Assessing the vulnerability of buildings to a tsunami in San Andrés de Canoa, Manabí, Ecuador an...

Running head: CANOA PTVA-3

Assessing the vulnerability of buildings to a tsunami in San Andrés de Canoa, Manabí,

Ecuador: an application of the PTVA-3 model

Liam Hartman, Oksana Bartosh, Filippo Dal’Osso

Abstract

San Andrés de Canoa (Canoa), located in the province of Manabí, Ecuador, is a coastal

community, with approximately 7,000 people, with an estimated 100,000 tourists per year. Due

to its geological proximity to the Carnegie Ridge, and the Nazca and South America tectonic

plates, the risk related to seismic activity is high. It is one of eighty-one communities in Ecuador

located to be in an area of tsunamigenic risk, identified in a report by the Ecuadorian

Oceanographic Navel Institute (INOCAR). There have been five major tsunami producing

earthquakes along the coast of Ecuador since 1906, the largest reaching a magnitude of M8.8

Richter (M). The Ecuadorian Oceanographic Navel Institute used the Tohoku University’s

Numerical Analysis Model for Investigation of Near-field tsunamis (TUNAMI) model to

conduct tsunami numerical simulations for the Ecuadorian coast. Using this simulation,

INOCAR was able to establish vulnerability levels for all the communities included in their

study.

The aim of this study is to access the vulnerability of buildings to damage from a tsunami

numerical simulation, used to identify vulnerable communities along the Ecuadorian coast. We

applied the Papathoma Tsunami Vulnerability Assessment Model – 3, to assess the vulnerability

of buildings for the community of San Andres de Canoa and produce thematic vulnerability

maps in regards to the projected tsunami scenario. The assessment allows us to make

recommendations about possible risk management and planning strategies for the community of

Canoa. Results show that the built form of Canoa has a high level of vulnerability to the

deterministic tsunami numerical simulation. This work has significant implications for

communities like Canoa along the western Ecuadorian coast.

1.0 Introduction

The Indian Ocean Tsunami of December 26, 2004 and the 2011 earthquake off the

Pacific coast of Tōhoku exposed the significance of tsunami threats to the world at large;

globally, nations have begun to take stock of their coastal vulnerability in order to safeguard their

own communities (Dall’Osso et al., 2012). These tsunamis were not unique, and it is known that

similar events have occurred in the past and will occur in the future (Dall’Osso & Dominey-

Howes, 2009). Tsunami warning systems, education, and disaster planning are becoming

common in at-risk areas around the world (Dall’Osso & Dominey-Howes, 2009). However,

detailed hazard, risk, and vulnerability assessments have not received the same amount of

attention (Dunning & Durden, 2013). Post et al. (2007) states that “the knowledge about

elements at risk, their susceptibility, coping and adaptation mechanisms are a precondition for

the setup of people centred warning structures, local specific evacuation planning and recovery

policy planning” (p.1). It is now imperative that tools that can forecast the physical impact of

TSUNAMI VULNERABILITY ASSESSSMENT AND CONSEQUENCE ANALYSIS

2

future tsunamis on our communities must be advanced and implemented (Dall’Osso et al., 2012).

Tools that can quantify vulnerability posed by tsunamis will be vital to emergency managers,

urban planners, and end users in order to implement mitigative measures that can reduce

vulnerability and risk and aid in developing response and recovery plans.

San Andrés de Canoa (Canoa) is located in the province of Manabí, centred along the

Coast of Ecuador (Figure 1). It is a coastal community, with approximately 7,000 people, with an

estimated 100,000 tourists per year, generally between the months of December and April

(Dayson Vite, personal communications, March 12, 2013; Ministerio del Turismo, 2012b).

During holidays and weekends, the population of Canoa increases to five times its general

population with tourists (Ministerio del Turismo, 2012b). Canoa is located in a complex

geodynamic location where the Carnegie Ridge interrupts the Nazca and South America tectonic

plates, which meet and collide forming a pit or trench that runs roughly parallel to the coast

between 50 and 70 km (figure 2) (Gutscher, Malavieille, Lallemand & Collot, 1999, Rentería,

Lizano, Benavidas, Arreaga, & Pino, 2011). The Carnegie Ridge extends 930km from the

Galapagos Islands to the Ecuadorian mainland, where it further extends an estimated 700km at a

constant dip of 25°–35° down to 200 km under the South American Plate (Gutscher et al., 1999,

Rentería et al., 2011). This extension inland is suggested by the seismic gap and the perturbed,

broad volcanic arc (Gutscher et al., 1999). The impact of the Carnegie Ridge collision on the

upper plate causes transpressional deformation, extending inboard to beyond the Ecuadorian

mainland volcanic arc with seismicity comparable to the San Andreas Fault system (Gutscher et

al., 1999). This location provides a main seismic activity for Ecuador, as it divides the Nazca

Plate, and slips under the South American continental Plate (Rentería et al., 2011). These

collisions are capable of generating large earth and submarine quakes (Rentería et al., 2011).

