Landslide Risk Assessment of the Santorini Volcanic Group

V. Antoniou

, S. Lappas

, Ch. Leoussis

and P. Nomikou

Faculty of Geology and Geoenvironment, National and Kapodistrian University of Athens, Panepistimioupoli, Zografou, Greece

National Cadastre & Mapping Agency S.A., Aigina’s Office, Athens, Greece

Keywords: Santorini, Greece, Μulticriteria Decision Αnalysis, AHP, Model Builder.

Abstract: The aim of this work is the assessment of landslide risk of the Volcanic Group of Santorini (Greece). The

methodology followed was based on the application of multi-criteria analysis (Multicriteria Decision

Analysis - MCDA) in a Geographical Information System (GIS) environment (ArcGIS 10.5). The original

data was converted to digital form, georeferenced in the national coordinate system GGRS’87, and

individual layers were processed through digitization. Furthermore, a geodatabase was created in order to

enrich the spatial information with the requisite descriptive information. The above was necessary for

further processing and analysis, which include the classification of factor’s data, according to the specific

requirements of the region. Model Builder, an ArcGIS tool, was used to produce a Model which resulted in

individual maps for each thematic layer, in addition to providing local authorities with an easy-to-use

adaptable tool for landslide susceptibility mapping. Finally, a ranking method was used to generate criterion

values for every factor. Then, each factor was weighted according to the estimated significance for causing

landslides and the final susceptibility map was produced, depicting vulnerable areas.

1 INTRODUCTION

Landslides are natural phenomena, which in many

cases turn into natural hazards, causing thousands of

casualties as well as extensive damages to

constructions and infrastructures. Varnes, IAEG

(1984) defined landslides as ‘almost all varieties of

mass movements on slope including some such as

rock falls, topples and debris flow that involve little

or no true sliding’. According to Brabb (1993),

approximately 90% of landslide damages can be

avoided by the early recognition of the problem.

Hence, there is a necessity for landslide hazard

assessment in different spatial scales (Pardeshi,

Autade & Pardeshi 2013).

An accurate susceptibility map can contribute to

landslide investigation and landslide risk

management/ assessment, helping in the prevention

and, if possible, the mitigation of the disasters. The

vulnerable areas can be identified directly and

indirectly with a method based on causative factors,

by analyzing the historical link between landslide-

controlling factors and the distribution of landslides.

Guzzetti et al (1999) makes the assumption that

landslide events may be repeated in the future due to

the existence of the same conditions that had

previously produced them. Therefore, susceptibility

assessments may help with the prediction of the

geographical location of upcoming events. Landslide

susceptibility does not forecast neither the time of

occurrence of a landslide nor the magnitude of the

destruction (Guzzetti et al. 2005).

Numerous techniques have been applied for

landslide susceptibility mapping such as inventory

based mapping, deterministic techniques,

probabilistic techniques, heuristic techniques,

statistical analysis techniques, and multi criteria

decision making techniques (Guzzetti et al. 1999;

Kouli et al. 2013; Pardeshi, Autade & Pardeshi

2013). All the above methods can be classified into

quantitative, semi-quantitative, and qualitative

(Kouli et al. 2013), of which the first critically

depends on expert opinions.

Most of the aforementioned techniques have

further evοlved in the GIS environment, especially

in the generation of causative layers, computation of

different analysis and assigning of weights,

integration of data, and production of landslide

susceptibility maps. Such GIS based models are

Weighted Overlay, Decision Tree model, Analytical

Hierarchy Process (AHP), and physically based

landslide hazard models.

These models can also be applied in landslide

affected areas in southern Europe and specifically in

Antoniou, V., Lappas, S., Leoussis, C. and Nomikou, P.

Landslide Risk Assessment of the Santorini Volcanic Group.

DOI: 10.5220/0006385801310141

In Proceedings of the 3rd International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2017), pages 131-141

ISBN: 978-989-758-252-3

131

Greece. According to the Landslide Hazard Zonation

Map of Greece (Koukis et al. 2005), which reflects

the recorded events in Greece, the majority of them

appear in the mainland, while many events occur in

the Aegean islands, which attract a large number of

tourist all year round.

