Tsunami hazard and risk assessment for Alexandria (Egypt) based on the maximum credible earthquake

Journal Pre-proof Tsunami hazard and risk assessment for Alexandria (Egypt) based on the maximum credible earthquake Hany M. Hassan, C. Frischknecht, Mohamed N. ElGabry, Hesham Hussein, Mona ElWazir PII:

S1464-343X(19)30390-5

DOI:

https://doi.org/10.1016/j.jafrearsci.2019.103735

Reference:

AES 103735

To appear in:

Journal of African Earth Sciences

Received Date: 30 August 2019 Revised Date:

12 November 2019

Accepted Date: 9 December 2019

Please cite this article as: Hassan, H.M., Frischknecht, C., ElGabry, M.N., Hussein, H., ElWazir, M., Tsunami hazard and risk assessment for Alexandria (Egypt) based on the maximum credible earthquake, Journal of African Earth Sciences (2020), doi: https://doi.org/10.1016/ j.jafrearsci.2019.103735. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier Ltd.

1

Tsunami Hazard and Risk Assessment for Alexandria (Egypt) Based on the

2

Maximum Credible Earthquake

3 4 5 6 7 8 9 10 11

Hany M. Hassana, b,e *, C. Frischknechtc, Mohamed N. ElGabrya,b, Hesham Husseina,b, Mona ElWazird

12 13 14 15

a

National Research Institute of Astronomy and Geophysics, 11421 Helwan, Cairo, Egypt (Affiliation ID: 60030681). b North Africa Group for Earthquakes and Tsunami Studies (NAGET), Ne t40/OEA ICTP, Italy. c Department of Earth Sciences, University of Geneva, 13, rue des Maraichers, 1205 Geneva, Switzerland. d Department of Architecture, Faculty of Engineering, Mansoura University, Egypt. e

*

Department of Mathematics and Geosciences, Via Weiss 4, I-34127, Trieste, Italy. Correspondent author: [email protected]; [email protected]

Abstract

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

Throughout its history, the northern coast of Alexandria has experienced devastating earthquakes and tsunamis impacts from tsunamigenic sources in the Eastern Mediterranean. The most hazardous tsunami events for Northern coast of Egypt were related to the earthquakes of 365 in Crete with Mw8.5, 1222 in Cyprus with Mw 7 - 7.5, and 1303 in Rhodes Island with Mw8.0.The 365 and 1303 earthquakes were accompanied by tsunamis that resulted in widespread destruction and subsequent fatalities along the coast of Alexandria as evidenced by the available historical reports for the Mediterranean region, geomorphologic and paleo-tsunami investigations accomplished recently. In this work, for each tsunamigenic source, i.e., West Hellenic Arc, East Hellenic Arc, and Cyprian Arc, the maximum credible earthquake (MCE) is defined then modeled with NAMI-DANCE. The comparison of the tsunami inundation maps of MCE scenarios computed for each tsunami source shows that the East Hellenic arc (EHA) is the most hazardous source for the Alexandria coast, with a tsunami wave height of 5.5-6.0 m, while the Cyprian arc (CA) is the least hazardous source. The aggregated tsunami inundation map for Alexandria defines the furthest boundary between inundated and non-inundated lands and is associated with hazard levels based on water heights. This map shows that the coasts of Al Amereya district and Borg Al Arab city are expected to be more affected than the others districts. It also indicates that the west side of Alexandria, called Al Sahel Al Shamally, is expected to have less impact than the eastern side, because it is naturally protected by a barrier of carbonate material parallel to the shoreline. The exposure analysis shows that more than 85,000 people are located in the tsunami hazard zone (representing 15.8% of Alexandria governorate’s population). Regarding buildings and infrastructure, the most exposed district is Al Amereya with 65.8% of total exposed objects. Among the buildings, the residential ones are the most exposed and about 21% can be classified as highly vulnerable. In terms of risk assessment, the Al Amereya district is identified as the area with the highest tsunami risk.

42

Tsunami hazard and risk assessment; Maximum Credible Earthquake; Alexandria, Egypt; Vulnerability

43

assessment; Residential buildings; Risk mitigation

44 45

1.

46

Although tsunamis in the Mediterranean Sea are less frequent than in the Pacific and

47

Indian oceans, their potential threat on population, buildings, lifelines, and

48

infrastructures (either existing or planned) of the coastal areas (North Africa and

49

South Europe regions) cannot be overlooked. Highly destructive tsunamis were

Keywords:

Introduction

1

50

reported at some locations along the Mediterranean Sea coast, but only a few of them

51

have been acknowledged to have affected the Alexandria governorate on the north

52

coast of Egypt (Eckert et al., 2012). According to historical documentation, the most

53

hazardous tsunami events for the northern coast of Egypt were caused by the

54

earthquakes of 365, Crete of Mw8.5, the 1222, Cyprus of Mw7.5, and the 1303,

55

Rhodes island of Mw8.0 (Ambraseys et al., 1995; Ambraseys 2009). The 365 and

56

1303 tsunami earthquakes are known to have strongly impacted Alexandria in

57

particular. Both caused widespread destruction along the coast, as evidenced by the

58

available historical reports (e.g., Ambraseys et al., 1995) and recently discovered

59

geomorphological and paleo-tsunami evidence (Shah-Hosseini et al., 2016; Salama et

60

al., 2018) whereas the 365 AD tsunami killed more than 50’000 people (Shah-

61

Hosseini et al., 2016).

62

More recent tsunami events, such as the Zemmouri earthquake on 21 May, 2003 with

63

Mw = 6.9 in Algeria (Alasset et al., 2006) and the Bodrum-Kos earthquake on 20 July,

64

2017 with Mw = 6.6 (Heidarzadeh et al., 2017), have reminded us about the capability

65

of Mediterranean earthquakes to trigger tsunamis. These events have also shown the

66

need for a proper evaluation of the tsunami hazard and associated risks.

67

Therefore, the objective of this work is to carry out a risk analysis for the Alexandria

68

governorate. To do so, the work has been divided into five steps: a) identification of

69

tsunamigenic earthquakes in the eastern Mediterranean region which could cause

70

significant impacts along the Alexandria coast; b) tsunami numerical modeling for the

71

Alexandria coast that considers the Maximum Credible Earthquake (MCE); c)

72

building an tsunami hazard map by combining all the information about inundation

73

heights obtained from individual scenarios into an aggregated inundation scenario that

74

can be viewed as the worst virtual scenario; d) analysisze of the exposure in terms of

75

population and built environment, and the associated vulnerability of residential

76

buildings located inside the tsunami inundated areas; e) development of a tsunami risk

77

map for residential buildings located in the inundation zone.

78 79

2. Study Area

80

Alexandria is the second largest governorate in Egypt in terms of population, as well

81

as represents a major economic and cultural heritage center for the country. The

82

governorate is divided into two main cities, which are Alexandria and Borg Al Arab 2

83

cities. Alexandria city extends about 80 km along the coast of the Mediterranean Sea

84

and is subdivided into 6 main districts, from east to west: Al Montazah, Sharq,

85

Wassat, Gomrok, Gharb, and Al Amereya (Fig. 1).

86

Since the last strong tsunami event affecting the area (i.e. 1303 earthquake according

87

to the available reports), Alexandria city has grown, both in terms of population and

88

in terms of the number of buildings and infrastructures. The permanent population in

89

the Alexandria governorate is about 5.4 million according to the official report of the

90

Egyptian Agency for Public Mobilization and Statistics (CAPMAS, 2017). This

91

number increases drastically during the summer season, due to the large influx of

92

tourists. Alexandria’s beaches represent the main resort and destination for millions of

93

local and international tourists each year (Tourism Promotion Authority of Alexandria

94

Governorate, 2019 “http://www.alexandria.gov.eg/”). In addition, it comprises many

95

important historic monuments belonging to different civilizations, which inhabited

96

Alexandria during its long history. The city also includes the Alexandria and Dekhila

97

harbors, which are the largest in Egypt, handling 80% of Egypt’s imports and exports

98

(http://www.alexandria.gov.eg/Alexandria/default.aspx, 2014). About 40% of all

99

Egyptian industry is located within the governorate of Alexandria.