Figure 1: Map of San Andres de Canoa, Ecuador, South America


3

Figure 2: Geodynamic map for Ecuador (Collot, Agudelo, Ribodetti, & Marcaillou,

2008)

1.1 Record of tsunamis generating earthquakes in Ecuador

In the past century there have been five registered tsunamigenic earthquakes off the coast

of Ecuador (Figure 3): one in 1906 (magnitude of M8.8 Richter), one in 1933 (M6.9), one in

1953 (M8.3), one in 1958 (M7.8), and one is 1979 (M7.9) (Espinosa, 1992, Rentería et al.,

2011). The 1906 event in particular was one of the strongest tsunamigenic earthquakes recorded

in history (Cruz De Howitt et al., 2005; Espinosa, 1992; Rentería et al., 2011). All of these

earthquakes would have devastated most of the Ecuadorian coastal communities if those areas

were developed and/or inhabited at those times (Cruz De Howitt et al., 2005).


4

Figure 3: Locations of the previously registered tsunamis off the coast of Ecuador (Rentería et

al., 2011)

Figure 4 shows the location of several "seismic swarms" registered off the Ecuadorian

coast between 2009 to 2014 (Cruz De Howitt et al., 2005, Rentería et al., 2011). The seismic

swarms are areas of high concentration of earthquakes, ranging from M4 to M6, however, the

magnitude of these earthquakes were insufficient to generate a tsunami.

Figure 4. Seismic swarms near Canoa (Cruz De Howitt et al., 2005)


5

On the basis of historical tsunamigenic records (Rentería et al., 2011) has estimated the

tsunami probability on the coast Ecuador to be high, While Espinoza`s study (1992) estimated

the next submarine earthquake to be at least 7.5M, and located off the coast of Jama, Manabí

province, 36 km north of Canoa. Given today’s high population density and the low engineering

standards of most dwellings, the impact of a major tsunami on the coast of Ecuador could be

catastrophic (Cruz De Howitt et al., 2005).

1.2 Ecuadorian Tsunami Hazard Identification and Risk Assessment

The Ecuadorian Oceanographic Navel Institute (INOCAR) began a coastal hazard

research program in the early 1990’s, conducting laboratory and field work comprised of

collecting and analyzing nautical and topographic maps and charts published by INOCAR and

the Instituto Geográfico Militar (IGM) for the Ecuadorian coast (Espinosa, 1990, Espinosa, 1991,

Espinosa, 1992). This study identified tectonic environments and vulnerable coastal populated

areas along the Ecuadorian coasts (Cruz De Howitt et al., 2005; Espinosa, 1992; Rentería et al.,

2011).

Firstly, the study identified three tectonic environments, which form trench runing

parallel to the coast roughly 50 to 70 km west of the continental coast (Espinosa, 1991). The

first located north of the Carnegie Ridge, between latitudes 1ºN and 7ºN where the main

submarine topographic feature is the Malpelo Ridge (3° 50' 00" N, 81° 13' 00" W ), the second

tectonic place it south of the Carnegie Ridge, among 2ºS and 4ºS latitudes, facing the Gulf of

Guayaquil (3.0000° S, 80.5000° W), among 1ºN and 2ºS, and the third environment

characterized by elevations of the Carnegie Ridge (1.0000° S, 83.0000° W), which is hitting the

American Continental Plate against Manabí Province; this ridge appears on the surface in the

Galapagos hot spot and Islands an located about 930 km from the Ecuadorian mainland

(Gutscher et al., 1999). Following the course of the Nazca Plate to the east, the Carnegie Ridge

is embedding 700km below the central coast of Ecuador at a constant dip of 25°–35° down to

200 km under the South American Plate (Renteria et al, 2011). This extension inland is suggested

by the seismic gap and the perturbed, broad mainland volcanic arc (Gutscher et al., 1999). The

dip of the Carnegie Ridge is manifested in the shallow depth of the pit or trench off the coast of

Ecuador, other manifestations can be found on the active lifting of the beach area and coastline

between 1ºN to 6ºS 83.0000° W in Ecuador and Peru (Gutscher et al., 1999, Renteria et al,

2011).

Secondly, the study used the Tohoku University’s Numerical Analysis Model for

Investigation of Near-field tsunamis (TUNAMI) model (Imamura, 1997; Imamura et al., 2006)

as the main program for numerical simulation of tsunamis. The model combines TUNAMI-N1,

Numerical Analysis Model for Investigation of Near-field tsunamis (linear theory with constant

grids), TUNAMI-N2 (linear theory in deep sea, shallow-water theory in shallow sea and runup

on land with constant grids), TUNAMI-N3 (linear theory with varying grids), TUNAMI-F1

(linear theory for propagation in the ocean in the spherical co-ordinates) and TUNAMI-F2

(linear theory for propagation in the ocean and coastal waters) (Imamura, 1997; Imamura et al.,

2006).

The tsunami numerical simulation used a projected submarine earthquake of a magnitude

of 2 on the Imamura scale (Renteria et al, 2011), equaling at M8 (Bryant, 2008), at depths of 100


6

metres for all study sites. The study placed the epicentres at coordinates of 1.685° N, 79.0975°W

in the first tectonic environment in the province of Esmeraldas, the second was projection was

located at 2.85° S, 81° W, effecting Santa Elena and Guayas province and the third tectonic

environment is projected to be located at 0.5° S, 80.5° W, slightly below Canoa (0.467° S,

80.450° W), in Manabí province (Figure 5) (Renteria et al, 2011). These locations were derived

from mean locations of previous large to great seismic events (Renteria et al, 2011).

