Abstract

This study focuses on index-based flood risk assessment in Metro Manila, the capital region of the Philippines and most densely populated region in the country. Its objective is to properly address urban characteristics in flood risk assessment by introducing a specific urban-type set of physical, social, economic and ecological indicators. Analytical hierarchy process (AHP) was used to quantify the optimal selection weights for each of the selected 14 indicators. Five levels of flood risk will be presented in spatial maps using geographic information system (GIS) ranging from Very Low Risk to Very High Risk. Results of this study are expected to aid in understanding flood hazard and risk in Metro Manila. Moreover, the resulting flood risk information can be used as a decision tool in policy making, land-use planning, developing guidelines and countermeasures and flood disaster insurance.

INTRODUCTION

Flood is one of the most destructive natural disasters in the Philippines (EM-DAT 2018), including Metro Manila, the National Capital Region, which is home to 12 million people. To reduce losses due to flooding, there is a need to know the extent of areas that are affected by floods and how vulnerable the people of these areas have become, thus spatial assessment of risk and identification of areas affected by floods would be effective.

Traditionally, flood risk is expressed in terms of expected damage and likelihood of occurrence. The flood damage is combined with information on the probability of the flood event and then plotted as a return period–damage curve (Apel et al. 2004; Meyer et al. 2009). However, the results obtained using this method provide neither sufficient information nor the required level of detail for input into flood risk reduction strategies. In addition, the use of damage to assess flood risk suffers from data scarcity, particularly in developing countries (Birkmann 2007; Gall 2007). According to Birkmann (2007), highly exposed regions, with high poverty levels and subject to repeated and catastrophic floods, may not necessarily register significant deaths and damage, although these factors make such places highly risky. Moreover, since mortality and damage figures are obtained from actual events, the use of damage assesses actual vulnerability, but potential vulnerability is ignored (Gall 2007).

Flood management cannot become technically controllable without a proper assessment of flood hazard mapping and flood hazard (Gigović et al. 2017). However, flood hazard itself only assesses the extent and depth of flood; it does not assess the consequences on the population, economy and environment, as flood risk assessment does (Rincón et al. 2018). In general, risk refers to the expected losses (in terms of fatalities, or in economic terms as damage to property) of a specific hazard to a specific element (e.g., evacuation center) at risk in a particular future time period or future scenarios (Albano et al. 2017).

In recent years, studies pertaining to flood risk assessment in the Philippines have been increasing. Pornasdoro et al. (2014) examined the flood-prone areas within Metro Manila to find out their degrees of disaster risk. Although the study was limited to population data and physical characteristics of barangays, the findings can be useful to urban and regional planners and government agencies involved in disaster risk reduction and mitigation management.

Siddayao et al. (2014) combined analytical hierarchy process (AHP) and geographic information system (GIS) to come up with a tool for evaluating flood risks in all areas in the municipality of Enrile, located in the province of Cagayan, northern Philippines. Three disaster criteria were considered in their estimation of flood risk: population density (F1), distance from the riverbank (F2) and site elevation (F3). Their study revealed that F2 is assessed by experts to be the largest contributing factor for disaster at 63.33% followed by F3 at 26.05% and F1 at 10.62%. Their developed tool is expected to be a very valuable resource for consulting, planning agencies and local governments in managing risk, land-use zoning, damage estimates, land tax valuation, life, and property insurance claim validation, good governance, lifeline emergency services and remediation efforts to mitigate risk.

Another study by Siddayao et al. (2015) incorporated the combination of AHP and GIS to evaluate flood risk in the Central Business District areas of Tuguegarao City, Philippines. They included four disaster criteria, namely population density (F1), distance from the riverbank (F2), site elevation (F3) and distance from ponds and creeks (F4).

In our study, AHP was used as the multi-criteria decision technique within a GIS mapping environment. A multi-criteria analysis method such as AHP provides a framework which can handle different views on the identification of the elements of a complex decision problem, organize the elements into a hierarchical structure, and study the relationships among components of the problem (Boroushaki & Malczewski 2010). Moreover, multi-criteria decision analysis within GIS may be used to develop and evaluate alternative plans that may facilitate compromise among interested parties (Malczewski 1999).

The objective of this study is to provide flood risk information considering urban characteristics by introducing physical, social, economic, and environmental indicators. The resulting different levels of flood risk will be presented in spatial maps using GIS.