100 101

Fig. 1: Location of the study area with the districts of Alexandria city.

102

3

103 104

3. Seismotectonic settings and tsunamigenic earthquake sources in the Eastern Mediterranean.

105

The Eastern Mediterranean (EM) is a small oceanic basin known for its complex

106

tectonics. The geodynamical studies in the EM basin has pinpointed that the African

107

plate subducts underneath Eurasia plate along the Hellenic Arc (HA) at a rate of about

108

1 cm/year, while the Aegean Sea represents an extensional basin with opening rates in

109

the order of 3.5 - 4.0 cm/year (for further reading see McClusky et al., 2000 and

110

references therein). In the far eastern side of the EM, the Africa plate collides with the

111

Anatolia micro-plate along the Cyprian arc (CA). In both arcs, the ongoing conversion

112

between African and Eurasian plates has been evidenced using bathymetry,

113

seismicity, and other geophysical data (e.g., seismic, gravity, magnetic). One main

114

tectonic feature in the EM is the Mediterranean Ridge, which is located southwest of

115

the Hellenic trench, whose seismotectonic nature is still under debate (see Le Pichon

116

et al., 2002). From the available historical tsunami catalog (e.g., Ambraseys et al.,

117

1995; Ambraseys & Synolakis, 2010; Maramai et al., 2014 and Papadopoulos et al.,

118

2014), it appears that earthquakes and submarine landslides are the two major

119

tsunamigenic sources in the EM. Several tsunamis may have originated as a

120

combination between these two different sources. The remaining 10 % of tsunamis in

121

the EM are the result of volcanic eruptions (Hassan et al., 2013).

122

Based on the spatial distribution of seismicity and tsunami activity, and considering

123

present-day tectonic activity, many authors (e.g., Tinti et al., 2005; Yolsal et al., 2007)

124

have divided the EM region into three different tsunamigenic sources: i.e., West

125

Hellenic (WHA) and East Hellenic (EHA) arcs forming the Hellenic Arc (HA), and

126

the Cyprian Arc (CA), arranged from west to east (Fig. 2).

4

127 128

Fig. 2: Tsunamigenic earthquakes reported in the Eastern Mediterranean during the last 4000 years

129

ago (plotted from the catalog of Ambraseys & Synolakis, 2010).

130

According to the historical record of earthquakes, the HA can be considered as the

131

source that has generated the largest earthquakes and basin-wide tsunamis occurring

132

in the EM. Its tsunamis are understood to be the most hazardous source in the EM. It

133

is well known to have produced relevant destruction along the Egyptian coast, in

134

particular the earthquake tsunamis of 365 and 1303. The HA tsunamis can be

135

generated by large shallow earthquakes associated with thrust faulting. Many studies

136

have discussed the earthquake and tsunami history of the Hellenic Arc and evaluated

137

its potentiality to generate future activity. For example, Galanopoulos (1960),

138

Papadopoulos and Chalkis (1984) and Papazachos et al. (1986) investigated tsunamis

139

along the Greek coasts. Papazachos and Dimitriu (1991) found that the most

140

devastating tsunamis occurred in areas of shallow earthquakes caused by normal and

141

thrust faulting style. Papadopoulos et al. (2007) found and updated about 18 tsunamis

142

having their sources in the EHA.

143

The Cyprian Arc is the closest collision zone to the Egyptian coast, but is smaller,

144

slower and less active than the HA. Large earthquakes triggering local tsunamis have

145

occurred throughout the area’s history, but so far no significant basin-wide tsunami

146

event is known to have originated by this source.

147 148

4.

Tsunamigenic earthquake modeling for Alexandria

5

149

For each of the three EM tsunamigenic sources, we have modeled what is called a

150

maximum credible earthquake (MCE) scenario, which is equal to the maximum

151

historical earthquake magnitude plus 2σ = 0.5, where sigma is the global magnitude

152

error (Dominique & Andre, 2012). Panza (2017) has defined the MCE as “the largest

153

event physically possible, whose magnitude can be tentatively and until proven

154

otherwise set equal to the maximum historical magnitude plus some multiple of the

155

standard deviation”.

156

In general, it may be assumed for MCE a magnitude Mdesign = Mmax + γEM σM (Rugarli

157

et al., 2018; Rugarli et al., 2019) where γEM is a tuning factor, that can vary, from

158

region to region, in the range 1.5-2.5. The adopted Mdesign of for each selected

159

earthquake scenario in EM tsunamigenic sources is set equal to the maximum,

160

historical or instrumental, recorded magnitude (Mmax) increased of 0.5. This value

161

represents twice the central value of magnitude standard deviation (σM) at the global

162

scale that usually varies in the range 0.2-0.3 (Båth, 1973, p.111).

163

The MCE represents an upper limit for all the available earthquake models and its

164

consideration is based on the lessons learned from the earthquakes of Sumatra 2004

165

and Tohoku 2011 and the ensuing tsunamis where their magnitude largely exceeded

166

the expected maximum magnitude inferred from the available data.

167

The fault dimension for each MCE model has been obtained using the scaling

168

relationship of Blaser et al. (2010) (see equations 1 and 2) that has been developed for

169

reverse faults database located in subduction tectonic setting. The choice of this

170

relationship is based on its low standard deviation (0.14) relative to the other scaling

171

relations, as well as the similarity between the tectonic setting of the region for which

172

the relationship was developed and the eastern Mediterranean region. Moreover, this

173

relation was selected due to the high-quality score given to this scaling relationship in

174

the study of Stirling and Goded (2012). The orientation of the fault rupture is adopted

175

based on the understanding of the present day tectonic setting of the region.

176

In terms of length: Log10L=-2.81+0.62Mw

1

177

In terms of width: Log10W=-1.79+0.45Mw

2

178

with L = length, W = width, Mw = Moment magnitude

179

One of the most challenging tasks in earthquake scenario modeling is the treatment of

180

uncertainties, since each of the key parameters has uncertainty and natural variability,

181

which are often not explicitly quantified. A possible way to handle this problem is to

6

182

systematically vary the modeling parameters. A severe underestimation of the tsunami

183

hazard could occur by fixing a priori some source characteristics,therefore, a

184

parametric study should analyze the effects of various fault mechanism parameters

185

(i.e., strike, dip, rake, depth, etc.) on the output. Consequently, we made parametric

186

tests to select the parameters that could produce the highest impact on the city. These

187

parameters will be considered in the definition of the MCE-scenarios starting from

188

fault rupture parameters that have been proposed by different authors for each

189

tsunamigenic source. The MCE scenario represents the upper limit of all the

190

developed scenarios for each tsunami source.

191

A tsunami hazard assessment for Alexandria has been carried out using a scenario-

192

based approach that incorporated information about earthquake sources and available

193

topo-bathymetry data. For this purpose, we used the NAMI-DANCE code developed to

194

serve as a computational tool for tsunami modeling (see Yalçiner et al., 2006 and its

195

validation and update in Kian et al., 2014).

196

In this code, the initial condition for the linear shallow water wave equation is

197

computed following Okada's (1985) work, which is the numerical result of the elastic

198

seabed displacement of the co-seismic deformation. The tsunami modeling using

199

NAMI-DANCE allows analysts to install a set of pseudo-tide gauges at selected sites

200

and located at a water depth of approximately 10 m, as recommended by the manual.

201

These tide gauge recordings will be used to show the change in tsunami wave height

202

as a function of time, producing what is called marigram.

203

To reduce the computational power and time requested, linear and non-linear

204

spherical shallow water equations have been used as a function of the sea depth

205

(Table 1). For depths greater than 500 m we have adopted a linear shallow water

206

equation, while the shallower part the non-linear equation has been used to

207

realistically simulate tsunami propagation, as well as study possible local effects.

208

Each of the two equations, i.e., linear and non-linear, is applied for the depth where it

209

works most efficiently. The choice of depth threshold, or the boundary between them,

210

is based on our insights regarding the available bathymetry map of the region and on

211

parametric tests we have done. The role of roughness on the propagation of flow has

212

been considered byusing a specific Manning’s coefficient value, taken from Chow

213

(1959), depending on the environment as shown in Table 1. The Manning’s

214

coefficient represents the roughness or friction applied to the flow. In our modeling,

215

the Manning’s value is used to characterize the bottom roughness. The choice of the 7

216

Manning’s value will affect the tsunami inundation extent and the arrival time. For

217

example, an increased roughness of the bottom will cause a delay in the arrival time

218

of the tsunami front and reduce the inundation zone inland.