Figure 5: Tsunami numerical Simulation epicenter for the Coast of the Province of

Manabí (Renteria et al, 2011).

In this study we focus on the last tectonic environment, encompassing the province of

Manabí for this study. The projected epicentre of tsunami numerical simulation approximates

tsunami arrival between ten to thirty minutes of its generation with wave propagation at heights

fluctuating from 1.1 to 9.1 meters above sea level (Figure 6), depending on coastal proximity,

bathymetric slope and tide level for the surrounding coastal area (Renteria et al, 2011).


7

Figure 6: The heights of the tsunami wave likely reach the coast of the province of

Manabí (2011).

The study by INOCAR placed the M8 projected tsunami numerical simulation at 100

metres deep and 6 km off the coast of Canoa (Renteria et al, 2011). It is estimated to receive a

wave projected to reach 2 metres at Mean Sea Level (MSL) tide (Renteria et al, 2011). Canoa is

located in a coastal plain surrounded by elevations up to 100 meters high (Renteria et al, 2011).

The occupied land is relatively low and flat between 2 and 3 metres above mean high tide level

of 2 metres, and is bordered by the Canoa River to the north of the town (Espinoza, 1992).

Canoa’s vulnerability to flooding from the projected tsunami waves is estimated to be low, due

to the bathymetric slope leading up to Canoa from the earthquake’s, which makes the wave loses

energy due to the processes of bottom friction (Renteria et al, 2011). However, depending on the

state of the tide at the time that the phenomenon occurs, waves could increase with astronomical

high tide.

Lastly, the study undertook a geomorphologic analysis and identified 81 coastal

communities’ vulnerability along the Ecuadorian coast, including the town of Canoa (Cruz De

Howitt, Acosta, & Vásquez, 2005; (Renteria et al, 2011). The research completed community

inundation and evacuation maps which estimated levels of vulnerability for coastal communities

located close to beaches with little slope, with presence of swamps, islands, sandy accumulations

and estuaries (Cruz De Howitt et al., 2005; SNGR, 2012). The Inundation and Evacuation Maps


8

identified zones with high and low inundation probability at highest astronomical tide levels

(HAT), as to identify all potential vulnerable areas within the community (SNGR, 2012).

Canoa’s location comprises many of the aforementioned geomorphologic characteristics

and was rated to be at moderate risk (Renteria et al, 2011). Highest astronomical tide levels for

Canoa are measured at 3 metres above MSL (SNGR, 2012). The high inundation probability

zone identifies topography at one metre above HAT levels, while the low inundation probability

is at three metres above HAT (SNGR, 2012). The application of HAT level effectively raises the

projected waves to a height of 5 metres high. Canoa`s inundation and evacuation map can be

found in below in Figure 7.

Figure 7: Tsunami Inundation Map of Canoa, Manabí, Ecuador (SNGR, 2012).

.

The study conducted by INOCAR did not included a vulnerability assessment (Renteria

et al, 2011), which is a critical component of a tsunami risk analysis (Jelınek & Krausmann,

2008). A vulnerability assessment of coastal building and infrastructure is essential in order to

understand the potential implications of a tsunami. Canoa has been identified as having low to

moderate vulnerability in the INOCAR study. Our study targets the gap of a tsunami

vulnerability assessment in this region. Cruz De Howitt et al. (2005) identified the area of Canoa

as being exposed to the highest tsunami risk in Ecuador; yet, little is known about the

vulnerability of the vulnerability of buildings and infrastructure.


9

1.4 Aim of this work

This study aims to apply the Papathoma Tsunami Vulnerability Assessment-3 model

(PTVA-3) (Dall’Osso & Dominey-Howes, 2009) to assess the vulnerability of existing buildings

in the San Andres de Canoa to tsunami. The PTVA-3 Model calculates a Relative Vulnerability

Index (RVI) for every inundated structure.

This work relies on study by INOCAR, which carried out lab and field work that used nautical

and topographic maps published by INOCAR and Instituto Geográfico Militar (IGM) for the

Ecuadorian Coast and which conducted a tsunami numerical simulation (Renteria et al, 2011)..

This analysis is used in order to establish vulnerability for the study area. The results of this

study will be used to make a series of recommendations for emergency services, community

planners, and end-users. It will assist them in planning for and managing tsunami risk, inform

land-use zoning, buildings standards and codes, and emergency and evacuation planning.

2.0 Applying the Papathoma Tsunami Vulnerability Assessment Model (PTVA-3 Model)

2.1 Methodological Approach

This study applied the Papathoma Tsunami Vulnerability Assessment Model – 3 (PTVA-

3) (Dall’Osso et al., 2009) to examine the structural vulnerability to a projected tsunamigenic

hazard for the community of Canoa, Manabí, Ecuador. According to the authors, it concentrates

on the physical effects of a hazard and includes the identification of the elements at risk.

