212-1218pp Lin, X.G. and Nadimpalli, K. (2005). Natural Hazard Risk in Perth: Chapter 3: Severe Wind Hazard Assessment in Metropolitan Perth, Geoscience Australia Report, GeoCat No. 63527. Knutson, T.R. McBride, J. L. Chan, J., Emanuel, K., Holland, G, Landsea, C., Held, I., P. Kossin, J., A. K. Srivastava and Masato Sugi, 2010: Tropical Cyclones and Climate Change. Nadimpalli, K., R.P. Cechet and M. Edwards, 2007.Severe Wind Gust Risk for Australian Capital Cities – A National Risk Assessment Approach. http://mssanz.org.au/MODSIM07/papers/26_s32/SevereWind_s32_Nadimpalli_.pdf Nadimpalli, K., M. Edwards and B. Cechet. Impacts of Severe Wind Gust Risk to the Australian Cities – A National Risk Assessment Approach. National Statistical Coordination Board, 2008, Metro Manila Congested at 18,650 Persons per Square Kilometer in 2007. Summons, Nicholas, 2012. TCRM: User Manual. Modelling the wind hazard from tropical cyclones, International Geological Congress. National Statistics Office, 2000, Population Density for the National Capital Region (NCR). Standards Australia/Standards New Zealand, 2011: AS/NZS 1170.2 Structural design actions, Part 2: Wind actions. Wehner, M. J. Ginger, J. Holmes, C. Sandland and M. Edwards, 2010: Development of Methods for Assessing Vulnerability of Australian Residential Buildings Stock to Severe Wind, 17th National Conference of Australian Meteorological and Oceanographic Society, 1OP Publishing 10P Conference Series: Earth Environmental Science 11 (2010).

56 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix A - Taguig City Severe Wind Hazard Maps

Figure A.1. Taguig City severe wind hazard map 5% AEP (20-year return period).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 57 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure A.2. Taguig City severe wind hazard map 2% AEP (50-year return period).

58 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure A.3. Taguig City severe wind hazard map 1% AEP (100-year return period).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 59 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure A.4. Taguig City severe wind hazard map 0.5% AEP (200-year return period).

60 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure A.5. Taguig City severe wind hazard map 0.2% AEP (500-year return period).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 61 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

62 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix B - GMMA Severe Wind Hazard Maps

Figure B.1. GMMA severe wind hazard map 5% AEP (20-year return period).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 63 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure B.2. GMMA severe wind hazard map 2% AEP (50-year return period).

64 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure B.3. GMMA severe wind hazard map 1% AEP (100-year return period).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 65 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure B.4. GMMA severe wind hazard map 0.5% AEP (200-year return period).

66 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure B.5. GMMA severe wind hazard map 0.2% AEP (500-year return period).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 67 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

68 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix C - Taguig City Damaged Floor Area Equivalent Maps

Figure C.1. 5%AEP (20-year return period) damaged floor area equivalent (ha/km2) in Taguig City.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 69 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure C.2. 2%AEP (50-year return period) damaged floor area equivalent (ha/km2) in Taguig City.

70 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure C.3. 1%AEP (100-year return period) damaged floor area equivalent (ha/km2) in Taguig City.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 71 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure C.4. 0.5%AEP (200-year return period) damaged floor area equivalent (ha/km2) in Taguig City.

72 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure C.5. 0.2%AEP (500-year return period) damaged floor area equivalent (ha/km2) in Taguig City.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 73 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

74 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix D - GMMA Damage Floor Area Equivalent Maps

Figure D.1. 5%AEP (20-year return period) damaged floor area equivalent (ha/km2) in GMMA

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 75 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure D.2. 2%AEP (50-year return period) damaged floor area equivalent (ha/km2) in GMMA.