219

Table 1: Types of equation being used and values of coefficient of bottom friction Type of

Water equation

environment

Depth

Distance from

(m)

shoreline (km)

Manning’s value

Deep water

Linear

>500

36

0.025

Shallow water

Nonlinear

500-0

0

0.03

Built

Nonlinear

<0

<0

0.06

environment

220 221 222

4.1 Data needed for tsunami modeling Three essential ingredients are needed to simulate the tsunami events which are a)

223

earthquake rupture process (source information); b) bathymetry of the path to the site

224

of interest (path information), and c) topography of the site (inland propagation).

225

For the location and configuration of tsunamigenic sources in EM, we used published

226

data,e.g., Ambraseys & Synolakis, (2010); Maramai et al. (2014) and Papadopoulos et

227

al. (2014); National Geophysical Data Center (NGDC).

228

For tsunami modeling both high-resolution bathymetry and topographic data are

229

needed for modeling tsunami generation and offshore and inland propagation. Local

230

topography and the travel direction of the tsunami waves have a great influence on the

231

run-up effect from earthquake-generated tsunamis. A nested bathymetry grid of

232

different resolutions that covers the area of interest was generated.

233

The bathymetry data for the deeper part of the eastern Mediterranean basin has been

234

extracted from the global bathymetric model (GEBCO_2014 Grid, version 20150318;

235

Weatherall et al., 2015) released by the General Bathymetric Chart of the Oceans

236

(GEBCO) with 30 arc-second spatial resolution (~ 460 m).For the shallower part a

237

local bathymetry a 244 m grid resolution has been created using interpolation (i.e.

238

resampling) approach of GEBCO data (Fig. 3).

8

239 240 241 242

Tsunami propagation can be accurately evaluated by using both high-resolution

243

bathymetry and topography data, which are pre-requisite for simulating tsunami

244

waves and describing tsunami wave interaction with topo-bathymetric features. The

245

topographic map of Alexandria (Digital Elevation Model, DEM),needed for the

246

accurate estimation of the boundary of inundation zone and the tsunami flow depth

247

categories (tsunami wave height – topography = flow depth), was obtained using the

248

Shuttle Radar Topography Mission (Farr et al., 2007) of 30 m resolution (SRTM-23-

249

SEP-2014 - https://earthexplorer.usgs.gov/) (Fig. 4). Fig. 4 shows lowland areas (now

250

used for salt production) just adjacent to build environment in Alexandria (marked

251

with red color, below the mean sea level or < 0 m and located in the south of the study

252

area) in which a tsunami could be trapped in the channels connecting these low lands

253

(artificial lakes) to the Mediterranean. However, in this work we do not quantify such

254

trapping.

255 256 257

4.2 Results of Tsunami Simulation 4.2.1 West Hellenic Arc Scenarios (WHA) For the WHA the earthquake source for the tsunami is located in the western end of

258

the Hellenic Trench. It comprises two-segment sources running parallel to the west

259

part of the trench that was very likely responsible for the 365 A.D., Mw = 8.5

260

earthquake and the related tsunami (Stiros 2001; Stiros 2001; Stiros and

261

Papageorgiou, 2001),. The magnitude of the MCE scenario is set equal to Mw =9.0.

262

The initial tsunami condition for this first case is plotted in Fig. A2.

Fig. 3: Bathymetry map of Eastern Mediterranean (GEBCO, 2015).

9

263 264

Looking at the results of the WHA tsunami modeling adopting the MCE scenario, we

265

found that most of the tsunami wave energy expressed in terms of tsunami wave

266

heights were polarized towards the Libyian coast, while a relatively small portion of

267

the energy, with lower wave height, was directed westward and eastward (Fig. A3).

268

The arrival time of tsunami waves originated by WHA source at Alexandria

269

shorelines ranges between 80 - 85 minutes, with the leading wave expected to have

270

negative amplitude (Fig. 5). The maximum expected tsunami wave heights at

271

Alexandria range between 4.0 - 4.4 meters above sea level (MSL) meters. The

272

inundation map of the MCE scenario with the associated tsunami wave heights is one

273

of the main outputs of the tsunami modeling code (Fig. 6).

274

10

275

P1)

276

p6)

11

277 278 279 280 281

p13 Fig. 4: DEM (SRTM-23-SEP-2014) with 30 m resolution (https://earthexplorer.usgs.gov/) of Alexandria province with 2D profiles at selected points in East, middle and west of the city (P1, P6 and P13, respectively). Lakes or low land are marked with red color (elevation <0m). All 2D profiles are included in Fig. A1.

12

Table 2: Different fault parameters compiled from different authors used in the simulation for the WHA source and the selected parameters for the MCE scenario

Western Hellenic Arc Fault parameters

Tinti et al, (2005) Northern segment

Tinti et al, (2005) Southern segment

Hamouda (2010)

Hamouda (2006)

Stiros (2010)

Shaw (2011)

Pagnoni et al, (2015)

Papadimitriou& Karakostas (2008)

MCE

L (km)

206

233

233

105

105

100

130

160

500

W (km)

38

38

38

100

100

100

86

80

180

Strike (degree)

326

312

312

292.5

292.5

315

314

315

315

Dip (degree)

20

20

20

40

40

30

35

35

20

Rake (degree)

90

90

90

90

90

90

90

90

90

Slip (m)

10

10

12

16

16

20

17.5

9

10

Depth (km)

1

1

15

70

70

45

5

5

15

13

1 2

Fig. 5: Synthetic time series of tsunami wave propagation computed at different points along Alexandria coast for the WHA source by adopting MCE scenario, located at

3

water depth of about 10m.

14

4 5

Fig. 6: The maximum flow depth and wave height in land for Alexandria computed using MCE

6

scenario for WHA source.

7 8

4.2.2. East Hellenic Arc Scenarios (EHA) The earthquake of 1303, which happened offshore of Rhodes island, is the largest

9

recorded event available from the historical catalog (Ambraseys & Synolakis, 2010)

10

with an estimated magnitude Mw =8.0. In our assessment, we considered this

11

earthquake as a reference scenario for the EHA. The magnitude of the MCE scenario

12

is set equal to Mw =8.5. Table 3 lists the fault parameters adopted by different existing

13

studies to model the potential tsunami threats from the EHA source.

14 15

Table 3: Different fault parameters compiled from different authors used in the simulation for the EHA source and the selected parameters for the MCE scenario. Eastern Hellenic Arc Fault parameters L (km) W (km) Strike (degree) Dip (degree) Rake (degree) Slip (m) Depth (km)

Tinti et al., (2005) 190 35 300 20 90 5 1

Hamouda (2010) 190 35 300 45 90 8 20

Yolsal and Taymaz, (2010) 100 30 115 45 110 8 20

Pagnoni et al., (2015). 100 46.7 236 20 90 8 5

MCE 290 108 80 20 90 6 15

16 17

Basic characteristics of the tsunami propagation at 20 min snapshots are shown in Fig.

18

A4. Most of the tsunami wave’s energy produced from the modeling of the 1303

19

scenario is directed towards Alexandria; only a small amount of energy travels east

20

and west of Alexandria (Fig. A5). Fig. 7 indicates that the expected travel time for a

21

tsunami wave to reach Alexandria’s coast would range between 45 - 50 min after the 15

22

earthquake occurrence and the leading tsunami waves are expected to have negative

23

amplitude. The maximum computed tsunami wave height for the MCE scenario for

24

the EHA varies between 5.5 and 6.0. The inundation map (Fig. 8) for the EHA source

25

shows more extended inundated areas and higher tsunami waves (up to 6 m) than

26

those caused by the WHA source.