(Dall’Osso & Dominey-Howes, 2009)

This model is inherently qualitative in its approach, but uses quantitative data to define

the initial scenario parameters and to calculate the vulnerability scores. The PTVA-3 assessment

uses qualitative data in the forms of researcher field observations of the built form of Canoa,

along with building reference images which were then scored using a scoring rubric, and the

resulting quantitative data was used to calculate a relative vulnerability index for the buildings

within the study area. The information was then described in detail and presented using thematic

maps.

2.2 The Papathoma Tsunami Vulnerability Assessment Model – 3

The PTVA-3 Model was developed using detailed information about the impacts of

historic tsunamis and the results of numerous post-tsunami damage surveys (Dall’Osso &

Dominey-Howes, 2009; Dall’Osso et al., 2009a, Dall’Osso et al., 2009b; Dall’Osso et al., 2010;

Dominey-Howes and Papathoma, 2007; Dominey-Howes; 2010; Papathoma, 2003; Papathoma

& Dominey-Howes, 2003; Papathoma et al., 2003).

The PTVA-3 provides an effective means of identifying vulnerable buildings in locations

where tsunami fragility curves or more sophisticated engineering models are not available or not

fully validated (Dall’Osso & Dominey-Howes, 2009; Tarbotton et al., 2012).


10

Dall’Osso and Dominey-Howes (2009) identified and ranked engineering and

environmental attributes that were reported to be responsible for controlling tsunami damage to

building structures, producing a model that offers a robust framework to explore building

vulnerability without certified engineering assessment models.

The main building attributes are grouped in three different categories: the physical

attributes of the building; the environment surrounding each building; and the depth of water

expected at the building location (Dall’Osso et al., 2010). The PTVA-3 Model requires all of the

necessary building attributes to be entered as numerical scores which are assigned according to

their contribution to the overall building vulnerability (Dall’Osso & Dominey-Howes, 2009;

Dall’Osso et al., 2010). To the PTVA-3 Model, each building is equally important, and since its

value is based on the structural-functional service, it does not account for the economic value of

buildings, or on the value of their contents (Dall’Osso et al., 2010; Dall’Osso & Dominey-

Howes, 2009). Also, contributions of individual engineering attributes are weighted using pair-

wise comparisons between attributes (Saaty, 1986). Using this technique, the contribution made

by separate attributes to the structural vulnerability of a building can be compared via a rigorous

mathematical approach, avoiding biases and reducing to a minimum the inevitable subjective

component of every decision making process (Dall’Osso & Dominey-Howes, 2009).

The PTVA-3 has been successfully implemented in two coastal areas of Sydney,

Australia (Dall’Osso & Dominey-Howes, 2009) and field tested in the Aeolian Island of Italy

(Dall’Osso et al., 2010). In the Aeolian study, the PTVA-3 model outputs were qualitatively

compared with post tsunami damage data from the 2002 Stromboli Tsunami, which showed

fairly accurate results (Dall’Osso et al., 2010).

2.3 Adapting the PTVA-3 Model to the construction standards of San Andrés de Canoa

As was stated in Dall’Osso et al. (2010), the PTVA-3 was developed to be applied

anywhere, though, RVI calculations rely upon type of architecture to be relatively consistent, but

may need to be modified to specific circumstances. For this study of Canoa, the researcher had to

adapt the building material factor slightly, as Dall’Osso et al. (2010) did in their application in

Italy. This paper uses the same building material adaptation as Dall’Osso et al. (2010), and added

bamboo and thatched roofs to the building material list where timber is located, and added

‘poorly cemented bricks’ in lieu of single brick, as much of the buildings in Canoa use these

(INEC, 2010) and the materials Dall’Osso et al. (2010) outlines in Italy (Figure 7).

-1

-

0.5

0 +.05 +1

Material

(M)

Reinforced

concrete

Double

Brick,

Single brick

(Ecuador - Poorly

cemented brick)

Timber

(Ecuador - Bamboo/

thatched Roof)

Table 1: Scores given to the attribute “m” (building material). These scores have been modified

with respect to the original PTVA-3 Model in order to fit with the common construction

practices used in Canoa, and along the Coast of Ecuador.


11

Figure 8. Typical buildings located in Canoa (Ministerio del Turismo, 2012a)

2.4 Data Gathering

The field site of Canoa was selected using the SNGR Inundation and Evacuation Map for

Canoa, which identified inundation zones with high and low inundation probability at highest

astronomical tide levels (HAT) (figure 7) (SNGR, 2012). Highest astronomical tide levels for

Canoa are measured at 3 metres above MSL (INOCAR, 2013). The high inundation probability

zone identifies topography at an average of one metre above HAT levels, while the low

inundation probability is at an average of 3 metres above HAT (Renteria et al, 2011; SNGR,

2012). The tsunami numerical simulation projected waves to reach heights at 2 metres high

(Renteria et al, 2011). Therefore, in order to be congruent with the HAT levels used in the

inundation and evacuation map for Canoa, we used a wave height of five metres in order to

calculate relative vulnerability for the PTVA-3 model.