METHODOLOGY

Study area

The National Capital Region (NCR) of the Philippines, more widely known as Metro Manila, has 17 local government units (LGUs). It is composed of 16 independent cities, classified as highly urbanized cities, and one independent municipality. NCR, with an area of 619.57 km2, has a population of 12,877,253, making it the second most populous region in the Philippines.

One of the most devastating flood disasters happened on September 2009 when Typhoon Ondoy (Typhoon Ketsana) struck southwest Luzon in the Philippines. Flood disasters caused by the continuous heavy rainfall affected 872,097 people throughout the entire Metro Manila region, causing 241 fatalities, 394 injuries and damaging 65,521 buildings (of which 12,562 were completely destroyed) (Sato & Nakasu 2011).

Conceptual framework

There are several conceptual frameworks for assessing the structure of flood risk. This study follows that of Davidson (1997), adopted by Bollin et al. (2003). This conceptual framework views risk as the sum of hazard, exposure, and vulnerability minus capacity measures as shown in Equation (1). This framework is very flexible and can be easily adapted to specific limitations of research such as data availability. This framework can also be easily updated whenever updated data are available. 
formula
(1)
The formulae used for defining the components of flood risk, namely hazard, exposure, vulnerability and coping capacity, are listed in Equations (2)–(5), where H= hazard index, E= exposure index, V= vulnerability index, C= coping capacity index; α, β, γ, δ= global weight for hazard, exposure, vulnerability and coping capacity, respectively; and ai, bi, ci, di= local weights for hazard, exposure, vulnerability and coping capacity, respectively. 
formula
(2)
 
formula
(3)
 
formula
(4)
 
formula
(5)
A set of indicators (Hi, Ei, Vi, and Ci) was selected based on availability of data and review of previous studies (Bollin et al. 2003; Siddayao et al. 2014, 2015) and these indicators are listed in Table 1. The global, local and indicator weights were calculated using the AHP by Saaty (1980).
Table 1

List of indicators and their corresponding descriptions

CriteriaIndicatorDescription
Hazard Flood depth (m) [H1Depth of flood for a 25-yr return period rainfall 
Total precipitation (mm) [H2Total precipitation calculated from the rainfall intensity duration frequency for a 25-yr return period 
Exposure Number of housing (unitless) [E1Number of housing units per municipality 
Locally sourced revenue (LSR) (PHP) [E2Real property tax + tax on business + other taxes + regulatory fees + service/user charges + receipts from economic enterprises 
Population density (person/km2) [E3Measurement of population per unit area of land 
Vulnerability Elevation (m) [V1Elevation in metres derived from the digital elevation model 
Poverty index (%) [V2Proportion of families with per capita income/expenditure less than the per capita poverty threshold to the total number of families 
Percentage of vulnerable population (%) [V3Ratio between vulnerable population (children aged 0–6, persons with disability, senior citizen, etc.) and the total population 
LSR dependency (%) [V4Locally sourced revenue/annual regular income 
Percentage of impermeable area (%) [V5Proportion of the total area with impermeable (paved) surface 
Coping capacity Literacy rate (%) [C1Percentage of the population ten years old and over, who can read, write and understand simple messages in any language or dialect 
Number of medical personnel (unitless) [C2Number of medical personnel per municipality 
Annual regular income (PHP) [C3Locally sourced revenue + internal revenue allotment (current year) + other shares from National Tax 
Disaster preparedness rating (%) [C4Disaster preparedness rating from the Government Assessment Report 
CriteriaIndicatorDescription
Hazard Flood depth (m) [H1Depth of flood for a 25-yr return period rainfall 
Total precipitation (mm) [H2Total precipitation calculated from the rainfall intensity duration frequency for a 25-yr return period 
Exposure Number of housing (unitless) [E1Number of housing units per municipality 
Locally sourced revenue (LSR) (PHP) [E2Real property tax + tax on business + other taxes + regulatory fees + service/user charges + receipts from economic enterprises 
Population density (person/km2) [E3Measurement of population per unit area of land 
Vulnerability Elevation (m) [V1Elevation in metres derived from the digital elevation model 
Poverty index (%) [V2Proportion of families with per capita income/expenditure less than the per capita poverty threshold to the total number of families 
Percentage of vulnerable population (%) [V3Ratio between vulnerable population (children aged 0–6, persons with disability, senior citizen, etc.) and the total population 
LSR dependency (%) [V4Locally sourced revenue/annual regular income 
Percentage of impermeable area (%) [V5Proportion of the total area with impermeable (paved) surface 
Coping capacity Literacy rate (%) [C1Percentage of the population ten years old and over, who can read, write and understand simple messages in any language or dialect 
Number of medical personnel (unitless) [C2Number of medical personnel per municipality 
Annual regular income (PHP) [C3Locally sourced revenue + internal revenue allotment (current year) + other shares from National Tax 
Disaster preparedness rating (%) [C4Disaster preparedness rating from the Government Assessment Report 