27 28

4.2.3 Cyprian Arc Scenarios (CA) The CA tsunamigenic source is considered as the least hazardous source in the eastern

29

Mediterranean basin based on the published historical earthquake catalog (Ambraseys

30

& Synolakis, 2010), which shows a minor activity. Though smaller in magnitude and

31

at a lower frequency than in the Hellenic arc, this source may be important for the

32

cities located along the western coast of Egypt (e.g., Port Said, Al Arish cities). The

33

CA source is tackled here by considering the MCE source model listed in Table 4.

34

The maximum reported earthquake at this zone in the available historical record is

35

about Mw =7.5. The MCE is set to Mw8.0. The maximum positive wave amplitudes

36

(water elevations) at every grid point computed during simulation duration (m) for

37

CA scenarios (Fig. A6:7) shows that extreme impacts could occur on the eastern side

38

of the Rosetta branch till Arish city. The first tsunami wave would arrive at the

39

Alexandria coast after 70 - 75 min from the earthquake occurrence with a negative

40

amplitude, as shown by the mariagrams (Fig. 9) and the snapshots (Fig. A6). The

41

maximum wave height obtained from the MCE scenario reached up to 2.6 - 3 meters.

42

The low wave height produced for the CA scenario, relative to the two other sources,

43

makes it the least hazardous source for Alexandria. The inundated area due to the CA

44

scenario (Fig. 10) is smaller than the two other sources.

45

Most of the tsunami wave energy originated by CA would be directed towards Suez

46

Canal, Port Said, Damietta, and Al Arish cities (Fig. A6 andA7 )

16

47 48

Fig. 7: Synthetic time series of tsunami wave propagation computed at different points along Alexandria coast from the MCE scenario from the EHA source.

17

49 50

Fig. 8: The maximum flow depth and wave height in land for Alexandria for the MCE scenario adopted

51

for EHA source.

52 53 54

Table 4: Different fault parameters compiled from different authors used in the simulation for the CA source and the selected parameters for the MCE scenario. Cyprian Arc Fault parameters

Hamouda, (2010)

Yolsal and Taymaz (2010)

Pagnoni et al. (2015)

Salamon et al., (2007)

MCE

L (km)

80

50

100

120

141

W (km)

30

25

46.7

40

70

Strike (degree)

305

305

280

300

305

Dip (degree)

35

35

20

15

35

Rake (degree)

110

110

90

90

110

Slip (m)

3

3

8

5

6

Depth (km)

15

15

5

10

20

55 56 57 58 59 60

4.3 An aggregated scenario using extreme tsunami wave height values for Alexandria In this analysis, an aggregated scenario was built using extreme tsunami wave height

61

values estimated through the modeling of the MCE scenarios. The aggregated

62

scenario for Alexandria was established by merging all wave height values estimated

63

at each site along the coast in one feature class using a geographic information system

64

(GIS) tool, and then the highest values were retained. The inundation extent is 18

65

obtained by propagating the water level from the shoreline inland, using a static

66

“bathtub-filling” approach. This means that all inland areas having an elevation less

67

than the water level on the shoreline (and being hydraulically connected to it) are

68

considered equally inundated. This process was undertaken using the same DEM with

69

30 m resolution.

70

The west coast of Alexandria, shows a less inundated area than the eastern side,

71

because it is naturally protected by a ridge of carbonate material that forms the

72

shoreline.

73

The aggregated scenario that defines the furthest boundary between inundated and

74

non-inundated land is presented in Fig. 11. Inside the inundated area, we estimated the

75

tsunami flow depth for different areas by taking into consideration the projection of

76

tsunami height inland and the local topography. We noted the fact that the boundary

77

between different tsunami flow depth categories inside the inundation map was not as

78

sharp as it appeared in the aggregated scenarios, but there was an area of overlapping

79

which represented the uncertainty in the obtained results. The inundation zone did not

80

show a uniform flow depth. The flow depth would depend on the initial tsunami wave

81

height inland and the topography of the area. Based on the obtained data, we divided

82

the inundated area into three hazard levels, as described in Table 5.

83

Table 5: Tsunami hazard category according to PREVIEW (2005).

Flow depth (m)

Low Hazard <0.5

Intermediate Hazard 0.5-2.0

High Hazard ≥2.0

84 85

According to the aggregated inundation scenario map (Fig. 11, Table 6), the potential

86

affected areas are located in the districts of Dekhila, Gomrok, and Al Montazah. The

87

area between the shoreline and the road running along the coast of Alexandria could

88

be considered as a highly hazardous zone. However, the west side of Alexandria’s

89

coast would be less hazardous due to the presence of the ridge.

90 91

Table 6: The maximum wave height and arrival time for MCE for the three tsunami sources Source

WHA

EHA

CA

4.0-4.4

5.5-6.0

2.6 – 3.0

80-85

45 – 50

70-75

Parameter Tsunami wave height (meter) Tsunami Arrival (minute)

92

19

93 94

Fig. 9: Synthetic time series of tsunami wave propagation computed at different points along Alexandria coast for the MCE scenario adopted for CA source.

20

95 96 97 98 99 100 101

5. Exposure and vulnerability analysis for Alexandria

102

(UNISDR, 2009), exposure is defined as “people, property, system, or other elements

103

(elements at risk) present in hazard zones that are thereby subject to potential losses.”

104

We consider that an element is exposed to a tsunami if it is present in the inundation

105

zone. For exposure analysis for buildings (residential, commercial, public service) and

106

infrastructure, a GIS formatted database for land use activities along Alexandria’s

107

coast with 4 km width has been used. Alternatively, the exposure analysis for people

108

in the tsunami inundation zone has been done using data obtained from the WorldPop

109

website (http://www.worldpop.org.uk), due to the absence of a uniformed dataset

110

from local sources. The dataset on population was produced in 2013 and provides the

111

number of people per grid square of 100m resolution. The database on buildings

112

contains information about building heights, number of units, number of floors, and

113

types of use.

114

According to the CAPMAS (2017), population in the Alexandria governorate has

115

reached ~5 million. Males represent about 51% of the total population, while about

116

49% are female. The trade, industry, and service sectors are the main activities and

117

represent about 32.7%, 30.9%, and 29.6% of manpower in Alexandria, respectively.

118

The rest of the labour pool is working in agriculture (6.1%) and others activities

Fig. 10: The maximum wave height in land for Alexandria for the MCE scenario adopted for CA source.

According to the United Nations International Strategy for Disaster Reduction

21

119

(0.8%). The number of illiterate people has decreased in 2017 and became less than

120

15% of the total population.

121 122

The zone of east Alexandria city is known as a residential zone and is considered to

123

be a destination for many local and international tourists because of its charming and

124

distinctive beaches. The zone of West Alexandria can also be considered as residential

125

(e.g., El Max, Dekhila and El Agamy) and contains many unique villas and cultural

126

heritage sites, as well as the west port of Alexandria. The zone of middle Alexandria

127

represents the old residential part of the city and includes museums, the library of

128

Alexandria, cinemas, and the University of Alexandria. Lastly, the coastal zone of

129

Alexandria is characterized by well-constructed buildings, such as hotels, towers, and

130

chalets. This coastal zone extends from the shoreline to hundreds of meters inland.

131

Also, it is subject to continuous maintenance and restoration activities due to its

132

economic value.

133

In the city of Alexandria, residential and industrial activities represent about 45.9%

134

and 18.9% of the total land use, respectively. The areas belonging to roads and train

135

services represent about 28.8%, while recreational activities and other services

136

represent about 3% only of the land use. Military areas represent 3.4% of the land use.

137

About 47.5% of Alexandria buildings are less than three floors, which is a relatively

138

high percentage. The most significant percentage of buildings with three stories or

139

less is presented in Al Amereya, while the least percentage is shown in the Sharq

140

(East) district 11.01% (Fig. 12).

22

141 142

Fig. 11: The aggregated scenario for Alexandria built from an envelope of the extreme values of tsunami flow depth produced by MCE scenarios from the WHA, EHA, and

143

CA sources.

23

100 80 60 40 20 0 Montaza District

East (Sharq) District

Central Gomrok (Wassat) (Customs) District District

West (Gharb) District

Ameriyah District

144 145 146 147 148 149

5.1. Populations inside the inundation area Data retrieved from the WorldPop database was combined with the tsunami

150

inundation map. The number of people in the inundation area (Fig. 11) is more than

151

800,000 . Fig. 13 shows the density of the exposed population.