To complete the PTVA-3 for Canoa, this study relied upon field surveys including an

examination of the building structures within the target area using filed observations and ground

truthing field surveys to develop a relative vulnerability index to be applied to thematic maps in

order to display structural vulnerability within the study area of Canoa.


12

We undertook field surveys to ground truth 506 buildings identified as being exposed to

the selected tsunami scenario. During field surveys we acquired data about the building attributes

as required by the PTVA-3 model. These include attributes of the building structure (i.e. number

of stories, building material, ground floor hydrodynamics, foundation type, the presence of

moveable objects, shape and orientation and preservation condition), and attributes of the

building surroundings (i.e. building row, presence of natural barriers, seawall height and shape,

and height of any brick walls around a building). Once collected, data were used to calculate the

Relative Vulnerability Index (RVI) of each building. A full description of the details regarding

the calculation of RVI scores can be found in Dall’Osso et al., (2009a). While the PTVA-3

model uses a GIS platform for undertaking the calculation, we obtained the same result using an

excel spreadsheet. The calculated RVI scores were used to generate a thematic building

vulnerability map for the area of Canoa, where buildings having different vulnerability levelS are

represented with different colours.

3. Results

Field survey results indicate that all the buildings in this area typically have shallow

foundation (one metre or less); and are not protected by seawalls or natural barriers. Out of 506

buildings assessed in this study, 115 (23%) buildings have been classified as having very high

vulnerability, 82 (16%) high vulnerability, 86 (17%) average vulnerability, 205 (40%) moderate

vulnerability, and only 18 (4%) minor vulnerability (Table 2). This means that over 50% of the

exposed buildings have average to very high vulnerability to the selected tsunami scenario.

(Table 3).

RVI ranking Number and Percentage of buildings

Very High 115 (23%)

High 82 (16%)

Average 86 (17%)

Moderate 205 (40%)

Minor 18 (4%)

Total Buildings Assessed 506

Table 2: Distribution of Relative Vulnerability scores


13

RVI ranking Portion of buildings (Total Buildings 506)

Average to Very High 283 (56%)

Minor to moderate 223 (44%)

Table 3: Distribution of building Vulnerability

Figure 9 illustrates the relative vulnerability of the exposed buildings in Canoa. Most of

the buildings rated as ‘very high’ or ‘high’ vulnerability are located within high inundation

probability zone. This area also included 95% (or 82 structures) of all the buildings ranked as

‘average’ and 22% of ‘moderate’ vulnerable structures. This area is projected to have an

inundated depth of four meters (designated by the light-orange area on Figure 9). The buildings

located within the low inundation probability zone (designated by the yellow-green area on

Figure 9) have with the projected inundation depth of two metres, included 78% of the moderate

risk, 4% of the average risk and 100% of the minor risk buildings (Table 4).

Figure 9: Relative Vulnerability Map of Canoa


14

RVI rating (High Inundation) (Low inundation)

Very High 115 (100 %) 0 (0%)

High 82 (100%) 0 (0%)

Average 82 (95%) 4 (5%)

Moderate 45 (22%) 160 (78%)

Minor 0 (0%) 18 (100%)

Table 4: Building Relative Vulnerability within high and low probability inundation

zones.

The exposed buildings were classified on the basis of their main use in the following

classes: private residences, hotels & hostels, restaurants, bars & discos and critically important

buildings (e.g. police stations, hospitals, schools, municipal offices). These are the official ‘use’

of buildings of locally owned tourist driven businesses. Many of these buildings provided

multiple services, such as hotel, bar and restaurant, but in this study, we grouped them into their

official business role. Residential buildings hosting unregistered commercial activities were

classified as private residences.

Table 5 presents the distribution of RVI scores across the five building use classes. These

include 374 private residences, 49 hotels/hostels, 57 restaurants, 20 bars/discos and 6 critically

important buildings.

Use of Building #buildings Very High High Average Moderate Minor

Private Residences 374 (74%) 48 (13%) 59(16%) 66 (18%) 184 (49%) 17 (4%)

Hotels & Hostels 49 (10%) 6 (12%) 14 (29%) 16 (33%) 12 (24%) 1 (2%)

Restaurants 57 (11%) 44 (77%) 6 (11%) 3 (5%) 4 (7%) 0 (0%)

Bars & Discos 20 (4%) 17 (85%) 2 (10%) 1 (5%) 0 (0%) 0 (0%)

Important buildings 6 (1%) 0 (0%) 1 (17%) 0 (0%) 5 (83%) 0 (0%)

Table 5: Buildings divided by class with Relative vulnerability rating.

Critically important buildings included the one police station, one health clinic, one

school, one gas station, one municipal building, one cemetery, and the church. The police

station, schools, gas station, municipal building, cemetery, and the church were all rated as

moderate. Their elevation and location far back in the community, along with other building

vulnerability and protection factors, contributed to these ratings. Consequently, the health clinic,

located within the high inundation zone received a high vulnerability rating also based on the

same elements of the PTVA-3 model.


15

4 Discussion

Results show that large portions of Canoa’s buildings have high levels of vulnerability.