Data collection

For the values of indicators, most of the data were gathered and collected from government agencies such as the Philippine Statistical Authority (PSA), Philippine Atmospheric, Geophysical and Astronomical Services Administration (PAGASA), Department of Education (DepEd), Commission of Higher Education (CHED) and Bureau of Local Government Finance (BLGF). Several available spatial data from the LiDAR Portal for Archiving and Distribution (LiPAD), PhilGIS and Humanitarian Data Exchange (HDX) were also used.

As for the data needed for the calculation of the global, local and indicator weights, a survey questionnaire was distributed to 12 respondents comprising engineers, municipal risk reduction officers, academics, government employees and medical personnel. In order to integrate their individual responses, a geometric mean was calculated for each indicator.

AHP analysis

Analytical hierarchy process was used to calculate the weights for the different indicators that were considered in this study. In order to perform the AHP analysis, a total of 12 respondents were selected from different fields including academics, engineering, government employees, local disaster risk reduction officers, rescue personnel and doctors. The survey was launched using both printed questionnaires and online survey forms.

In AHP, a pairwise comparison method (PCM) is used to obtain the weight or priority vector of the criteria. Saaty (1980) employed a numerical scale from 1 to 9 in order to evaluate the relative importance between two criteria. The respondents' judgment is then transferred to a pairwise comparison matrix A. Each numerical value rij of A represents the relative importance of the ith indicator in comparison with the jth indicator. The numerical values satisfy the condition given by Equation (6): 
formula
(6)
After building the matrix A, a normalized pairwise comparison matrix was derived by dividing each value of rij by the sum of all values of that column. Finally, the relative weights (wAHP) vector was estimated by calculating the average values on each row of the normalized pairwise comparison matrix.
The AHP method makes it possible to check the consistency of the estimated weights. This is done with the consistency ratio (CR) shown in Equation (7): 
formula
(7)
where CI is the consistency index and is calculated using Equation (8): 
formula
(8)
where λmax is the largest eigenvalue (Malczewski 1999) of the matrix and n is the number of indicators. RI is a constant that depends on n as shown in Table 2. When CR <0.1, the evaluation is consistent, and reliable results can be expected from the AHP model.
Table 2

Random index (RI) adapted from Saaty (1980) 

n12345678910
Random Index (RI) 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 
n12345678910
Random Index (RI) 0.58 0.90 1.12 1.24 1.32 1.41 1.45 1.49 

Spatial mapping of flood risk

The calculated indicator weights from the AHP analysis is then used to spatially distribute the flood risk. Since the selected indicators have different units of measurement, an initial standardization is performed using the z-score method. This pre-processing method for the raw indicator values thus made the range of mean and standard deviation within 0 to 1.

After calculating the flood risk using Equation (1), the different levels of flood risk were spatially mapped in GIS using the raster calculator and overlay function.

RESULTS AND DISCUSSION

Computation AHP pairwise matrix and consistency ratio

Table 3 shows the calculated global, local and indicator weights from the AHP analysis. Based on the survey responses, exposure has the highest global weight with a value of 0.37. It is followed by vulnerability, hazard and coping capacity with values of 0.26, 0.21 and 0.16, respectively. For the local weights, the value is in the range 0.10–0.67. The indicator weight was calculated by multiplying the local weights with their corresponding global weights. The resulting indicator weights are in the range 0.0256–0.1998.