76 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure D.3. 1%AEP (100-year return period) damaged floor area equivalent (ha/km2) in GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 77 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure D.4. 0.5%AEP (200-year return period) damaged floor area equivalent (ha/km2) in GMMA.

78 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure D.5. 0.2%AEP (500-year return period) damaged floor area equivalent (ha/km2) in GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 79 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix E - Taguig City Damage Building Cost Maps

Figure E.1. 5% AEP (20-year return period) building damage costs (million Peso/km2) in Taguig City.

80 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure E.2. 2% AEP (50-year return period) building damage costs (million Peso/km2) in Taguig City.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 81 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure E.3. 1% AEP (100-year return period) building damage costs (million Peso/km2) in Taguig City.

82 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure E.4. 0.5% AEP (200-year return period) building damage costs (million Peso/km2) in Taguig City.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 83 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure E.5. 0.2% AEP (500-year return period) building damage costs (million Peso/km2) in Taguig City.

84 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix F - GMMA Damage Building Cost Maps

Figure F.1. 5% AEP (20-year return period) building damage costs (million Peso/km2) in GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 85 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure F.2. 2% AEP (50-year return period) building damage costs (million Peso/km2) in GMMA.

86 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure F.3. 1% AEP (100-year return period) building damage costs (million Peso/km2) in GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 87 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure F.4. 0.5% AEP (200-year return period) building damage costs (million Peso/km2) in GMMA.

88 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure F.5. 0.2% AEP (500-year return period) building damage costs (million Peso/km2) in GMMA.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 89 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix G - Sample Case: Typhoon Milenyo (International Name: Xangsane - 0615)

Tropical cyclone is a perennial problem in the Philippines and the building structures exposed to tropical cyclone severe winds are at risk of being damaged due to its impact. GMMA is exposed to various intensity of wind speed as experienced during the occurrence of Typhoon Milenyo where the estimated maximum wind speed ranged from 89 to 156 kilometers per hour (km/h) (Figure 5.17). The maximum wind speed values are within the radius of maximum winds near the center relative to the typhoon track. The wind hazard varies considerably in the surrounding areas due to local terrain roughness, the shielding due to upwind structures and topographic factors. A single typhoon will not result in wind speeds of the same return period across the Greater Metro Manila Area (as demonstrated by the Milenyo example).

The estimated maximum wind speed from TCRM was compared and showed fairly good agreement with the observed maximum wind speed from the different PAGASA Synoptic stations. It developed into a tropical depression in the vicinity of 11.5 N/129.0E on the morning of the 25th (250000) then intensified into a tropical storm 18 hours late. The observed wind speed from NAIA Stations recorded 143 km/h (40 m/s) and the estimated wind speed in Paranaque was 140 km/h (39 m/s) and in Pasay it was 124 km/h (34 m/s) while in Port Area, the observed maximum wind speed was 108 km/h (30 m/s) and the estimated wind speed was 104 km/h (29 m/s).

Typhoon Milenyo (Xangsane) passed very close to the South of Metro Manila on the 28th of September 2006 before it moved towards West Philippine Sea. It was the worst tropical cyclone (in terms of strong winds) to impact Metro Manila since Typhoon Rosing in November of 1995. Wind- damaged impacts were enormous which include toppled down billboards and electrical posts causing power disruptions and damaged buildings and school houses (Figure 5.18). As can be noted, most of the damage is within the few kilometers near the center of Typhoon Milenyo, big trees were either downed or uprooted and several houses of medium and light materials were damaged (many unroofed) unroofed houses of medium-built materials and damage to G.I. roofing.

90 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure G.1. Wind field and typhoon track of Typhoon Milenyo, September 28, 2006. Local wind speeds are shown within GMMA, regional wind speeds otherwise.

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 91 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Figure G.2. Damage from Typhoon Milenyo 2006 over Metro Manila, Albay and Sorsogon Areas. Source: PAGASA STRIDE Team.