Most buildings are made of wood, bamboo, and other light materials, with roofs of corrugated

iron or zinc, tile or thatched leaves. Many of these dwellings do not comply with local law

construction requirements and safety standards, which further increases their vulnerability

(Rentería et al., 2011; INEC, 2010). Some of the houses are raised off the ground by two or three

feet and supported by wooden pillars, increasing hydrodynamics, but most cases did not and are

one to two storeys, in poor preservation condition (INEC, 2010). Additionally, most of the

buildings generally have wood, straw or dirt floors and use recycled materials such as

newspapers, cardboard, magazines; branches, asbestos, cans, and plastic as building materials

(INEC, 2010). Moreover, buildings in Canoa typically have shallow foundation levels, a

common rectangle building shape, which contributes to a poor hydrodynamics; and have further

increased vulnerability due to the lack of a seawall, or any type of natural barriers.

The PTVA-3 model results clearly show a generalised high vulnerability level of the

study area. Over 50% of the exposed buildings were classified as having average to very high

vulnerability scores. An overwhelming portion of the buildings (i.e. 197 buildings out of 506) to

have high to very high vulnerability, however 48 of the Very High rated buildings are

improvised restaurants known locally as ‘chozas’ or huts located on the main beach, and the

majority of them are only used often only occasionally for national holidays.

There are some anomalies to the pattern of vulnerability revealed in the findings. Ninety-

six percent of the buildings classified as average vulnerability are scattered throughout the high

inundation zone, a location that is predominantly occupied with high and very high vulnerability.

This has revealed some major differences in the construction of these buildings compared to the

general practice within the rest of the community, as their classification showed considerable

resiliency to the tsunami scenario. The differences in construction standards are reflected by the

PTVA-3 model scores. In most cases, buildings with lower RVI scores are generally built with

better standards for tourism purposes and have better preservation conditions. This is consistent

with what Calgaro and Lloyd (2008) observed in Thailand after the 2004 Indian Ocean Tsunami;

while smaller wooden structures were demolished by the tsunami waves, larger foreign owned

open concept/well-constructed buildings remained structurally intact.

Another interesting finding was the visible difference between the high and low probable

inundation zones, where one can see a remarkable difference in levels of vulnerability in

buildings. It is documented that buildings located within the low inundation probability zone and

faired considerably better. This visible rift in ratings is predominantly based on which row they

were located in, in combination with their elevation level; acquiring lower vulnerability levels

the further back a building was located.

The historic record of tsunami impacting Ecuador’s coast shows that 5 tsunamis have

occurred in the area, thus indicating a significant tsunami hazard threat for this region (Cruz De

Howitt et al., 2005; Espinosa, 1992; Rentería et al., 2011).


16

The main limitations of this work are:

- The tsunami hazard assessment is based on a single tsunami event triggered by

a submarine earthquake of 8.0M at a depth of 100 metres. This examples does

not represent the worst case scenario for Canoa and further research should

address this issue.

- The impact of cascade effects, such as the tsunami flow triggering landslides or

sandy liquefaction, is not considered.

- Tsunami flow velocity is not considered.

- The effect of debris is qualitatively considered only for large movable objects

such as cars or trucks through the attribute “mo” of the PTVA-3 model. The

impact of smaller debris trapped in the tsunami flow is not considered.

- The PTVA-3 model is solely a building vulnerability assessment tool. The

vulnerability of the population or other socio-economic impacts is not

considered in this work and should be addressed by future research.

5.0 Recommendations

Based on the PTVA-3 assessment results, the following recommendations have been

made to the municipality of San Vicente, where Canoa is located:

- The design of new buildings should comply with safety codes and standards for flood-

prone areas. Generally speaking, Canoa is known to be an area where building codes are ignored

and not enforced. The municipality needs to reinstate guidelines for floor layouts, heights of

buildings, numbers of floors, material, and orientation which are generally followed in most

urban centres of Ecuador. Solid materials such as reinforced concrete, steel, or 30m think

masonry walls should be preferred over timber, bamboo or other temporary materials. Deep

foundations and open ground floors would further decrease the vulnerability to inundation.

- The Municipality should make sure that public buildings, particularly those providing

important services such as schools, health clinics, utility and government buildings, have the

necessary engineering requirements to be fully operative during marine inundations caused by

tsunamis or storms.

- Financial support should be made available for retrofitting existing private buildings to

comply with the national construction and safety standards. It is suggested that the SNGR work

with the local community to reduce the vulnerability of those structures.

- Land use planners should ensure that existing natural areas along the coast are

maintained and new green areas are included in future urban plans. Natural barriers would not

only act as buffers against the hydraulic forces of tsunami inundation, but would add value to

area and help support the tourism sector.

- Tsunami evacuation maps and plans should be shared with the local community and

visibly posted in buildings that operate in tourism, recreation or cultural sector, such as

restaurants, hotels/ hostels, travel agencies and recreational or cultural centers. This is especially


17

important for this community, as it receives tourists from all over the world, and the temporary

residents of these buildings change on a regular basis.

- Tsunami emergency management educational activities should be organised in schools

and cultural centres.

- This study primarily focused on the structural vulnerability of Canoa’s built form.