Table 3

Resulting global, local and indicator weights from the AHP analysis

CriteriaIndicatorGlobal weightLocal weightIndicator weightConsistency check
Hazard H1 α= 0.21 a1 = 0.67 0.1407 Not applicable 
H2 a2= 0.33 0.0693 
Exposure E1 β= 0.37 b1= 0.54 0.1998 λmax= 3.0036
CI = 0.0018
CR = 0.0031 
E2 b2= 0.28 0.1036 
E3 b3= 0.18 0.0666 
Vulnerability V1 γ= 0.26 c1= 0.42 0.1092 λmax = 5.2681
CI = 0.0670
CR = 0.0599 
V2 c2= 0.13 0.0338 
V3 c3= 0.22 0.0572 
V4 c4= 0.13 0.0338 
V5 c5= 0.10 0.0260 
Coping capacity C1 δ = 0.16 d1= 0.37 0.0592 λmax = 4.1458
CI= 0.0486
CR = 0.0540 
C2 d2= 0.20 0.0320 
C3 d3= 0.27 0.0432 
C4 d4= 0.16 0.0256 
CriteriaIndicatorGlobal weightLocal weightIndicator weightConsistency check
Hazard H1 α= 0.21 a1 = 0.67 0.1407 Not applicable 
H2 a2= 0.33 0.0693 
Exposure E1 β= 0.37 b1= 0.54 0.1998 λmax= 3.0036
CI = 0.0018
CR = 0.0031 
E2 b2= 0.28 0.1036 
E3 b3= 0.18 0.0666 
Vulnerability V1 γ= 0.26 c1= 0.42 0.1092 λmax = 5.2681
CI = 0.0670
CR = 0.0599 
V2 c2= 0.13 0.0338 
V3 c3= 0.22 0.0572 
V4 c4= 0.13 0.0338 
V5 c5= 0.10 0.0260 
Coping capacity C1 δ = 0.16 d1= 0.37 0.0592 λmax = 4.1458
CI= 0.0486
CR = 0.0540 
C2 d2= 0.20 0.0320 
C3 d3= 0.27 0.0432 
C4 d4= 0.16 0.0256 

The consistency ratios listed in the last column of Table 3 indicate that the integrated respondent's judgment is very consistent since the CR values are all less than 0.10. Since there are only two indicators for hazard, the consistency check is not applicable. This is because, as seen in Table 2, the RI for n= 2 indicators is equal to zero (0). With zero as a denominator, the value of CR will be undefined or undetermined. This only emphasizes that since there are only two choices, the responses will automatically be consistent.

Flood risk mapping

The calculated indicator weights were used to spatially lay out the components of flood risk using linear combination. Figure 1 shows the spatial maps of flood hazard, exposure, vulnerability and coping capacity of Metro Manila. Figure 1(a) shows that the areas near the rivers and some areas in Marikina and Pasig have the highest flood hazard. This map is very similar to the flood maps developed by Lagmay et al. (2017) since the values for the flood depth indicator have been derived from their results. In terms of flood exposure, Manila and Quezon City have the highest values while Pateros, Mandaluyong, Malabon and Navotas have the lowest exposure (Figure 1(b)). For flood vulnerability, most areas in Manila and Quezon City score the highest values as implied in Figure 1(c). As for the coping capacity, Figure 1(d) illustrates that Manila and Quezon City have the highest value. This can be attributed to the fact that these cities both have high annual regular income and a sufficient number of medical personnel.

Figure 1

Spatial map of Metro Manila: (a) hazard, (b) exposure, (c) vulnerability and (d) coping capacity.

Figure 1

Spatial map of Metro Manila: (a) hazard, (b) exposure, (c) vulnerability and (d) coping capacity.

Table 4 shows the average value of the calculated hazard, exposure, vulnerability, coping capacity and flood risk for each LGU. Values highlighted in bold are the maximum values for each column. It can be observed that Marikina has the highest value of hazard among the LGUs while Manila has the highest exposure. For vulnerability and coping capacity, Quezon City has the highest value. Due to the extremely high value of exposure for Manila, it is also the municipality with the highest risk. It can also be noted that although Marikina has the highest hazard, it has low value of flood risk since its exposure and vulnerability are also low. These findings suggest that although Marikina has the highest hazard, it is the municipality of Manila which has the highest possible flood risk, which was calculated considering physical, social, economic and ecological indicators.

Table 4

Mean values of criteria and flood risk per LGU

LGU nameHazardExposureVulnerabilityCoping capacityFlood risk
Manila 0.1018 0.6538 0.1623 0.1610 1.5552 
Mandaluyong 0.1196 0.1143 −0.0374 −0.0253 0.0027 
Marikina 0.2728 −0.1432 −0.1306 −0.0648 −0.4456 
Pasig 0.2127 −0.0095 −0.0146 0.0558 −0.2418 
Quezon City 0.1282 0.5873 0.1962 0.3711 0.9184 
San Juan 0.1429 −0.1667 −0.1117 −0.0719 −0.8209 
Caloocan 0.1012 0.1807 0.1462 0.0595 0.4302 
Malabon 0.2379 −0.2249 −0.1091 −0.0882 −0.6497 
Navotas 0.0402 −0.2110 −0.0941 −0.0989 −1.1159 
Valenzuela 0.1540 −0.1122 −0.1047 −0.0171 −0.7813 
Las Piñas −0.1134 −0.1453 −0.0856 −0.0670 −1.4405 
Makati 0.1319 0.0802 −0.0308 0.0791 −0.3430 
Muntinlupa −0.0535 −0.1206 −0.0421 −0.0181 −1.2176 
Parañaque 0.0065 −0.0550 −0.0961 −0.0314 −0.9654 
Pasay 0.0324 −0.0007 −0.0858 −0.0303 −0.7071 
Pateros 0.1534 −0.4421 −0.1323 −0.1882 −1.3088 
Taguig 0.1239 0.0150 −0.0883 −0.0321 0.3759 
TOTAL 0.1013 0.1805 0.0323 0.1039 0.0250 
LGU nameHazardExposureVulnerabilityCoping capacityFlood risk
Manila 0.1018 0.6538 0.1623 0.1610 1.5552 
Mandaluyong 0.1196 0.1143 −0.0374 −0.0253 0.0027 
Marikina 0.2728 −0.1432 −0.1306 −0.0648 −0.4456 
Pasig 0.2127 −0.0095 −0.0146 0.0558 −0.2418 
Quezon City 0.1282 0.5873 0.1962 0.3711 0.9184 
San Juan 0.1429 −0.1667 −0.1117 −0.0719 −0.8209 
Caloocan 0.1012 0.1807 0.1462 0.0595 0.4302 
Malabon 0.2379 −0.2249 −0.1091 −0.0882 −0.6497 
Navotas 0.0402 −0.2110 −0.0941 −0.0989 −1.1159 
Valenzuela 0.1540 −0.1122 −0.1047 −0.0171 −0.7813 
Las Piñas −0.1134 −0.1453 −0.0856 −0.0670 −1.4405 
Makati 0.1319 0.0802 −0.0308 0.0791 −0.3430 
Muntinlupa −0.0535 −0.1206 −0.0421 −0.0181 −1.2176 
Parañaque 0.0065 −0.0550 −0.0961 −0.0314 −0.9654 
Pasay 0.0324 −0.0007 −0.0858 −0.0303 −0.7071 
Pateros 0.1534 −0.4421 −0.1323 −0.1882 −1.3088 
Taguig 0.1239 0.0150 −0.0883 −0.0321 0.3759 
TOTAL 0.1013 0.1805 0.0323 0.1039 0.0250 

Note: Values highlighted in bold are the maximum values for each column.

Figure 2 shows the spatial map for the flood risk index of Metro Manila. The resulting flood risk index was standardized using the z-score method and was divided into five classifications, namely Very Low Risk, Low Risk, Moderate Risk, High Risk and Very High Risk. The range of values for this index was classified as shown in Table 5.

Table 5

Flood risk index classification

Range of flood risk indexClassification
>1.75 Very High Risk 
0.75 to 1.75 High Risk 
−0.75 to 0.75 Moderate Risk 
−1.75 to −0.75 Low Risk 
<− 1.75 Very Low Risk 
Range of flood risk indexClassification
>1.75 Very High Risk 
0.75 to 1.75 High Risk 
−0.75 to 0.75 Moderate Risk 
−1.75 to −0.75 Low Risk 
<− 1.75 Very Low Risk 
Figure 2

Index-based flood risk map for Metro Manila.

Figure 2

Index-based flood risk map for Metro Manila.