152 153

5.2 Buildings, infrastructures and lifelines inside the tsunami inundation area Using the inundation map of different water height categories of the aggregated

154

scenario (Fig. 11), side by side with the available land use database of Alexandria, we

155

were able to extract the typology of buildings located within the expected flood area

156

by taking advantage of GIS tools. About 100,000 buildings (Table 7) are located in

157

this area. These buildings have different uses e.g., infrastructure, cultural heritage

158

sites, public services, lifelines, commercial, and residential. Table 7 shows the number

159

of buildings per district and the distribution per category. The first conclusion is that

160

the residential buildings are the most exposed buildings over all other uses (53%),

161

followed by public services (21%) and commercial buildings (about 14%) (Fig. 14).

162

The number of exposed residential buildings is the highest overall occupation in the

163

districts of Al Montazah, Al Amereya, and Borg Al Arab city, with the public services

164

and lifelines structures prevail over the other occupations in the affected buildings in

165

the districts of Sharq and Wassat. In the districts of Gomrok and Gharb most of the

166

exposed buildings belong to the commercial type (Fig. 12). The highest exposed

167

district is Al Amereya 65.8% of total exposed objects followed by Borg Al Arab city

168

(27.2 %) and Al Montazah (3%) (Fig. 15).

Fig. 12: The distribution of building with different number floors for Alexandria’s districts. Blue= less than 3 floors; Red= between 3 to 5 floors; Green= more than 5.

169

24

170 171 172 173 174

Fig. 13: The density of exposed population for the study area. Number of persons per 100 square meters.

Occupation

Table 7: Number and percentage of exposed buildings per district of Alexandria city. Borg Al AlAlTotal Sharq Wassat Gharb alMontaza Gomrok Almereya Arab 2239 139 11 20 126 612 1269 62

Infrastructure buildings Residential buildings Public service and lifeline buildings Commercial buildings

1537

53

20621

21

14144

14

9872

10

27080

99287

100%

27.2

100%

58

583

532

33580

16432

255

23

110

107

78

11732

8316

734

12

82

934

783

11156

443

Others

106

0

6

19

42

7866

1827

Total

2271

47

276

1769

2047

65291

percentage of buildings at risk

3

0.1

0.3

1.6

2.0

65.8

25

2

52723

1

175 176

%

2.3

Infrastructure

9.9 Residential

14.2 52.8 20.8

Public services and lifelines Commercial Others

177 178

Fig. 14: The percentage of the exposed buildings per category (over all districts).

0.0 2.8

0.3 1.8

2.1 Montaza

27.3

Sharq Wassat Gharb 65.8

Gomrok Ameriyah Borg Al Arab

179 180

Fig. 15: The percentage of buildings per district relative to the total number of buildings exposed

181

5.3 Vulnerability of the residential buildings in the tsunami inundation zone

182

The second important component in the risk assessment process is the vulnerability

183

analysis of a given element at risk. According to UNISDR (2009), vulnerability is

184

defined as “The characteristics and circumstances of a community, system or asset

185

that make it susceptible to the damaging effects of a hazard.” There are many aspects

186

of vulnerability arising from various physical, social, economic, and environmental

187

factors. The factors that control physical vulnerability related to a tsunami may

188

include poor design and construction of buildings, low-rise buildings, inadequate 26

189

protection of assets, lack of public information and awareness, limited official

190

recognition of risks and preparedness measures, and disregard for comprehensive

191

environmental management. This section focuses on physical vulnerability of

192

residential buildings.

193

Vulnerability of residential buildings is a function of a number of parameters that in

194

case of a tsunami hazard includes physical parameters, e.g., building construction

195

conditions, building height, and the presence of structural mitigation measures.

196

According to our knowledge, there are no tsunami mitigation defenses installed in the

197

study area, only wave breakers that are located in some places to reduce the impacts

198

of the seasonal sea surges that affect the shoreline of the city,. However, it should be

199

noted that these wave breakers proved inadequate during recent surges, which

200

happened in Alexandria (Soliman et al., 2014).

201

For the calculation of the physical vulnerability of residential buildings in Alexandria,

202

which are exposed to tsunamis, the following three parameters were used: 1) building

203

elevation, 2) building materials and condition, and 3) number of floors. Elevation and

204

number of floors were retrieved from the CAPMAS (2017) database that was used for

205

the first risk assessment for the city. We classify the residential buildings situated

206

along the coastal zone, up till 4 km inland, into three general classes: good, moderate,

207

and poor, based on a field survey for the building conditions and type of construction

208

materials. The buildings built from concrete, according to the national norms, which

209

show no cracks and are subject to continuous maintenance, are considered as good,

210

while the poor buildings are those built without management, built of masonry bricks,

211

and suffer from deteriorations.

212

Since it is difficult to perform a building classification for more than fifty thousand

213

exposed residential structures one by one (Table 7), due to limited financial and

214

human resources, we decided to attribute a specific building condition to an area,

215

where most of buildings (>80%) in that particular zone/area show a similar building

216

condition. Most of buildings located very close to the shoreline are considered to be in

217

good condition. Only the buildings located along the coast of Gomrok and Al

218

Amereya districts show intermediate building conditions (Fig. 16). This could be due

219

the continuous rehabilitation of coastal buildings because of their economic value and

220

revenue, as they serve as hostels for thousands of local people that visit Alexandria

221

during the summer season. On the other hand, most of the buildings in poor

222

conditions sit further away from the coast. 27

223

224 225 226 227

Following Eckert et al. (2012), each vulnerability parameter was attributed the same

228

weight, as no post-tsunami impact assessments on buildings exist on past events for

229

the area under study. Each residential building located in tsunami inundation area has

230

been attributed a vulnerability level, using a modified vulnerability matrix of Eckert et

231

al. (2012) based on the vulnerability parameters we selected. Fig. 17 a, b and c show,

232

as an illustration, the vulnerability assessment of residential buildings in Al Montazah

233

(Box A) and Al Amereya (Box B) districts.

Fig. 16: Classification of buildings conditions for the coastal area.

28

234 235

a)

29

b)

236

30

c)

237 238 239 240 241 242

The residential buildings of high vulnerability represent about 21% of the total

243

exposed buildings while the low vulnerablity buildings account for about 24% of the

244

total exposed buildings. The percentage of residential buildings in the medium

245

vulnerability class is about 55% of total residential buildings, as shown in Table 8.

Fig. 17: Physical vulnerability assessment for Alexandria’s residential buildings (a); Box A comprises Al Montazah district (b) while the box B comprises Al Amereya (c) districts.

246

31

247 248

Table 8: Number of residential buildings for each vulnerability class.

Vulnerability class

No of buildings

Percentage (%)

Low

11003

21

Medium

28954

55

High

12766

24

Total

52723

100%

249 250 251 252

5.4 Tsunami risk assessment for residential buildings As we have carried out vulnerability and hazard analyses, it could be of interest to

253

combine information and produce a risk map for the study area. In this section, we

254

analyzed the tsunami risk to the residential buildings by applying the risk equation as

255

defined by UNDRO (1991; Equation 3), which indicates that risk is a convolution of

256

hazard, vulnerability, and elements at risk. The tsunami risk to residential buildings

257

was estimated in a qualitative way using a risk matrix (Table 9) relating hazard and

258

vulnerability. Fig. 18 (a, b and c) shows the tsunami risk maps for two selected

259

districts, Al Montazah and Al Amereya. The risk maps indicate that the district of Al

260

Amereya shows the highest risk for residential buildings relative to the rest of the

261

districts, due to the exposure and the level of vulnerability of the residential buildings.

262 263 264

Risk = Element at risk*Hazard*Vulnerability

265

Table 9: Tsunami Risk Matrix.

266

Hazard Levels Low

Medium

High

Low

1

2

3

Medium

2

4

6

High

3

6

9

Vulnerability level

267 268

3

32

269 270

a)

33

b)

271 272 273 274 275

34

c)

276 277

Fig. 18: Tsunami risk map for Alexandria’s residential buildings (a); Box comprises Al Montazah

278

district (b) while the box B comprises Al Amereya (c) districts.