Further research into the examination of risk perceptions, awareness and preparedness of a

tsunami or seismic hazard within the community of Canoa as well as emergency management

staff and responders, using either quantitative or qualitative methods, could greatly benefit the

SNGR as well as the community itself.

- Given the high tsunami risk for Ecuador, we recommend undertaking vulnerability

assessment using the latest PTVA model available or other validated assessment tools in

different Ecuadorian coastal locations.

- This study adopted the tsunami hazard assessment undertaken by INOCAR, which was

undertaken by running the TUNAMI numerical simulation of the tsunami generation,

propagation and inundation. We recommend repeating this study using an accurate DEM of the

study area to improve the accuracy of the hazard assessment, as elevations below 10 metres were

dubious and used only as averages.

6 Conclusions

Historical records show that Ecuador is exposed to relatively frequent tsunamis triggered

by earthquakes along the [insert name of subduction zone here]. The exposure of Ecuador coastal

areas has significantly increased in the last century due to intense urbanisation and elevated

population growth. Yet, in part for lack of resources, no comprehensive tsunami vulnerability

assessment has been undertaken anywhere in Ecuador. This work demonstrates how a

comprehensive building vulnerability assessment can be undertaken in Ecuador using the

available data and minimal resources. For the first time, we used the PTVA-3 model without GIS

support and generated thematic vulnerability maps for the area of Canoa.

Results confirmed that Canoa’s built environment is very vulnerable to tsunamis. Multi-

storey reinforced concrete buildings, with open ground floor plans or good hydrodynamic shape

were assessed as being less vulnerable than typical local dwellings in timber or bamboo, even if

most of the resilient buildings were located close to the beach, where inundation depth would be

larger.

This research informed the municipal disaster plan for Canoa, developed by the SNGR

and broad tsunami research in Ecuador. Results may be used to help local disaster and

emergency managers determine preparedness and response strategies, as well as to develop


18

suitable mitigation actions through an improved land use planning and new building codes and

regulations. This analysis of building vulnerability to tsunamigenic risk can enable local

authorities and emergency planners to focus their limited resources in the most effective way.

Moreover, this analysis can help the SNGR and other governmental or non-governmental

agencies tasked with the responsibility of managing and responding to actual disasters and

preplanning mitigative measures. The study results may enable various end-users to produce a

series of maps that can help those agencies respond better, by knowing where the most

vulnerable building and, potentially, persons, households and businesses are located (Dall’Osso,

et al., 2009; Dall’Osso, et al., 2009b; Dominey-Howes & Papathoma, 2003).

References

Arreaga, Patricia. (1996) Estudio de los tsunamis en la costa sur del Ecuador (golfo de

Guayaquil), Guayaquil, Departamento de ciencias del mar centro de alerta de tsunamis,

Instituto Oceanográfico de la Armada [INOCAR], Ecuador.

Bryant, E. (2008). Tsunami: the underrated hazard. Springer.

Calgaro, E., & Lloyd, K. (2008). Sun, sea, sand and tsunami: Examining disaster vulnerability in

the tourism community of Khao Lak, Thailand. Singapore Journal of Tropical

Geography, 29(3), 288-306.

Collot, J.-Y., W. Agudelo, A. Ribodetti, and B. Marcaillou , 2008. Origin of a crustal splay fault

and its relation to the seismogenic zone and underplating at the erosional north Ecuador–

south Colombia oceanic margin, J. Geophys. Res., 113, B12102,

doi:10.1029/2008JB005691, 2008. 005691, 2008.

Cruz De Howitt, M. A., Acosta, M.C., & Vásquez, N.E. (2005). Riesgos por tsunami en la costa

Ecuatoriana. [Tsunami hazards in the Ecuadorian coast]. Sangolquí, Sección Nacional

del Ecuador del Instituto Panamericano de Geografía e Historia.

Dall’Osso, F., & Dominey-Howes, D. (2009). A method for assessing the vulnerability of

buildings to catastrophic (tsunami) marine flooding, unpublished report, 139 pp.


19

Dall'Osso, F., Gonella, M., Gabbianelli, G., Withycombe, G., & Dominey-Howes, D. (2009a). A

revised (PTVA) model for assessing the vulnerability of buildings to tsunami

damage. Natural Hazards and Earth System Science, 9(5), 1557-1565.

Dall'Osso, F., Gonella, M., Gabbianelli, G., Withycombe, G., & Dominey-Howes, D. (2009b).

Assessing the vulnerability of buildings to tsunami in Sydney. Natural Hazards and

Earth System Science, 9(6), 2015-2026.

Dall’Osso, F., Bovio, L., Cavalletti, A., Immordino, F., Gonella, M., & Gabbianelli, G. (2010). A

novel approach (the CRATER method) for assessing tsunami vulnerability at the regional

scale using ASTER imagery. Rivista Italiana di Telerilevamento, special issue:

Geomatics technologies for coastal environment observation, 42(2), 55-74.

Dall’Osso, F., Maramai, A., Graziani, L., Brizuela, B., Cavalletti, A., Gonella, M., & Tinti, S.

(2010). Applying and validating the PTVA-3 Model at the Aeolian Islands, Italy:

Assessment of the vulnerability of buildings to tsunamis. Nat. Hazards Earth Syst. Sci.

Dominey-Howes, D., & Papathoma, M. (2007). Validating a tsunami vulnerability assessment

model (the PTVA Model) using field data from the 2004 Indian Ocean tsunami. Natural

Hazards, 40(1), 113-136.

Dunning, C. M., & Durden. S. (2013). Social vulnerability analysis: A comparison of tools.

Alexandria, VA: U.S. Army Corps of Engineers, Institute for Water Resources.

Espinoza, J. (1990) Posibles efectos de un tsunami en las costas de la península de Santa Elena -

Ecuador [Potential effects of a tsunami on the north coast of the province of Esmeraldas -

Ecuador]. Guayaquil, Departamento de ciencias del mar centro de alerta de tsunamis,

Instituto Oceanográfico de la Armada [INOCAR], Ecuador.


20

Espinoza, J. (1991) Efectos potenciales de un tsunami en la costa norte de la provincia de

Esmeraldas - Ecuador [Potential effects of a tsunami on the north coast of the province of

Esmeraldas - Ecuador]. Guayaquil, Departamento de ciencias del mar centro de alerta de

tsunamis, Instituto Oceanográfico de la Armada [INOCAR], Ecuador.

Espinoza, J. (1992) Informe de Evaluación del Riesgo de Tsunamis de las Poblaciones de la

Costa Central del Ecuador - Ecuador [Risk Assessment Report Tsunami Stocks Central

Coast of Ecuador - Ecuador]. Guayaquil, Departamento de ciencias del mar centro de

alerta de tsunamis, Instituto Oceanográfico de la Armada [INOCAR], Ecuador

Geist, E. L., Titov, V. V., & Synolakis, C. E. (2006). Tsunami: Wave of Change. Scientific

American, 294(1), 56-63.

Gutscher, M. A., Malavieille, J., Lallemand, S., & Collot, J.Y. (1999) Tectonic segmentation of

the North Andean margin: impact of the Carnegie Ridge collision. Earth and Planetary

Science Letters 168, 255–270

Instituto Oceanográfico de la Armada del Ecuador. (2013). Tabla de Mareas [Tide Table].

http://www.inocar.mil.ec/mareas/mareas.php

Instituto Nacional de Estadística y Censos. (2010). Plan de Desarrollo y Ordenamiento

Territorial Parroquia rural de San Andrés de Canoa. [Rural Land and Plan development

of San Andrés de Canoa]. San Vicente, Manabí.

Imamura, F. (1997). IUGG/IOC Time Project: Numerical method of tsunami simulation with the

leap-frog scheme. UNESCO IOC.

Imamura, F., Ozyurt, G., & Yalciner, A. (2006). Tsunami modelling manual. UNESCO IOC

International Training Course on Tsunami Numerical Modelling.

Jelınek, R., & Krausmann, E (2008). Approaches to tsunami risk assessment. Luxembourg:

European Communities, OPOCE.


21

Ministerio del Turismo. (2012a). Catastro de alojamientos, bares, discotecas y restaurantes de

San Andrés de Canoa. [Registration of accommodation, bars, clubs and restaurants of

San Andrés de Canoa]. Quito, Ecuador.

Ministerio del Turismo. (2012b). Provincia Manabí Cantón San Vicente Playa de Canoa.

[Canoa beach, municipality of San Vicente, Manabí]. Quito, Ecuador.

Papathoma, M. (2003). Assessing tsunami vulnerability using GIS with special reference to

Greece. Unpublished PhD thesis, Coventry University (UK).

Post, J., Zosseder, K., Strunz, G., Birkmann, J., Gebert, N., Setiadi, N., ... & Siagian, T. (2007,

July). Risk and vulnerability assessment to tsunami and coastal hazards in Indonesia:

conceptual framework and indicator development. In A paper for the international

symposium on disaster in Indonesia, Padang, Indonesia (pp. 26-29).

Rentería, W., Lizano, M., P., Benavidas, Arreaga, P. & Pino, L. (2010). Diagnóstico de la

amenaza tsunamigenica de las costas Ecuatorianas. [Diagnosis of the tsunamigenic

threat of the Ecuadorian coasts]. Guayaquil Instituto Oceanográfico se la Armada

(INOCAR).

Saaty, T.L. (1986). Axiomatic foundation of the Analytic Hierarchy Process. Management

Science, 32, 841-855

Secretaria Nacional Gestión de Riesgos. (2012). Mapa de inundación y ruta de evacuación. [Map

Flood and evacuation route map]. Guayaquil. SNGR.

Tarbotton, C., Dominey-Howes, D., Goff, J. R., Papathoma-Kohle, M., Dall'Osso, F., & Turner,

I. L. (2012). GIS-based techniques for assessing the vulnerability of buildings to tsunami:

current approaches and future steps. Geological Society, London, Special

Publications, 361(1), 115-125.

Assessing the vulnerability of buildings to a tsunami in San Andrés de Canoa, Manabí, Ecuador an...

Documents

Transcript of Assessing the vulnerability of buildings to a tsunami in San Andrés de Canoa, Manabí, Ecuador an...