For the current study, the area of the Santorini

Volcanic Group was chosen, as it presents a

spreading urbanization along with increasing tourist

visits in a strained environment due to caldera cliffs.

Few landslide events have occurred along the

caldera causing significant damages on paths and

roads and, rarely, causalities (eg. landslide in Oia,

2011).

To sum up, the aim of this study is the

assessment of landslide risk of the Volcanic Group

of Santorini (Greece) by applying the Analytical

Hierarchy Process (AHP) (a multi criteria decision

making process) in a Geographical Information

System (GIS) environment. Besides the benefit of

providing a landslide susceptibility map, local

authorities will also gain a handy tool in dealing

with landslide hazard.

2 STUDY AREA

The Volcanic Group of Santorini (Thera) is located

in the southern Aegean Sea (Figure 1) and currently

represents the most active volcanic field of the

Hellenic Volcanic Arc (Nomikou et al. 2013). Its

most recent large eruption, known as the Minoan

Eruption (Late Bronze), occurred 3600 years ago

and is particularly well known for the significant

impact on the Minoan civilization (Friedrich 2000).

The volcanic complex includes the islands of Thera,

Therasia, Nea Kameni, Palaia Kameni, and

Aspronisi, arranged in a circular shape.

The morphology of the Santorini Volcanic Group

is composed of: (i) the internal rocky and steep

slopes of Thera, Therasia, and Aspronisi islands,

forming the aforementioned caldera ring,

characterized by high morphological dip values that

approach vertical values in certain locations, and

consisting of impressive morphological

discontinuities, and (ii) the external sections of the

islands, characterized by smooth surfaces of

relatively low dipping angles and radial distribution

to the volcanic centre, representing the remnant

outer slopes of the volcanic cone. The unique

morphological plays an important role in landslide

rockfall occurrence and determine the landslide

hazard in a great extent (Antoniou & Lekkas 2011).

Druitt et al. (1999) described the geological

Figure 1: Location of the Volcanic Group of Santorini

(Cyclades, Greece).

evolution of the volcanic field of Santorini. The

young Akrotiri volcanic strata (650–550 ka BP) are

hydrothermally altered silicic tuffs and lava flows. A

stratocone complex emerged in the northern half of

the volcanic field between 530 and 430 ka BP.

Major explosive activity began around 360 ka BP.

Since that time, about 12 large (few cubic kilometers

or more) explosive eruptions have occurred,

alternating with periods of constructional intra-

caldera activity (Druitt 2014). Repeated effusion in

the northern part of the volcanic field constructed a

10 km

shield volcano (the Skaros shield) that by 54

ka BP had reached a height of 350 m above present-

day sea level. This volcano was capped, between 45

and 23 ka BP, by a 2 km

dyke-fed succession of

dacitic lavas (Therasia dome complex) that reached

a thickness of up to 200 m on the western flank of

the edifice. At 22 ka, a >10 km

dacitic explosive

eruption (Cape Riva eruption) collapsed the Skaros–

Therasia edifice, forming a caldera (Druitt &

Francaviglia 1992; Druitt et al. 1999; Fabbro, Druitt,

& Scaillet 2013). About 18 ky of reduced magmatic

output then ensued, prior to the 3.6 ka Minoan

eruption. The 22 ka caldera dominated the

morphology of northern Santorini in the late Bronze

Age prior to the Minoan eruption. Several eruptive

phases followed the birth of Kameni islands (Palaia

and Nea), both onshore and offshore, and the latest

GISTAM 2017 - 3rd International Conference on Geographical Information Systems Theory, Applications and Management

132

one occurred in 1950 AD (Pyle & Elliott 2006;

Nomikou et al. 2014 and references therein). The

LBA caldera is 10x7 km wide, comprises of three

flat-floored basins around the Kameni edifices, and

is connected to the sea by three straits (one to the

NW and two to the SW) (Nomikou et al. 2016).

There are no officially recorded landslides from

any local authority (i.e. Municipality of Santorini),

however a number of events have been found

through the use of bibliography.

Although landslide hazard is not uniformly

distributed along the caldera slopes (Lekkas 2009),

Stergianos (2016) has made a serious effort to

document as many landslides occurring throughout

the island as possible. Two main regions can be

distinguished due to frequent rockfall events:

Teleferik area and Athinios port (Lekkas, Alexoudi

& Lialiaris 2013; Lekkas 2009), with the lateral

movement causing damages to roads and vehicles.

Oia is another area where rockfalls occur (Lekkas et

al. 2010), and it is the only one where a fatal

incident was documented in 2011, in Oia port, when

a rockfall led to the loss of a tourist’s life. One last

event worth mentioning occurred in the Red Beach

area, south of Akrotiri, which was closed to the

public due to extensive rockfalls.

Most of the aforementioned events are located

along the caldera’s cliffs, as expected. It is of utmost

importance to note that the recorded landslides

mainly affect human activities and infrastructures,

such as residential areas and road/path networks.

Moreover, in many cases there was no exact location

of the events identified, but only a description of

them, which leads us to allocate them approximately

(Figure 2).

Figure 2: Approximate location of documented landslide

events found through bibliography.

3 METHODOLOGY

The Methodology that was used can be divided into

two stages.

In the first one, all necessary vector and raster

data were collected and imported in a Geographical

Information System, specifically in ArcGIS -

ArcMap 10.5 (https://goo.gl/k9tAW1). Then, once

the factors that would be taken into account were

defined, a categorization of the data according to

their susceptibility to landslides was made. Each

category was normalized to 100 per cent, so that

calibration would have the same scale in all factors.

In order to produce the maps that represent each

factor, Model Builder, an ArcToolbox tool of

ArcMap 10.5 (ArcGIS platform, ESRI) was used.

The Pairwise Comparison Method is used in

Stage 2 in determining the weights for the criteria

(Saaty 1980) and in producing the susceptibility

map. The criterion for choosing the pairwise

comparison matrix was that it takes the pairwise

comparisons as an input and produces the relative

weights as an output, and the Analytic Hierarchy

Process (AHP) provides a mathematical method of

translating this matrix into a vector of relative

weights for the criteria.

4 CONDITIONING FACTORS

In order to investigate the landslide risk of the

Santorini Volcanic Group, the morphology of the

area was taken into account because it influences

both initiation and runout of landslides.

For this reason, a Digital Elevation Model (DEM)

was used to derive topographic factors other than

simply elevation, including slopes, aspects,

curvature, and hillshade. The DEM was derived

from the area’s Orthophoto map (2012) of the

National Cadastre & Mapping Agency S.A., having

a ground resolution of 5m.

Apart from the morphology factors of the area,

the lithology, land cover, road network, mean annual

rainfall, drainage network, fault zones, and soil

thickness were used as follows (Table 1):

Landslide Risk Assessment of the Santorini Volcanic Group

133

Table 1: Conditioning factors that were taken into account

for landslide susceptibility mapping.

4.1 Lithology

Lithology is a major controlling factor for

landslides. In the present study the geological

formations of the Santorini Volcanic Group were

derived from the geological map of Druitt et al., 1999.

Formations were grouped in regards to the

stratigraphy, geotechnical characteristics, and the

stability of each formation.

Even though lava formations are considered as

hard and stiff rocks, due to the fragmentation they

are favorable to rock falls.

Loose and high erodible formations like scoriae

and scree are prone to landsliding, since they are

unconsolidated formations. Also, dykes, subvertical

sheet-like intrusion of magma which are located

mainly across the northern cliffs of the caldera, were

given a high risk value. On the contrary, tuff is an

impermeable formation, thus it is not likely that

landslides will occur. In the same category

metapelites are also included; although they

normally have good geotechnical characteristics,

they are weathered in the study area.

All the formations mentioned above dominate the

caldera cliffs, making them susceptible to landslides.

The rest of the Santorini Volcanic Group is formed

by volcanic formations, with the exception of the

Mesozoic limestones in the northeastern part of

Profitis Ilias, which are presented with a low risk.

4.2 Slope Gradient

As slope increases shear stress in unconsolidated soil

cover, increases as well. Generally, landslides are

not expected to occur on gentle slopes due to lower

shear stress (Ladas, Fountoulis & Mariolakos 2007).

According to statistical approach (Mpliona

2008), landslides in Greece occur mostly in 16-35

slopes, while >35

slopes include both landslide and

rockfall phenomena.

A slope map was processed from the 5m-DEM

(in degrees) with values ranging from 0° to 85°.

Slope values were classified into four classes

according to Table 1. The caldera cliffs and the

southeastern part of the island (composed of Profitis

Ilias limestones) are characterized by steep slopes.

On the contrary, most parts of the island are

characterized by gentle slopes (0

– 15

4.3 Land Cover

Land cover is considered to be an important

landslide-controlling factor since it affects the

hydrological conditions and the soil strength (Ladas,

Fountoulis & Mariolakos 2007). Additionally, urban

areas reduce infiltration, increasing runoff erosion.

In the Santorini Volcanic Group, settlements on top

of caldera cliffs add an extra weight, weakening the

bedrock structure and increasing the susceptibility of

landsliding. Sparsely vegetated areas are generally

prone to erosion and present greater instability.

In this study, The Corine Land Cover map 2012

(CLC2012) of the European Environment Agency

GISTAM 2017 - 3rd International Conference on Geographical Information Systems Theory, Applications and Management

134

(EEA, Copenhagen, 2012; http://www.eea.europa.eu)

was adopted for the assignment of the land use

classes, which was updated with the help of the

orthophoto map. The evaluation of cover types was

conducted according to Table 1. Highest values were

given to areas covered by urban fabric and bare

rocks which dominate the caldera cliffs and Kameni

Ιslands. Lower weights were assigned to natural and

semi-natural areas. Cultivations/Agricultural areas

distributed randomly in Santorini, were given

moderate values.

4.4 Distance from Road Network

Another evaluation factor related to the occurrence

of landslides is the distance from the road network.

It is considered that landslide susceptibility

decreases when increasing the distance from the

main roads located on steep slopes (Ladas,

Fountoulis & Mariolakos 2007).

Road construction is also related to extensive

excavations, application of static and dynamic loads,

vegetation removal etc. along natural and engineered

slopes. These landslide triggering actions (WP/WLI

1994) were considered in the design of the landslide

susceptibility maps by introducing a road network

buffer zones data layer.

The road network was derived from

OpenStreetMaps, an open data source platform

(https://goo.gl/ozDBYc), and updated with the help

of the orthophoto map. Roads in steep slope areas

increase the proneness to mass movements.

Documented landslides in Santorini island occurred

in areas close to road networks (either asphalt roads

or paths). Three distance classes, 3, 5, and 10 meters

were used according to the type of road as shown in

Table 1. Roads which appear inside the urban fabric

were not taken into account, as they were included

in the Land Cover factor.

4.5 Curvature

Curvature values represent the morphology of the

topography. They can be calculated in terms of plan

curvature (which is perpendicular to the direction of

the maximum slope), profile curvature (which is in

the direction of the maximum slope) and general

curvature (Ladas, Fountoulis & Mariolakos 2007).

In this study the general curvature was used since

it combines the characteristics of the first two.

Curvature was selected as a factor as it affects the

hydrological conditions of the soil cover (Ladas,

Fountoulis & Mariolakos 2007) and it was derived

from the 5m-DEM. Potentially, the soil cover on a

concave slope can contain more water and retain it

for a longer period than a convex slope, thus the first

ones are favorable to landslides.

In the Santorini Volcanic Group curvature ranges

from -653 to 633 m

-1

. However, the majority of

values are within -30 to 10 m

-1

and they are

classified based to natural breaks classification

method. Negative values of the general curvature

represent convex slopes while positive values

correspond to concave slopes. The first ones are

mostly observed on the caldera cliffs and at Profitis

Ilias hill, as well as along the drainage network.

Curvature was divided into 7 classes accordingly to

Table 1.

4.6 Mean Annual Rainfall

Rainfall is an important factor, because as water

passes through discontinuities, it widens the holes

and leads to their destabilization, triggering

landslides. Although the mean annual rainfall of the

surrounding area is about 460mm (SSW, 2015), in a

few extreme events rainfalls of approximately

570mm have been recorded, which they could not be

ignored.

To interpolate the mean annual rainfall across the

Santorini Volcanic Group, the Thira meteorological

station data (National Weather Service, 36° 25' 3'' N,

25° 25' 55'' E, altitude 37m, gauging period 1931-

2002) were processed and rainfall information was

calibrated with the altitude using the 5m-DEM

(Karamesouti 2011), dividing the area into zones per

200m of altitude (Table 1) based on low

precipitation gradient. High risk areas are located

mainly in the SE part of the Volcanic Group.

4.7 Distance from Streams

It is assumed that the seasonal flow regime of

streams and gullies in the basin presents noteworthy

erosive processes, leading to superficial mass

wasting phenomena in areas adjacent to drainage

channels (Kouli et al., 2013). Fluvial erosion is one

of the most common triggering factors of the

landslides (WP/WLI 1994), and it usually affects the

slope’s toe.

The drainage network was derived from the 5m-

DEM and then corrected using the orthophoto map.

It was classified in two distance classes; 20 meters

away from streams, as they present periodic flow

therefore reduced corrosivity and the rest of the

areas (Table 1). Streams are located mainly at the

eastern and southern part of Thera Ιsland and at the

western part of Thirasia Island, while the absence of

Landslide Risk Assessment of the Santorini Volcanic Group

135

streams along the caldera cliffs is noticeable, as well

as the presence of water erosion that formed narrow

channel-like canyons on volcanic rock surfaces.

4.8 Proximity to Fault Zones

Landslides can be produced by tectonic factors such

as thrusts and major faults, which create steep slopes

and sheared zones of weakened and fractured rocks,

leading to landslide potential (Ladas, Fountoulis &

Mariolakos 2007).

Fault lines were digitized in vector format from

the geological map of Druitt et al. (1999), where the

presence of the major NE-SW Kolumbo fault zone

in the northern part of Santorini island is noticeable.

The buffered layer was classified into two distance

classes; 50 meters away from faults, as they present

low activation rate and the rest of the areas (Table

1). Other major fault zones are not distinguishable

on the island due to the existence of several volcanic

formations.

4.9 Aspect

Considering the influence of the aspect it can be

assumed that it has an effect on the degree of

saturation and evapotranspiration of the slopes. It is

generally considered that the N and NW facing

slopes in Greece are prone to landsliding due to their

shadier and colder conditions that favor the

accumulation and preservation of soil moisture.

Thus, landslides are expected to be more common

on the North and NW-facing slopes, due to larger

water accumulation (Ladas, Fountoulis &

Mariolakos 2007).

The aspect map was derived from the 5m-DEM

and has been divided into two classes (Table 1).

Relatively the northern part of the study area is

dominated by high risk aspect slope faces.

4.10 Soil Thickness

The Soil factor corresponds to soil thickness and

was derived from the Soil Associations Map of

Greece, in a scale of 1:500.000 (Yassoglou 2004).

Due to the fact that soil thickness is less than 15m all

over the Volcanic Group, which is considered as not

prone to landslides, it was decided not to take the

value into account when calculating the final

landslide susceptibility map.

5 FACTOR ANALYSIS

Model Builder, an ArcToolbox tool of ArcMap 10.5

(ArcGIS platform, ESRI) was used to create the

spatial distribution of each factor. Model Builder is a

visual programming language for building

geoprocessing workflows to create, edit, and manage

geoprocessing models that automate those tools

(https://goo.gl/D4JlSL).

A toolbox was created, containing a unique

model for each factor (Figure 3). Each model

consists of a workflow that strings together

sequences of the appropriate geoprocessing tools,

feeding the output of one tool into another as input,

taking into account the respective data variables.

Having completed all the contributed factor

maps, the relative contribution of each one was

taken into account.

Landslides are influenced by many factors

(preparatory or triggering) which vary significantly

from region to region. It is therefore difficult to

determine each factor’s relative contribution in

landslide occurrence and assign the respective

weights.

Consequently, the multi criteria decision making

approach is of high importance in susceptibility

mapping. The Analytical Hierarchy Process (AHP)

is a popular semi-qualitative method which

introduces objectivity in weight assignment through

pair wise comparisons and relies on the judgment of

the experts to derive priority scales (Barredo et al.

2000, Ayalew et al. 2005. Akgun & Bulut 2007,

Saaty 2008).

AHP is divided into two stages: (a) construction

of a pair wise comparison matrix, and (b)

determining weights and obtaining overall priority.

Absolute numbers (from 1 to 9) were assigned to

each landslide related factor based on its relative

importance and comparison matrices were

constructed to compute the Consistency Ratio (CR)

and the Consistency Index (CI).

The resulting maps of 5m resolution for each

factor (Figure 4) were used as input in the Analytic

Hierarchy Process tool extAhp 20, developed for

ArcGIS by Marinoni (2014), in order to determine

the weight of each factor and to produce the

susceptibility map (Table 2).

In this table, factors are listed in order of

importance. For example, lithology is the most

important one, slope is the second one, etc. The

numbers in each cell correspond to the degree of

preference between the two relevant factors. In the

last column the assigned weights for each factor and

the computed consistency ratio (CR) are shown.

GISTAM 2017 - 3rd International Conference on Geographical Information Systems Theory, Applications and Management

136

Figure 3: The models that created through Model Builder (ArcMap 10.), in order to produce the final map for each factor.

6 THE SUSCEPTIBILITY MAP

All the aforementioned factors processed with the

AHP methodology resulted in the generation of the

Landslide Risk Assessment Map of the Santorini

Volcanic Group, in which the study area is classified

according to the degree of landslide risk, pinpointing

the problematic regions for further analysis (field

observations, detailed mapping, etc.) (Figure 5).

The final map illustrates that the spatial

distribution of landslide risk was classified into five

categories: very high, high, moderate, low, and very

low. Classification was based on a frequency

histogram of landslide risk values, of which the

higher the value, the more susceptible the area is to

landsliding.

Generally, caldera cliffs display high risk values.

The internal northern cliffs of Thera Island and the

eastern ones of Thirasia Island display high landslide

risk due to the effect of the three most important

factors (lithology, slope, and land cover), with very

high risk values to be scattered along Thera’s cliffs.

Moderate to high values occurred mainly on Nea

Kameni island (slope and lithology factors), along

Landslide Risk Assessment of the Santorini Volcanic Group

137

the caldera rim (land cover and road network

factors) and scattered throughout Thera and Thirasia

island due to land cover and drainage network. Low

and very low risk areas dominate the rest of the

Santorini Volcanic Group.

7 CONCLUSIONS

In order to produce a landslide risk assessment map

of the Santorini Volcanic Group, nine causative

factors were taken into account by using the AHP

methodology in a GIS environment.

The susceptibility map shows that landslide risk

is not uniformly distributed along the caldera cliffs

and it is compatible with the distribution of

Figure 4: Contribution factors for Santorini susceptibility mapping, categorized according to the normalization rates of

Table 1.

GISTAM 2017 - 3rd International Conference on Geographical Information Systems Theory, Applications and Management

138

Table 2: The contribution factors and the corresponding weights, according to the AHP method.

documented past landslide events. The main factors

which contribute to the phenomena are the

following:

- Geotechnical conditions of the lithological

composition of the geological formations, which

vary perpendicularly along the caldera.

- The large morphological dip values, although

they are not uniformly distributed along the

caldera.

Human interventions with negative impact, which

are represented mainly by expanding the urban

fabric.

As Santorini constitutes an active volcano, it is

subjected to periodic tremors which strain on the

already congested areas. However, it was impossible

to model the spatial distribution of this specific

factor.

Considering the need for local authorities to

manage the landslide risk, a model was built,

providing an easy-to-use tool for landslide

susceptibility mapping.

In case a single factor or a number of factors

change through time, the model gives the

opportunity to decision makers to reevaluate the

landslide risk according to future conditions. In

addition, the model can be easily edited in order to

add any other determinant or triggering factor. These

options make the model extremely adaptable.

Furthermore, it makes it very easy for other

researchers to realize the preliminary landslide risk

assessment in other areas, simply by changing either

the controlling factors, or their respective weights.

AKNOWLEDGEMENTS

We would like to thank Viktor Vereb, PhD Student

from the Department of Physical Geography, Eötvös

Loránd University, who digitized and shared with us

the geological map of Druitt et al. (1999) in vector

format.

REFERENCES

Akgun, A & Bulut, F 2007, ‘GIS-based landslide

susceptibility for Arsin-Yomra (Trabzon, North

Turkey) region’, Environ. Geol., vol. 51, no. 8, pp.

1377-1387.

Antoniou, A & Lekkas, E 2010, ‘Rockfall susceptibility

map for Athinios port, Santorini Island, Greece’,

Geomorphology, vol. 118, pp. 152–166.

Ayalew, L, Yamagishi, H, Marui, H & Kanno, T 2005,

Landslides in Sado Island of Japan: Part II. GIS-based

susceptibility mapping with comparisons of results

from two methods and verifications. Engineering

Geology, vol. 81, pp. 432-445.

Barredo, JI Benavides, A Hervas, J Van Westen, CJ 2000,

‘Comparing heuristic landslide hazard assessment

techniques using GIS in the Trijana basin, Gran

Canaria Island, Spain’, Int. J. Appl. Earth. Obser.

Geoinf. vol. 2, pp. 9-23.

Brabb, E 1993, Proposal for worldwide landslide hazard

maps, Proceedings of 7th International Conference

and field workshop on landslide in Czech and Slovak

Republics, pp. 15–27.

Landslide Risk Assessment of the Santorini Volcanic Group

139

Figure 5: Landslide susceptibility map of the Santorini Volcanic Group.

Druitt, TH 2014, ‘New insights into the initiation and

venting of the Bronze-Age eruption of Santorini

(Greece), from component analysis’, Bull. Volcanol.,

76:794. doi:10.1007/s00445-014-0794-x.

Druitt, TH & Francaviglia, V 1992, ‘Caldera formation on

Santorini and the physiography of the islands in the

late Bronze Age’, Bull. Volcanol., vol. 54, pp. 484–

493.

Druitt, TH, Edwards, L, Mellors, R, Pyle, DM, Sparks,

RSJ, Lanphere, M, Davies, M, & Barriero, B 1999,

‘Santorini volcano’, Geol. Soc. Lond. Mem. vol. 19, p.

165.

Fabbro, G, Druitt, TH, Scaillet, S, 2013, ‘Evolution of the

crustal magma plumbing system during the build-up to

the 22-ka caldera-forming eruption of Santorini

(Greece)’, Bull. Volcanol., 75: 767.

doi:10.1007/s00445-013-0767-5.

Friedrich, W, 2000, Fire in the Sea, the Santorini

Volcano: Natural History and the Legend of Atlantis,

Cambridge University Press, Cambridge.

GISTAM 2017 - 3rd International Conference on Geographical Information Systems Theory, Applications and Management

140

Guzzetti, F, Carrara, A, Cardinali, M & Reichenbach, P

1999, ‘Landslide hazard evaluation: a review of

current techniques and their application in a

multistudy, Central Italy’, Geophys J Roy Astron Soc

vol. 31, pp. 181–216.

Guzzetti, F, Reichenbach, P, Cardinali, M, Galli M &

Ardizzone, F 2005, ‘Probablistic landslide hazard

assessment at the basin scale’, Geophys J Roy Astron

Soc vol. 72 pp. 272–299.

Karamesouti, M, 2011, ‘Environmentally sensitive areas

to desertification in Greece’, MSc thesis, Agricultural

University of Athens, Postgraduate Course

“Exploitation of Natural Resources and Agricultural

Engineering”, <https://goo.gl/FIQ5s6> (accessed

28/11/2016, in greek).

Koukis, G, Sabatakakis, N, Nikolaou, N & Loupasakis C,

2005. ‘Landslide hazard zonation in Greece’, In:

Proceedings of the first general assembly of the

international consortium on landslides, pp. 291–296.

Kouli, M, Loupasakis, C, Soupios, P, Rozos, D and

Vallianatos, F 2013, ‘Comparing multi-criteria

methods for landslide susceptibility mapping in

Chania Prefecture, Crete Island, Greece’, Natural

Hazards Earth Syst. Sci. Discuss., vol. 1, pp. 73–109.

Ladas, I, Fountoulis, I. & Mariolakos I 2007, ‘Using GIS

and Multicriteria Decision Analysis in landslide

susceptibility mapping - case study in Messinia

prefecture area (SW Peloponnesus, Greece)’ Proc.

11th International Conference of the Geological

Society of Greece, Athens, 24-24 May 2007, Bull.

Geol. Soc. Greece, vol. XXXX, no. 4, pp. 1973-1985.

Lekkas, E 2009, ‘Landslide hazard and risk in geologically

active areas. The case of the caldera of the Santorini

(Thera) volcano island complex (Greece)’, 7th Asian

Regional Conference for IAEG, International

Association for Engineering Geology (IAEG),

Chengdu pp. 417-423 .

Lekkas E, Papanikolaou, D, Nοmikou, P and

Papanikolaou, M 2010, Urban and land planning of

settlements in geodynamically hazardous regions of

particular cultural, environmental and natural

significance. The case of Oia-Santorini, Greece,

Geologically Active – Williams et al. (eds), Taylor and

Francis Group, London, ISBN 978-0-415-60034-7.

Lekkas E, Alexoudi V and Lialiaris I 2013, ‘Reduction of

rockfall risk of the Teleferik area of Santorini -

Greece’, Proc. 13th International Conference of the

Geological Society of Greece, Chania, Sept. 2013,

Bull. Geol. Soc. Greece, vol. XLVII, no. 3, pp. 1731-

1738.

Marinoni, O 2014, Analytic Hierarchy Process Tool

extAhp 20, <https://goo.gl/IlQ25Q> (accessed

10/1/2017).

Mpliona, M 2008, Development of landslide database in

Greece. MSc thesis, University of Patras, Postgraduate

Course “Geosciences and Environment”, in greek.

Nomikou, P, Papanikolaou, D, Alexandri, M, Sakellariou

D. & Rousakis G, 2013, ‘Submarine volcanoes along

the Aegean Volcanic Arc’, Tectonophysics vol. 507-

508, pp. 123-146.

Nomikou, P, MM, Parks, D, Papanikolaou, DM, Pyle, TA,

Mather, S, Carey, AB, Watts, M, Paulatto, ML,

Kalnins, I, Livanos, K, Bejelou, E, Simou, & I, Perros,

2014, ‘The emergence and growth of a submarine

volcano: The Kameni islands, Santorini (Greece)’,

GeoResJ vol. 1-2, pp. 8–18.

Nomikou, P., Druitt, T.H., Hübscher, C., Mather, T.A.,

Paulatto, M., Kalnins, L.M., Kelfoun, K.,

Papanikolaou, D., Bejelou, K., Lampridou, D., Pyle,

D.M., Carey, S., Watts, A.B., Weiß, B. and Parks,

M.M. 2016, ‘Post-eruptive flooding of Santorini

caldera and implications for tsunami generation’,

Nature Communications, vol. 7, p. 13332. doi:

10.1038/ncomms13332.

Pardeshi, S, Autade, S, & Pardeshi, S, 2013, ‘Landslide

hazard assessment: recent trends and techniques’,

Springer Plus, 2:523.

Pyle, DM & Elliott, JR, 2006, ‘Quantitative morphology,

recent evolution, and future activity of the Kameni

Islands volcano, Santorini, Greece’, Geosphere vol. 2,

pp. 253–268.

Saaty, T.L, 1980, The Analytic Hiearchy Process,

McGraw-Hill, New York.

Saaty T, 2008, ‘Decision making with the analytical

hierarchy process’, Int J Services Sci vol. 1, no.1 pp.

83–98.

Stergianos, EP 2016, ‘Online application using Geomatics

(webGIS) for the management of natural disasters and

delineation of hazard zones on the island of Santorini’,

MSc thesis, National and Kapodistrian University of

Athens, Postgraduate Course “Prevention and

Management of Natural Disasters”, in greek.

Yassoglou, N, 2004, Soil Association Map of Greece,

1:500.000, Greek National Committee for Combating

Desertification, Agricultural University of Athens,

Athens.

Special Secretariat for Water (SSW), 2015, River basin

management Plan RBMP of the Islands Aegean river

basin district RBD (GR14) <https://goo.gl/Zgf0g6>

(access 28/11/2016, in greek).

Varnes D, IAEG, 1984. Landslide hazard zonation: a

review of principles and practice, United Nations

Scientific and Cultural Organization, Paris, 1–6.

WP/WLI — International Geotechnical societies'

UNESCO Working Party on World Landslide

Inventory, 1994. A suggested method for reporting a

landslide, International Association Engineering

Geology Bulletin, vol. 50, pp. 71–74.

Landslide Risk Assessment of the Santorini Volcanic Group

141