279 280

6. Discussion and Conclusions Effective tsunami mitigation for areas at risk needs close horizontal and vertical

281

cooperation between fundamental and applied sciences, and also between

282

international and local level authorities and policy makers. Tsunami hazard modelling 35

283

and risk analysis are necessary for a better understanding of the issue and in defining

284

appropriate mitigation measures.

285

We provided a detailed tsunami hazard assessment for Alexandria from three potential

286

sources located in EM basin using numerical modelling. For each tsunami source an

287

MCE, which is equal to the maximum historical earthquake magnitude plus 2σ, where

288

sigma is the global magnitude error, was defined then adopted for modelling the

289

tsunami impacts that could be generated at three different known tsunami sources

290

(i.e., WHA, EHA, and CA). After modelling each scenario alone, we combined all the

291

information about tsunami wave heights obtained from all single scenarios into an

292

aggregated scenario that can be viewed as the worst virtual scenario. Fault rupture

293

parameters were compiled from different authors and parametric tests carried out to

294

select the parameters to be considered in the MCE-scenarios. The MCE scenarios

295

trigger bigger tsunamis in terms of wave height and could be considered as

296

conservative scenarios. If we compared the results of this study to other studies that

297

have been conducted for the area of interest (e.g. Hamouda, 2006, 2010; Pagnoni et

298

al., 2015), our results are midrange. These differences can be linked to the

299

topographic data used, the modelling approach, and the fault parameters.

300

The inundation maps obtained for the MCE of the WHA, EHA, and CA sources show

301

that the more extensive potential impacts to Alexandria is due to the EHA scenario,

302

which triggers higher tsunami waves height and would inundate the largest area, as

303

also pointed out by Pagnoni et al., (2015). The CA scenario is a relatively low hazard

304

source for the tsunamis and ranked third in terms of its capability to impact extensive

305

areas in Alexandria. However, the tsunami wave propagates towards the Suez Canal,

306

Port Said, Damietta, and Arish cities. This then indicates the importance of this source

307

to other cities and should not be ignored when planning further developments along

308

the Egyptian coast.

309

According to the aggregated inundation map, the most affected areas in Alexandria

310

are the districts of Al Amereya and Borg Al Arab. The west of the city, called Al

311

Sahel Al Shamally, showed fewer impacts than the eastern side, as it is naturally

312

protected by a ridge of carbonate material parallel to the shoreline. Our exposure

313

analysis shows that the number of people located in the tsunami inundation zone

314

would be more than 800,000 in the off-season. In terms of buildings, residential

315

structures are the most exposed building type. The most exposed district is Al

316

Amereya, which is also in agreement with the study by Pagnoni et al. (2015), with 36

317

65.8% of total exposed buildings However, our numbers differ from existing studies

318

(e.g. Eckert et al., 2012; Pagnoni et al., 2015 and El-Hattab et al., 2018), due to our

319

approach and data used for the tsunami hazard assessment. The datasets used for

320

characterizing the elements at risk were taken from the most officially up to date

321

database available and therefore more up to date. Based on the available database for

322

the city of Alexandria, from the report of CAPMAS (2017) about the location, the

323

number of floors, and heights, as well as the building type developed by a general

324

field survey for residential buildings located along the coastal zone, we were able to

325

evaluate the vulnerability of residential buildings in the areas under risk. The presence

326

of about 79% of the total exposed buildings in the high and medium vulnerability

327

classes points to the importance of finding the proper procedure and knowledgeable

328

strategies for decreasing the vulnerability of residential buildings within the coastal

329

zone.

330

The risk maps indicated that the district of Al Amereya shows the highest risk for

331

residential buildings relative to the rest of the districts. This is due to the high number

332

of elements at risk and the level of vulnerability of the residential buildings in this

333

area. The high risk of the buildings in Al Amereya district highlights the importance

334

of developing proper prevention measures to safeguard human life in the settled area.

335

Priority for risk prevention and mitigation measures should be given to buildings

336

located within the high hazard zone. The presence of highly vulnerable buildings and

337

people inside the areas of high and intermediate hazard risk further raises the

338

importance of adopting structural and non-structural prevention measures for the coast

339

of Alexandria.

340

According to our knowledge, the Alexandria coast has neither artificial tsunami

341

defences, an early warning system, or evacuation and emergency plans, which

342

increases the vulnerability of people and assets along the coast, to tsunami hazard. In

343

the west of the Alexandria governorate, there is a natural barrier of carbonate ridge

344

parallel to the shoreline that acts as a natural defence. However, this barrier has

345

already been destroyed at some locations because it interfered with the sea view,

346

without fully considering its role as an ecosystem protection against tsunamis.

347

Most of the tsunamigenic earthquake sources known to affect Alexandria’s coastal

348

zone are far-field sources, while the potential of the continental margin of Egypt to

349

generate significant tsunami is not fully knownand requiresfurther investigation. This

350

margin occupies the southern part of the folded arc, forming the Mediterranean ridge, 37

351

and is characterized by a narrow continental shelf that extends for about 15 - 20 km

352

seaward. During the last decades, some intermediate to strong earthquakes have

353

occurred, e.g., the Mw 5.2 in 2012 (Abu El-Nader et al., 2013) and the offshore

354

Alexandria earthquake Ms~6.7 in September 1955 (Korrat et al., 2005). So far, no

355

signatures or historical documentation of tsunamis that could have been generated by

356

such earthquakes have been discovered, which may be due to the lack of previous

357

investigation. However, the proximity of earthquakes that could happen along this

358

margin from the sedimentary cone of the Nile poses the potential for tsunamis

359

triggered by slope failure (Salamon et al., 2007) or cascading events between

360

earthquakes and submarine landslides.

361

To define the MCE, we mainly used observations from past earthquakes. Additional

362

information about possible earthquake sources could be obtained from active faults

363

databases (e.g. European database of seismogenic faults, Basili et al., 2013) and

364

Morphostructural Zonation (MZ) analyses (Gorshkov et al., 2009). A recent study

365

using the MZ approach was applied for the northeastern part of Egypt providing

366

information on potential earthquake sources in the area (Gorshkov et al., 2019). This

367

approach could be applied to assess the capability of the continental margin of Egypt

368

to generate tsunamigenic earthquakes.

369

To conclude, this study can be considered as instalment in increasing the awareness

370

about tsunami risk for the city of Alexandria. However, further investigations should

371

be carried out in order to improve the risk assessment and the implementation of

372

mitigation measures. Attention could focus on the characterization of physical

373

vulnerabilities, not only for residential buildings, but also for critical facilities, such as

374

hospitals and schools, as well as infrastructure such as road networks and harbours. In

375

addition, preservation and maintenance of the ridge, which acts as a natural tsunami

376

barrier, should be promoted. Finally, the vulnerability of people should also be further

377

analysed in order to identify which mitigation strategies could be put in place in order

378

to reduce impacts on life.

379 380

Acknowledgement

381

We would like to thank Prof. Giuliano Panza, Prof. Fabio Romanelli and Ahmed

382

Osama for their support and help in improving this work. Grateful thanks to Hazem

383

BadrEldeen for the help and the census data provided. Similar thanks and appreciation

384

go to Professor Ahmet Cevdet Yalçıner, Head of North-Eastern Atlantic and 38

385

Mediterranean Tsunami Information Center of UNESCO for the provision of the user-

386

friendly Tsunami simulation package NAMI-DANCE with which the results of this

387

study were obtained. The CERG-C Program of the University of Geneva is also

388

thanked for providing the framework to carry out this project and the first author was

389

financially supported by a Geneva private foundation during his stay in Geneva. Last

390

but not the least, we would like to thank Mesha Richard for revising the paper

391

linguistically.

392

39

393 394

References Abu El-Nader, I. F., El Gabry, M. N., Hussein, H. M., Hassan, H. M., & Elshrkawy,

395

A. (2013). Source characteristics of the Egyptian Continental margin earthquake, 19

396

October 2012. Seismological Research Letters, 84(6), 1062-1065.