Figure G.3. Wind-related damage from Typhoon Milenyo in September 2006 (Source: flickr.com).

92 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Appendix H - Risk calculation

In Figure H.1, we can see that the wind speed across a single exposure polygon can vary due to the changes in the landscape (e.g. hills and ridges, open spaces, trees, etc.). For the selected polygon, we calculate the mean wind speed over the entire polygon, a value of 122 km/h.

Figure H.1. Peak wind speed averaged across each polygon. The highlighted polygon has a mean wind speed of 122 km/h.

For this example, we have chosen a single polygon in the Taguig area. The polygon has the following attributes:

1. Formal Settlement, with Mixed Residential and Small Commercial land use; 2. Total land area of 32565 m2; 3. Total floor area of 18767 m2; 4. There are 13662 m2 of 1-storey buildings, 4330 m2 of 2-storey buildings and 777 m2 of medium rise buildings (3–8 storeys);

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 93 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Of course, this case is much simpler than are most polygons in the exposure database. In reality, there will be multiple building types within each polygon; with different heights (e.g. around 10 different building types would be common).

However, we will show later on how to account for this issue, by adapting the calculation from the simple case of a single building type. Thus, it is most important to understand the calculation for the simple case first.

Measures of damage

The measures of damage that we use are:

1. Damage fraction: the ratio of the cost of repairs to the total construction cost. This can apply to individual building types, a group of buildings (such as in one exposure polygon) or across an area such as a barangay or city. 2. Damaged floor area equivalent: This is equal to the total floor area (m2) of a given building type multiplied by the damage fraction for that building type, divided by the total land area of the polygon. Intuitively, this measures the intensity of physical damage to buildings, irrespective of their monetary value. 3. Building damage cost: This is equal to the damaged floor area equivalent multiplied by the associated building construction cost. Intuitively, this measures the physical damage to buildings in terms of the cost of construction.

Risk calculation process

Following is a step-by-step process for calculating the measures of damage for a single polygon.

1. Calculate a representative wind speed for the polygon. This is usually the mean wind speed across the polygon. For our polygon of interest, this is 122.3 km/h.

2. Define the combinations of building type and storey categories from those available I the exposure databases. These are populated in the exposure database, with an estimated floor area for each of the combinations. In the polygon selected, there are 20 combinations of building type and storey category with floor areas greater than zero (W1-L-1, N-L-1, CHB-L-1, URA-L-1, URM-L-1, C1-L-1, S1-L-1, S3-L-1, W1-L-2, N-L-2, CHB-L-2, URA-L-2, URM-L-2, MWS-L-2, CWS-L-2, C1-L-2, S1-L-2, S3-L-2, C1-M and S1-M). Refer to the documentation for the exposure database for an explanation on how these values are calculated.

3. Using the building type, land use category and the building cost database, calculate the value of each building type/storey category class within the polygon. For each building type/storey category combination:

(Building value (PHP)) = (Floor area (m2)) x (Construction cost (PHP/m2)).

Add these to obtain the total value of built assets in the polygon (the total value is stored for future use). For the polygon chosen, there are 18769.1 m2 of floor area, with a total value of PHP 181,218,885 (Table H.1). The largest contribution to this value is from 1-storey C1 buildings (C1-L-1), of which there are 6793 m2, with a value of PHP 89,667,600.

94 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table H.1. Building type/storey category combinations, corresponding vulnerability model floor area of each combination (m2), construction cost (PHP/m2) and total value (PHP). Total floor area and total value for the polygon is given at the bottom.

Building Vulnerability Floor area (m2) Construction cost Building value (PHP) type/storey model (PHP/m2) category W1-L-1 W1-L 3205.6 PHP 5,600.00 PHP 17,951,360.00

N-L-1 N-L 256 PHP 1,200.00 PHP 307,200.00

CHB-L-1 CHB-L-W 3348.5 PHP 6,500.00 PHP 21,765,250.00

URA-L-1 CHB-L-W 8.2 PHP 7,150.00 PHP 58,630.00

URM-L-1 CHB-L-W 8.8 PHP 7,150.00 PHP 62,920.00

C1-L-1 C1-L-W 6793 PHP 13,200.00 PHP 89,667,600.00

S1-L-1 S1-L 22.3 PHP 30,000.00 PHP 669,000.00

S3-L-1 S3-L 19.7 PHP 15,000.00 PHP 295,500.00

W1-L-2 W1-L 770.5 PHP 5,600.00 PHP 4,314,800.00

N-L-2 N-L 61.5 PHP 1,200.00 PHP 73,800.00

CHB-L-2 CHB-L-W 804.9 PHP 6,500.00 PHP 5,231,850.00

URA-L-2 CHB-L-W 2 PHP 7,150.00 PHP 14,300.00

URM-L-2 CHB-L-W 2.1 PHP 7,150.00 PHP 15,015.00

MWS-L-2 MWS-L-W 235.7 PHP 6,000.00 PHP 1,414,200.00

CWS-L-2 CWS-L-W 810.4 PHP 9,000.00 PHP 7,293,600.00

C1-L-2 C1-L-W 1632.8 PHP 13,200.00 PHP 21,552,960.00

S1-L-2 S1-L 5.4 PHP 30,000.00 PHP 162,000.00

S3-L-2 S3-L 4.7 PHP 15,000.00 PHP 70,500.00

C1-M C1-M 774.5 PHP 13,200.00 PHP 10,223,400.00

S1-M S1-M 2.5 PHP 30,000.00 PHP 75,000.00

Total 18769.1 PHP 181,218,885.00

4. For each building type/storey category combination, use the representative wind speed and the corresponding vulnerability model to calculate the damage fraction for each combination (Table H.2). The N-L-1 and N-L-2 building classes experience the greatest damage (these are the Makeshift/Informal buildings).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 95 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table H.2. Damage fraction for each building type/storey category combination.

Building Representative wind Damage fraction type/storey speed (km/h) category W1-L-1 122.295 0.00003 N-L-1 122.295 0.29768 CHB-L-1 122.295 0.01298 URA-L-1 122.295 0.01298 URM-L-1 122.295 0.01298 C1-L-1 122.295 0.00963 S1-L-1 122.295 0.03182 S3-L-1 122.295 0.00092 W1-L-2 122.295 0.00003 N-L-2 122.295 0.29768 CHB-L-2 122.295 0.01298 URA-L-2 122.295 0.01298 URM-L-2 122.295 0.01298 MWS-L-2 122.295 0.03174 CWS-L-2 122.295 0.03174 C1-L-2 122.295 0.00963 S1-L-2 122.295 0.03182 S3-L-2 122.295 0.00092 C1-M 122.295 0.00486

5. For each building type/storey category combination:

(Cost of damage (PHP)) = (damage fraction) x (Building construction cost (PHP/m2)) x

(Building floor area (m2))

The greatest cost of damage is incurred by the C1-L-1 building type, with a cost of PHP 863,461 (Table H.3). This is due to the large amount of that building type in the polygon (see Table H.1).

96 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

Table H.3. Damage fraction, building value and cost of damage.

Building type/storey category Damage fraction Building value (PHP) Cost of damage (PHP) W1-L-1 0.00003 PHP 17,951,360.00 PHP 469.68 N-L-1 0.29768 PHP 307,200.00 PHP 91,446.36 CHB-L-1 0.01298 PHP 21,765,250.00 PHP 282,459.20 URA-L-1 0.01298 PHP 58,630.00 PHP 760.87 URM-L-1 0.01298 PHP 62,920.00 PHP 816.55 C1-L-1 0.00963 PHP 89,667,600.00 PHP 863,461.88 S1-L-1 0.03182 PHP 669,000.00 PHP 21,287.52 S3-L-1 0.00092 PHP 295,500.00 PHP 273.18 W1-L-2 0.00003 PHP 4,314,800.00 PHP 112.89 N-L-2 0.29768 PHP 73,800.00 PHP 21,968.56 CHB-L-2 0.01298 PHP 5,231,850.00 PHP 67,896.49 URA-L-2 0.01298 PHP 14,300.00 PHP 185.58 URM-L-2 0.01298 PHP 15,015.00 PHP 194.86 MWS-L-2 0.03174 PHP 1,414,200.00 PHP 44,890.31 CWS-L-2 0.03174 PHP 7,293,600.00 PHP 231,517.45 C1-L-2 0.00963 PHP 21,552,960.00 PHP 207,546.08 S1-L-2 0.03182 PHP 162,000.00 PHP 5,154.82 S3-L-2 0.00092 PHP 70,500.00 PHP 65.17 C1-M 0.00486 PHP 10,223,400.00 PHP 49,723.36 Total PHP 181,218,885.00 PHP 1,890,767.26

6. The cost of damage from all building type/storey category classes is summed to obtain the total cost of damage in the polygon. The total cost of damage for this polygon is PHP 1,890,767.26.

7. For the entire polygon:

(Damage fraction) = (Total cost of damage (PHP)) / (Total building value (PHP)).

For this polygon

(Damage fraction) = PHP 1,890,767.26 / PHP 181,218,885.00 = 0.01043.

8. For each building type/storey category class:

(Damaged floor area equivalent (ha/km2)) = (Damage fraction) x (floor area (m2)) x 100 /

(Total land area of polygon (m2)).

Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the 97 Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis FOR OFFICIAL USE ONLY

This gives the damaged floor area equivalent in units of hectares/km2 (the factor of 100 converts from m2/m2 to ha/km2).

Table H.4. Damaged floor area equivalent, and per square kilometre of land area by building class.

Building type/storey Floor area Damage fraction Damaged floor area Damaged floor area category (m2) (m2) equivalent (ha/km2) W1-L-1 3205.6 0.00003 0.084 0.00025756 N-L-1 256 0.29768 76.205 0.23401315 CHB-L-1 3348.5 0.01298 43.455 0.13344351 URA-L-1 8.2 0.01298 0.106 0.00032678 URM-L-1 8.8 0.01298 0.114 0.00035070 C1-L-1 6793 0.00963 65.414 0.20087428 S1-L-1 22.3 0.03182 0.710 0.00217901 S3-L-1 19.7 0.00092 0.018 0.00005593 W1-L-2 770.5 0.00003 0.020 0.00006191 N-L-2 61.5 0.29768 18.307 0.05621800 CHB-L-2 804.9 0.01298 10.446 0.03207666 URA-L-2 2 0.01298 0.026 0.00007970 URM-L-2 2.1 0.01298 0.027 0.00008369 MWS-L-2 235.7 0.03174 7.482 0.02297505 CWS-L-2 810.4 0.03174 25.724 0.07899440 C1-L-2 1632.8 0.00963 15.723 0.04828316 S1-L-2 5.4 0.03182 0.172 0.00052765 S3-L-2 4.7 0.00092 0.004 0.00001334 C1-M 774.5 0.00486 3.767 0.01156756 Total 18769.1 267.8 0.822

9. Add the damaged floor area equivalent values for all building type/storey category classes to obtain the total damaged floor area equivalent for the polygon. For this polygon, the total damaged floor area equivalent is 0.822 ha/km2.

This process is repeated for all polygons and for all available return periods. The cost of damage, damaged floor area equivalent and damage fraction measures are integrated across all the return periods to obtain annualised values for these measures of risk. Damage measures are aggregated to barangay and LGU level as outlined in Section 3.6.3.

98 Enhancing Risk Analysis Capacities for Earthquake, Tropical Cyclone Severe Wind and Flood for the Greater Metro Manila Area – Tropical Cyclone Severe Wind Risk Analysis