The percentage area of the LGUs under each flood risk classification are presented in Table 6. Values that are highlighted in bold indicate the highest percentage area, which identifies the flood risk classification of the greater area of the LGU. Only Manila is classified under Very High Risk and only Quezon City is under High Risk. Mandaluyong, Marikina, Pasig, Caloocan, Malabon, and Makati are under Moderate Risk. Low Risk was observed in the greater areas of San Juan, Navotas, Valenzuela, Las Piñas, Muntinlupa, Parañaque, Pasay, Pateros and Taguig. As a region, 42.69% of Metro Manila is under Moderate Risk.

Table 6

Percentage area of LGUs under each flood risk classification

LGU nameVery Low RiskLow RiskModerate RiskHigh RiskVery High Risk
Manila 0.00 0.00 0.03 44.09 50.13 
Mandaluyong 0.00 0.03 94.47 5.08 0.05 
Marikina 0.00 31.06 62.65 0.03 0.00 
Pasig 0.00 15.19 78.53 0.09 0.00 
Quezon City 0.00 0.01 41.63 49.93 4.21 
San Juan 0.00 69.20 30.64 0.12 0.05 
Caloocan 0.00 0.04 65.06 13.53 0.29 
Malabon 0.00 43.99 46.32 0.00 0.00 
Navotas 0.20 58.20 17.27 0.02 0.01 
Valenzuela 0.00 49.08 23.90 0.02 0.00 
Las Piñas 24.35 69.34 5.95 0.00 0.00 
Makati 0.00 1.18 98.48 0.26 0.01 
Muntinlupa 2.92 84.32 9.35 0.00 0.00 
Parañaque 0.42 73.62 25.14 0.00 0.00 
Pasay 0.00 51.05 48.36 0.01 0.00 
Pateros 0.00 98.60 1.40 0.00 0.00 
Taguig 0.00 88.40 6.19 0.08 0.00 
NCR 1.56 25.69 42.69 17.96 4.68 
LGU nameVery Low RiskLow RiskModerate RiskHigh RiskVery High Risk
Manila 0.00 0.00 0.03 44.09 50.13 
Mandaluyong 0.00 0.03 94.47 5.08 0.05 
Marikina 0.00 31.06 62.65 0.03 0.00 
Pasig 0.00 15.19 78.53 0.09 0.00 
Quezon City 0.00 0.01 41.63 49.93 4.21 
San Juan 0.00 69.20 30.64 0.12 0.05 
Caloocan 0.00 0.04 65.06 13.53 0.29 
Malabon 0.00 43.99 46.32 0.00 0.00 
Navotas 0.20 58.20 17.27 0.02 0.01 
Valenzuela 0.00 49.08 23.90 0.02 0.00 
Las Piñas 24.35 69.34 5.95 0.00 0.00 
Makati 0.00 1.18 98.48 0.26 0.01 
Muntinlupa 2.92 84.32 9.35 0.00 0.00 
Parañaque 0.42 73.62 25.14 0.00 0.00 
Pasay 0.00 51.05 48.36 0.01 0.00 
Pateros 0.00 98.60 1.40 0.00 0.00 
Taguig 0.00 88.40 6.19 0.08 0.00 
NCR 1.56 25.69 42.69 17.96 4.68 

Note: Values that are highlighted in bold indicate the highest percentage area, which identifies the flood risk classification of the greater area of the LGU.

This flood risk map can be used to assist decision makers in policy making, land-use planning, developing guidelines and countermeasures and flood disaster insurance. One concrete example would be to assess the adequacy of the available numbers of evacuation centers in each LGU. Areas classified under Very High Risk must have enough evacuation centers to accommodate affected residents. However, these evacuation centers must not be located in low-lying areas, which could also be verified using the resulting flood risk map. Otherwise, evacuation centers will forfeit their main purpose. Another possible application of the developed risk map will be to identify impacts of flooding on several sectors such as the education sector (Cadag et al. 2017), medical sectors and private sectors. Additionally, the developed flood risk map can also be used to appropriately assist stakeholders of the government in prioritizing budget and type of countermeasure, either structural or non-structural. Badilla et al. (2014) emphasized that understanding large-scale patterns in flood hazard and food risk in Metro Manila should be done with appreciation of the limitations of the underlying datasets, methods and models used.

CONCLUSION

This study used the combination of multi-criteria decision making (MCDM) with geographical information system (GIS) which allows the integrating of the four components of risk assessment (hazard, exposure, vulnerability, and coping capacity), wherein the physical, social, economic, and/or environmental factors can be considered. This comprehensive flood risk assessment was performed for the local government units of Metro Manila. A total of 14 indicators considering urban characteristics including physical, social, economic, and environmental factors were selected and the AHP method was applied to calculate the weights of each indicator. Among the four criteria weights, exposure was the highest with a value of 0.37, followed by vulnerability, hazard and coping capacity with values of 0.26, 0.21 and 0.16, respectively. Although the greater area of Metro Manila is under Moderate Risk, about 42.69%, there were areas under Very High Risk and High Risk. The results of this study are expected to aid in understanding flood hazard and risk in Metro Manila. Moreover, the resulting flood risk information can be used as a decision tool in policy making, land-use planning, developing guidelines and countermeasures and flood disaster insurance.

REFERENCES

REFERENCES
Apel
H.
Thieken
A. H.
Merz
B.
Blöschl
G.
2004
Flood risk assessment and associated uncertainty
.
Natural Hazards and Earth System Sciences
4
,
295
308
.
Badilla
R. A.
Barde
R. M.
Davies
G.
Felizardo
J. C.
Hernandez
E. C.
Ordonez
M. G.
Umali
R. S.
2014
Enhancing Risk Analysis Capacities for Flood, Tropical Cyclone Severe Wind and Earthquake for the Greater Metro Manila Area, Component 3 – Flood Risk Analysis
.
Philippine Atmospheric Geophysical and Astronomical Services Administration (PAGASA)/Geoscience Australia (GA)
.
Bollin
C.
Cárdenas
C.
Hahn
H.
Vatsa
K. S.
2003
Disaster Risk Management by Communities and Local Governments
.
Inter-American Development Bank
,
Washington, DC, USA
.
Boroushaki
S.
Malczewski
J.
2010
Using the fuzzy majority approach for GIS-based multi-criteria group decision-making
.
Computers & Geosciences
36
,
302
312
.
Cadag
J. R. D.
Petal
M.
Luna
E.
Gaillard
J. C.
Pambid
L.
Santos
G. V.
2017
Hidden disasters: recurrent flooding impacts on educational continuity in the Philippines
.
International Journal of Disaster Risk Reduction
25
,
72
81
.
Davidson
R. A.
1997
An Urban Earthquake Disaster Risk Index
.
Report No. 121, The John A. Blume Earthquake Engineering Center, Department of Civil Engineering
,
Stanford University
,
Stanford, CA, USA
.
EM-DAT
2018
The Emergency Events Database – Universite Catholique de Louvain (UCL) – Centre for Research on the Epidemiology of Disasters (CRED). https://www.emdat.be/emdat_db/ (accessed 28 August 2018).
Gall
M.
2007
Indices of Social Vulnerability to Natural Hazards: A Comparative Evaluation
.
PhD thesis
,
University of South Carolina
,
Columbia, SC
,
USA
.
Lagmay
A. M. F. A.
Racoma
B. A.
Aracan
K. A.
Alconis-Ayco
J.
Saddi
I. L.
2017
Disseminating near-real-time hazards information and flood maps in the Philippines through Web-GIS
.
Journal of Environmental Sciences
59
,
13
23
.
Malczewski
J.
1999
GIS and Multi-criteria Decision Analysis
.
Wiley
,
New York, NY
,
USA
.
Pornasdoro
K. P.
Silva
L. C.
Munárriz
M. L. T.
Estepa
B. A.
Capaque
C. A.
2014
Flood risk of Metro Manila barangays: a GIS based risk assessment using multi-criteria techniques
.
Journal in Urban and Regional Planning
1
(
1
),
51
72
.
Saaty
T. L.
1980
The Analytical Hierarchy Process: Planning, Priority Setting, Resource Allocation
.
McGraw-Hill
,
New York, NY
,
USA
.
Sato
T.
Nakasu
T.
2011
2009 Typhoon Ondoy flood disasters in Metro Manila
.
Natural Disaster Research Report
45
,
63
74
.
Siddayao
G. P.
Valdez
S. E.
Fernandez
P. L.
2014
Analytic hierarchy process (AHP) in spatial modeling for floodplain risk assessment
.
International Journal of Machine Learning and Computing
4
(
5
),
450
457
.
Siddayao
G. P.
Valdez
S. E.
Fernandez
P. L.
2015
Modeling flood risk for an urban CBD using AHP and GIS
.
International Journal of Information and Education Technology
5
(
10
),
748
753
.