397

Alasset, P. J., Hébert, H., Maouche, S., Calbini, V., & Meghraoui, M. (2006). The

398

tsunami induced by the 2003 Zemmouri earthquake (Mw= 6.9, Algeria): modelling

399

and results. Geophysical Journal International, 166(1), 213-226.

400

Ambraseys, N. N., Melville, C. P., & Adams, R. D. (1995). The seismicity of Egypt,

401

Arabia and the Red Sea, pp. 201. ISBN 0521391202. Cambridge, UK: Cambridge

402

University Press, January 1995., 201.

403

Ambraseys, N., & Synolakis, C. (2010). Tsunami catalogs for the Eastern

404

Mediterranean, revisited. Journal of Earthquake Engineering, 14(3), 309-330.

405

Båth, M. (1973) “Introduction to Seismology”, Birkhäuser Verlag, Basel, 395 pp.

406

Basili R., Kastelic V., Demircioglu M. B., Garcia Moreno D., Nemser E. S., Petricca

407

P., Sboras S. P., Besana-Ostman G. M., Cabral J., Camelbeeck T., Caputo R., Danciu

408

L., Domac H., Fonseca J., García-Mayordomo J., Giardini D., Glavatovic B., Gulen

409

L., Ince Y., Pavlides S., Sesetyan K., Tarabusi G., Tiberti M. M., Utkucu M.,

410

Valensise G., Vanneste K., Vilanova S., Wössner J. (2013). The European Database

411

of Seismogenic Faults (EDSF) compiled in the framework of the Project SHARE.

412

http://diss.rm.ingv.it/share-edsf/, doi: 10.6092/INGV.IT-SHARE-EDSF

413

Blaser, L., Krüger, F., Ohrnberger, M., & Scherbaum, F. (2010). Scaling relations of

414

earthquake source parameter estimates with special focus on subduction

415

environment. Bulletin of the Seismological Society of America, 100(6), 2914-2926.

416

Chow, V. T. (1959). Open-channel hydraulics. McGraw-Hill civil engineering series.

417

Dominique, P., & Andre, E. (2012). Probabilistic seismic hazard map on the French

418

national territory. In Proceedings of 12 world conference on earthquake engineering.

419

CAPMAS (Central Agency for Public Mobilization and Statistics Arabic Republic of

420

Egypt) report (2017). Census of Population, Housing and Establishments, internal

421

report.

422

Eckert, S., Jelinek, R., Zeug, G., & Krausmann, E. (2012). Remote sensing-based

423

assessment of tsunami vulnerability and risk in Alexandria, Egypt. Applied

424

Geography, 32(2), 714-723.

40

425

El-Hattab, M. M., Mohamed, S. A., & El Raey, M. (2018). Potential tsunami risk

426

assessment to the city of Alexandria, Egypt. Environmental monitoring and

427

assessment, 190(9), 496.

428

Esri, Garmin International, Inc., U.S. Central Intelligence Agency, 2019, World

429

Administrative Divisions, 15 May

430

2019,https://www.arcgis.com/home/item.html?id=9bc93fabc8d94747a22e811642ca4

431

26d [accessed on 17 July, 2019]

432

Farr, T.G., Rosen, P.A., Caro, E., Crippen, R., Duren, R., Hensley, S., Kobrick, M.,

433

Paller, M., Rodriguez, E., Roth, L., Seal, D., Shaffer, S., Shimada, J., Umland, J.,

434

Werner, M., Oskin, M., Burbank, D., Alsdorf, D. (2007), The Shuttle Radar

435

Topography Mission, Rev. Geophys., 45, RG2004.

436

https://doi.org/10.1029/2005RG000183

437

Galanopoulos, A. G. (1960). Tsunamis observed on the coasts of Greece from

438

antiquity to present time. Annals of Geophysics, 13(3-4), 369-386.

439

GEBCO (2015), The GEBCO_2014 Grid, version 20150318,

440

https://www.gebco.net/data_and_products/gridded_bathymetry_data/gebco_30_secon

441

d_grid/ [accessed on 2017]

442

Hamouda, A. Z. (2010). Worst scenarios of tsunami effects along the Mediterranean

443

coast of Egypt. Marine Geophysical Research, 31(3), 197-214.

444

Hamouda, A. Z. (2006). Numerical computations of 1303 tsunamigenic propagation

445

towards Alexandria, Egyptian Coast. Journal of African Earth Sciences, 44(1), 37-44.

446

Hassan, H. M., Badawy., A., Yalçiner., A., Abdel-Hafez., N. (2013). Tsunami hazard

447

assessment of the Mediterranean-Egyptian coastal zone by using the Worst case

448

Credible Tsunami Scenario Analysis, petrophysical journal, 2013.

449

Heidarzadeh, M., Necmioglu, O., Ishibe, T., & Yalçiner, A. C. (2017). Bodrum–Kos

450

(Turkey–Greece) Mw 6.6 earthquake and tsunami of 20 July 2017: a test for the

451

Mediterranean tsunami warning system. Geoscience Letters, 4(1), 31.

452

Kian, R., Yalçiner, A. C., & Zaytsev, A. (2014). Evaluating the performance of

453

tsunami propagation models. Proc. Bauhaus Summer School in Forecast Engineering:

454

Global Climate Change and the Challenge for Built Environment, August 17–29,

455

2014, Weimar, Germany, 1-10.

456

Korrat, I. M., El Agami, N. L., Hussein, H. M., & El-Gabry, M. N. (2005).

457

Seismotectonics of the passive continental margin of Egypt. Journal of African Earth

458

Sciences, 41(1), 145-150. 41

459

Le Pichon, X., Lallemant, S. J., Chamot-Rooke, N., Lemeur, D., & Pascal, G. (2002).

460

The Mediterranean Ridge backstop and the Hellenic nappes. Marine Geology, 186(1),

461

111-125.

462

Maramai, A., Brizuela, B., Graziani, L., 2014. The Euro-Mediterranean

463

tsunami catalogue. Ann. of Geophys. 57, S0435; doi:10.4401/ag-6437.

464

Okada, Y., (1985), Surface deformation due to shear and tensile faults in a half-space,

465

Bull. Seism. Soc. America, 75, 1135-1154, 1985.

466

McClusky, S., Balassanian, S., Barka, A., Demir, C., Ergintav, S., Georgiev, I., &

467

Kastens, K. (2000). Global Positioning System constraints on plate kinematics and

468

dynamics in the eastern Mediterranean and Caucasus. Journal of Geophysical

469

Research: Solid Earth, 105(B3), 5695-5719.

470

Pagnoni, G., Armigliato, A., & Tinti, S. (2015). Scenario-based assessment of

471

buildings damage and population exposure due to tsunamis for the town of

472

Alexandria, Egypt. Natural Hazards & Earth System Sciences Discussions, 3(8).

473

Panza, G. F. (2017). NDSHA: robust and reliable seismic hazard assessment. arXiv

474

preprint arXiv:1709.02945.

475

Papadimitriou, E., and V. Karakostas. (2008). Rupture model of the great AD 365

476

Crete

477

Geophysica 56(2): 293–312.

478

Papadopoulos, G. A., & Chalkis, B. J. (1984). Tsunamis observed in Greece and the

479

surrounding area from antiquity up to the present times. Marine Geology, 56(1), 309-

480

317.

481

Papadopoulos, G. A., Daskalaki, E., Fokaefs, A., & Giraleas, N. (2007). Tsunami

482

hazards in the Eastern Mediterranean: strong earthquakes and tsunamis in the East

483

Hellenic Arc and Trench system. Natural Hazards and Earth System Science, 7(1), 57-

484

64.

485

Papadopoulos, G. A., E. Gràcia, R. Urgeles, et al. (2014). Historical and pre-historical

486

tsunamis in the Mediterranean and its connected seas: Geological signatures,

487

generation

488

j.margeo.2014.04.014.

489

Papazachos, B. C., & Dimitriu, P. P. (1991). Tsunamis in and near Greece and their

490

relation to the earthquake focal mechanisms. In Tsunami Hazard (pp. 161-170).

491

Springer Netherlands.

earthquake

in

mechanisms

the

south-western

and

coastal

part

impacts.

42

of

the

Mar.

Hellenic

Geol.,

Arc. Acta

DOI:10.1016/

492

Papazachos, B. C., Koutitas, C., Hatzidimitriou, P. M., Karacostas, B. G., &

493

Papaioannou, C. A. (1986). Tsunami hazard in Greece and the surrounding area. Ann.

494

Geophys, 4, 79-90.

495

PREVIEW (2005). Prevention, Information and Early Warning preoperational

496

services to support the management of risks, Damage assessment based on damage

497

intensity scales: service specification. http://www.preview-risk.com/. [accessed on

498

2017]

499

Rugarli, P., Amadio, C., Peresan, A., Fasan, M., Vaccari, F., Magrin, A., Romanelli,

500

F. and Panza, G.F. (2018). Neo-Deterministic Scenario-Earthquake Accelerograms

501

and Spectra: a NDSHA Approach to Seismic Analysis. Jia, J. and Paik, J.K (Eds)

502

Structural Engineering in Vibrations, Dynamics and Impacts, CRC press, Taylor &

503

Francis Group, Abingdon, UK, in press.

504

Rugarli,

505

viable alternative to seismogenic zones and observed seismicity for the definition of

506

seismic hazard at regional scale. Vietnam Journal of Earth Sciences 41 (4), 289–

507

304. Doi:10.15625/0866-7187/41/4/14233

508

Salama, A., Meghraoui, M., El Gabry, M., Maouche, S., Hussein, M. H., & Korrat, I.

509

(2018). Paleotsunami deposits along the coast of Egypt correlate with historical

510

earthquake records of eastern Mediterranean. Natural Hazards & Earth System

511

Sciences, 18(8).

512

Salamon, A., Rockwell, T., Ward, S. N., Guidoboni, E. and, Comastri, A., (2007).

513

Tsunami hazard evaluation of the eastern Mediterranean: Historical analysis and

514

selected

515

10.1785/0120060147.

516

Shah-Hosseini, M., Saleem, A., Mahmoud, A. M. A., & Morhange, C. (2016). Coastal

517

boulder deposits attesting to large wave impacts on the Mediterranean coast of

518

Egypt. Natural Hazards, 83(2), 849-865.

519

Shaw, B. (2011). The AD 365 Earthquake: Large Tsunamigenic Earthquakes in the

520

Hellenic Trench. In Active tectonics of the Hellenic subduction zone (pp. 7-28).

521

Springer Berlin Heidelberg.

522

Soliman, A., Elsharnouby, B., & Elkamhawy, H., (2014). Shoreline Changes Due to

523

Construction of Alexandria Submerged Breakwater, Egypt. Abstract book, ICHE

524

2014.

P., Vaccari,

modeling,

F.,

Bull.

Panza,

G.F.,

Seismol.

Soc.

43

(2019). Seismogenic

Am..

97

(3),

nodes

705-724,

as

a

doi:

525

Stirling, M., & Goded, T. (2012). Magnitude scaling relationships. Report Produced

526

for the GEM Faulted Earth & Regionalisation Global Componets, GNS Science

527

Miscellaneous Series, 42, 35.

528

Stiros, S. C. (2010). The 8.5+ magnitude, AD365 earthquake in Crete: Coastal uplift,

529

topography

530

International, 216(1), 54-63.

531

Stiros, S. C. (2001). The AD 365 Crete earthquake and possible seismic clustering

532

during the fourth to sixth centuries AD in the Eastern Mediterranean: a review of

533

historical and archaeological data. Journal of Structural Geology, 23(2), 545-562.

534

Stiros, S. C., & Papageorgiou, S. (2001). Seismicity of Western Crete and the

535

destruction of the town of Kisamos at AD 365: Archaeological evidence. Journal of

536

Seismology, 5(3), 381-397.

537

Tinti, S., Armigliato, A., Pagnoni, G., & Zaniboni, F. (2005). Scenarios of giant

538

tsunamis of tectonic origin in the Mediterranean. ISET Journal of Earthquake

539

Technology, 42(4), 171-188.

540

UNDRO (1991). Mitigating Natural Disasters. Phenomena, Effects and Options. A

541

Manual for Policy Makers and Planners, United Nations, New-York, 94p

542

UNISDR (2009), 2009 UNISDR Terminology on Disaster Risk Reduction, United

543

Nations, Genva, 35p,

544

http://www.unisdr.org/files/7817_UNISDRTerminologyEnglish.pdf [accessed on

545

2019]

546

Weatherall, P., K. M. Marks, M. Jakobsson, T. Schmitt, S. Tani, J. E. Arndt, M.

547

Rovere, D. Chayes, V. Ferrini, and R. Wigley (2015), A new digital bathymetric

548

model of the world's oceans, Earth and Space Science, 2, 331–345,

549

doi:10.1002/2015EA000107.

550

Yalçiner, A. C., Pelinovsky, E., Zaytsev, A., Kurkin, A., Ozer, C., & Karakus, H.

551

(2006). NAMI DANCE Manual. Middle East Technical University, Civil Engineering

552

Department, Ocean Engineering Research Center, Ankara, Turkey, http://namidance.

553

ce. metu. edu. tr/pdf/NAMIDANCE-version-5-9-manual.pdf.

554

Yolsal, S., & Taymaz, T. (2010). Sensitivity analysis on relations between earthquake

555

source rupture parameters and far-field tsunami waves: case studies in the Eastern

556

Mediterranean Region. Turkish Journal of Earth Sciences, 19(3), 313-349.

changes,

archaeological

and

44

historical

signature. Quaternary

557

Yolsal, S., Taymaz, T., & Yalçiner, A. C. (2007). Understanding tsunamis, potential

558

source

559

Mediterranean. Geological Society, London, Special Publications, 291(1), 201-230.

560

Yalçıner, A., Annunziato, A., Papadopoulos, G., Güney-Doğan, G., Gökhan-Güler,

561

H., Eray-Cakir, T., & Kanoğlu, U. (2017). The 20th July 2017 (22: 31 UTC) Bodrum-

562

Kos Earthquake and Tsunami: Post Tsunami Field Survey Report. Online report at:

563

http://users. metu. edu. tr/yalciner/july-21-2017-tsunami-report/Report-Field-Survey-

564

of-July-20-2017-Bodrum-Kos-Tsunami. pdf.

regions

and

tsunami-prone

mechanisms

in

the

Eastern

565 566

Websites:

567 568 569 570 571 572 573 574 575

http://www.alexandria.gov.eg/Alexandria/default.aspx, 2014, Alexandria our home.

45

Appendix A:

P1)

P2)

1

P3)

P4)

2

P5)

P6)

3

P7

P8)

4

P9)

P10

5

P11)

P12

6

P13 Fig 1: Eleven 2D profiles at selected points arranged from East to west of the city (P1:P13) see figure 4 of the manuscript. Distance is represented by horizontal axis while elevation is on the vertical one.

a)

7

b)

8

c)

d)

e) Fig. 2: Snapshots of tsunami wave propagation for the MCE scenario for the WHA source plotted every 20 min (a, b, c,d and e).

Fig. 3: The maximum positive wave amplitudes (water elevations) at every grid point computed during simulation duration (m) for MCE scenario for WHA source.

10

a)

b)

11

c)

d) Fig. 4: Tsunami wave propagation for the MCE scenario for the EHA source plotted every 20 min (a, b, c, and d).

12

Fig.5: The maximum positive wave amplitudes (water elevations) at every grid point computed during simulation duration (m) for MCE scenario for EHA source.

13

a)

b)

14

c)

d) Fig.6: Snapshots of the tsunami elevation fields computed for MCE scenario for the CA source (a, b, c, and d) .

15

Fig.7: The maximum positive wave amplitudes (water elevations) at every grid point computed during simulation duration (m) for MCE scenario for the CA source.

16

We provided a detailed tsunami hazard assessment for Alexandria using numerical modelling. Tsunami flow depth for Alexandria was computed using topography and tsunami wave height. Vulnerability of residential buildings in the areas under risk has been evaluated. Tsunami risk map for Alexandria indicate that the district of Al Amereya is the highest risk. There is a need of adopting structural and non-structural prevention measures for Alexandria.

Declaration of interests ☐ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: