ChapterPDF Available

The recovery strategy of second homeowners and tourists after a disaster: insights from the 2016 central Italy earthquake

Authors:
Geographies of the Anthropocene
Open Access and Peer-Reviewed series
Editor-In-Chief: Francesco De Pascale (CNR Research Institute for Geo-
Hydrological Protection, Italy).
Co-Editors: Marcello Bernardo (Department of Culture, Education and
Society, University of Calabria, Italy); Charles Travis (School of Histories
and Humanities, Trinity College Dublin; University of Texas, Arlington).
Editorial Board: Mohamed Abioui (Ibn Zohr University, Morocco), Andrea
Cerase (Sapienza University of Rome, Italy), Valeria Dattilo (University of
Calabria, Italy), Chair, Dante Di Matteo (Polytechnic University of Milan,
Italy); Jonathan Gómez Cantero (University of Alicante, Spain; Young
Scientists Club, IAPG), Nguvulu Chris Kalenge (University School for
Advanced Studies IUSS Pavia, Italy), Battista Liserre (Aix-
Marseille University, Campus ESSCA, France), Giovanni Messina
(University of Palermo, Italy), Gaetano Sabato (University of Catania, Italy),
Carmine Vacca (University of Calabria, Italy).
International Scientific Board: Marie-Theres Albert (UNESCO Chair in
Heritage Studies, University of Cottbus-Senftenberg, Germany), David
Alexander (University College London, England), Loredana Antronico (CNR
Research Institute for Geo-Hydrological Protection, Italy), Lina Maria
Calandra (University of L’Aquila, Italy); Salvatore Cannizzaro (University of
Catania, Italy), Fabio Carnelli ((Polytechnic University of Milan, Italy); Carlo
Colloca (University of Catania, Italy), Gian Luigi Corinto (University of
Macerata, Italy); Roberto Coscarelli (CNR Research Institute for Geo-
Hydrological Protection, Italy), Sebastiano D’Amico (University of Malta,
Malta), Armida de La Garza (University College Cork, Ireland), Elena
Dell’Agnese (University of Milano-Bicocca, Italy; Vice President of IGU),
Piero Farabollini (University of Camerino, Italy), Giuseppe Forino
(University of Newcastle, Australia), Virginia García Acosta (Centro de
Investigaciones y Estudios Superiores en Antropología Social, CIESAS,
México); Cristiano Giorda (University of Turin, Italy), Giovanni Gugg
(University of Naples “Federico II”, Italy, University of Nice Sophia
Antipolis, France), Luca Jourdan (University of Bologna, Italy), Francesca
Romana Lugeri (ISPRA, University of Camerino, Italy), Fausto Marincioni
(Marche Polytechnic University, Italy), Cary J. Mock (University of South
Carolina, U.S.A.; Member of IGU Commission on Hazard and Risk),
Francesco Muto (University of Calabria, Italy), Gilberto Pambianchi
(University of Camerino, Italy; President of the Italian Association of
Physical Geography and Geomorphology), Silvia Peppoloni (Istituto
Nazionale di Geofisica e Vulcanologia, Italy; Secretary General of IAPG;
Councillor of IUGS), Isabel Maria Cogumbreiro Estrela Rego (University of
the Azores, Portugal), Andrea Riggio (University of Cassino and Southern
Lazio, Italy; President of the Association of Italian Geographers), Bruno
Vecchio (University of Florence, Italy), Masumi Zaiki (Seikei University,
Japan; Secretary of IGU Commission on Hazard and Risk).
Editorial Assistant, Graphic Project and Layout Design: Franco A.
Bilotta;
Website: www.ilsileno.it/geographiesoftheanthropocene;
The book series “Geographies of the Anthropocene” edited by Association
for Scientific Promotion “Il Sileno” (Il Sileno Edizioni) will discuss the new
processes of the Anthropocene epoch through the various worldviews of
geoscientists and humanists, intersecting disciplines of Geosciences,
Geography, Geoethics, Philosophy, Socio-Anthropology, Sociology of
Environment and Territory, Psychology, Economics, Environmental
Humanities and cognate disciplines.
Geoethics focuses on how scientists (natural and social), arts and humanities
scholars working in tandem can become more aware of their ethical
responsibilities to guide society on matters related to public safety in the face
of natural hazards, sustainable use of resources, climate change and protection
of the environment. Furthermore, the integrated and multiple perspectives of
the Environmental Humanities, can help to more fully understand the cultures
of, and the cultures which frame the Anthropocene. Indeed, the focus of
Geoethics and Environmental Humanities research, that is, the analysis of the
way humans think and act for the purpose of advising and suggesting
appropriate behaviors where human activities interact with the geosphere, is
dialectically linked to the complex concept of Anthropocene.
The book series “Geographies of the Anthropocene” publishes online
volumes, both collective volumes and monographs, which are set in the
perspective of providing reflections, work materials and experimentation in
the fields of research and education about the new geographies of the
Anthropocene.
“Geographies of the Anthropocene” encourages proposals that address one or
more themes, including case studies, but welcome all volumes related to the
interdisciplinary context of the Anthropocene.
Published volumes are subject to a review process (double blind peer
review) to ensure their scientific rigor.
The volume proposals can be presented in English, Italian, French or Spanish.
The choice of digital Open Access format is coherent with the flexible
structure of the series, in order to facilitate the direct accessibility and
usability by both authors and readers.
EARTHQUAKE RISK PERCEPTION,
COMMUNICATION AND
MITIGATION STRATEGIES ACROSS
EUROPE
Piero Farabollini
Francesca Romana Lugeri
Silvia Mugnano
Editors
“Earthquake risk perception, communication and mitigation strategies
across Europe”,
Piero Farabollini, Francesca Romana Lugeri, Silvia Mugnano (Eds.)
is a volume of the Open Access and peer-reviewed series
“Geographies of the Anthropocene”
(Il Sileno Edizioni), ISSN 2611-3171.
www.ilsileno.it/geographiesoftheanthropocene
Cover: Norcia, Piazza San Benedetto. On the left, the civic tower of the Town Hall; on the right, the
safety intervention of the facade of the Basilica of San Benedetto, heavily damaged as a result of the
seismic events that affected central Italy starting from 24 August 2016.
Copyright © 2019 by Il Sileno Edizioni
Scientific and Cultural Association “Il Sileno”,
Via Pietro Bucci, Università della Calabria, 87036 - Rende (CS), Italy.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs
3.0 Italy License.
The work, including all its parts, is protected by copyright law. The user at the time of
downloading the work accepts all the conditions of the license to use the work, provided
and communicated on the website
http://creativecommons.org/licenses/by-nc-nd/3.0/it/legalcode
ISBN 978-88-943275-6-4
Vol. 2, No. 2, December 2019
6
CONTENTS
Preface 8
Introduction 12
Section I
Mitigation Strategies of Seismic Risk
1. Urban Seismic Risk Reduction and Mitigation Strategies in Turkey
Ahmet Anıl Dindar, Cüneyt Tüzün and Aybige Akinci 19
2. A Collection of Statistical Methods for Analysis of the Disaster
Damages and the Seismic Regime
Vladilen Pisarenko, Mikhail V. Rodkin 43
3. Turkey’s Earthquake History and Institution Based Earthquake
Reduction Policies and Strategies
Alper Uzun, Burak Oğlakci 64
4. Risk Mitigation through Local Building Knowledge: Turkish Van
Region Case Study
Chiara Braucher, Mattia Giandomenici 84
Section II
Communication and Prevention Strategies of Seismic Risk
5. Communication-Based Prevention Strategies: A Draft Model
Proposal
Andrea Volterrani 105
6. Geoscientists’ Voice in the Media: Framing Earth Science in the
Aftermath of Emilia 2012 and Amatrice 2016 Seismic Crises
Andrea Cerase 123
7. The 2016 Earthquake in Central Italy. The Alphabet of Reconstruction
Piero Farabollini 145
7
8. Food Management in Disasters: the Case Study of the Earthquakes
of 24 August 2016 in Central Italy
Fausto Marincioni, Eleonora Gioia, Mirco Zoppi,
Elena Vittadini 172
Section III
Resilience and Post-Disaster Recovery
9. An Historical Flight and Some Open Questions towards a Pluralistic
but Holistic View of Resilience
Maurizio Indirli 194
10. Earthquakes and Society: the 2016 Central Italy Reverse Seismic
Sequence
Piero Farabollini, Serafino Angelini, Massimiliano Fazzini,
Francesca Romana Lugeri, Gianni Scalella,
GeomorphoLab 249
11. Second Home Holidays Makers Recovery After a Disaster: Insights
from the 2016 Central Italy Earthquake
Silvia Mugnano, Fabio Carnelli, Sara Zizzari 267
12. Assessing Resilience of Mountain Communities Hit By The Central
Italy Earthquakes of 2016
Teresa Carone, Giulio Burattini, Fausto Marincioni 285
The Authors 302
8
Preface
Sebastiano D’Amico1
This volume aims at collecting some contributions presented in the S41
“Earthquake risk perception, communication and mitigation strategies”
session of the 36th General Assembly of the European Seismological
Commission, held in Valletta, Malta, of which I had the honor of being
Chairperson and Organizer. The ESC mission is to promote the science of
Seismology within the scientific community of the European and
Mediterranean countries (encompassing the area from the Mid-Atlantic Ridge
to the Ural Mountains and from the Arctic Ocean to northern Africa), by
promoting research studies, to extend and enhance scientific co-operation and
to train young scientists.
Session S41 was very welcome during the 36th ESC General Assembly
since it covered issues related to the perception and communication of seismic
risk which are certainly worthy of consideration. In fact, every disaster caused
by physical and natural phenomena, such as the earthquake, or deriving from
human causes, always represents a cross-section of the life of a community,
of society and of the affected place, bringing out the latent vulnerabilities that
cause a catastrophe, the resources available and the qualities of the pre-
existing (sometime complex) relationships between the population and the
authorities. Indeed, it is important to consider also the social aspects also
because the latter can be strongly related to the several legal aspects with
particular regards to the reconstruction phase after a major earthquake as well
as to analyze the communication processes during the emergency, to
understand the evolution of local governance, to account of the
transformations that take place in the daily life of the involved actors. Thus,
social approaches can play a key role in disaster studies. In this context, the
session “Earthquake risk perception, communication and mitigation
strategies”, proposed and coordinated by Francesco De Pascale, Francesca
Romana Lugeri, Elena Dell’Agnese, Fausto Marincioni, Francesco Muto and
Piero Farabollini, had an excellent response with twenty one contributions
presented by scholars from various countries: Spain, Romania, Russia, United
1 Department of Geosciences, University of Malta, Msida Campus, Malta; Vicepresident of
European Seismological Commission 2016-2018; Chairperson 36th General Assembly of
ESC, Valletta, Malta, 2-7 September 2018, e-mail: sebastiano.damico@um.edu.mt.
9
States, Hungary, France, Taiwan, Turkey, Portugal, Armenia, Greece,
Bulgaria, Cyprus, Italy. It was an important occasion that offered original and
interesting studies and reflections in the field of earthquake perception,
resilience and risk communication, the use of new technologies for seismic
risk investigations, with the presentation of several case studies with
interdisciplinary and integrated approaches. Resilience, as pointed out by
Indirli (2019), in this volume, is another important topic which needs to be
addressed. Indirli presents an interesting study on the etymology of the word
resilience, which in the last years has assumed ever-increasing importance,
both in the academic context and in that relating to policy, to the point of
being in competition with the concept of sustainability. From a social
perspective, resilience can be defined as an interactive and multidirectional
social process, consisting of a set of pre-existing response behavior learned
during the event itself (Lucini, 2014).
A key role also was also given by the contributions that deal with the
recent earthquakes of Central Italy in 2016 tackling the disaster from different
perspectives: from the analysis of food management in the disaster
(Marincioni et al., 2019) to the study of social resilience in mountain
communities (Carone et al., 2019); from the topic of the second home tourism
recovery (Mugnano et al., 2019) to the role of communication spread by the
media in the context of the earthquakes of Emilia and Amatrice (Cerase,
2019). In addition, there are two contributions in which the Extraordinary
Commissioner for post-earthquake reconstruction of Central Italy
(Farabollini, 2019; Farabollini et al., 2019) is involved presenting “a
provisional alphabet of reconstruction” and an analysis of the seismic
sequence of 2016. Three contributions, on the other hand, present different
seismic risk mitigation strategies in Turkey (Braucher and Giandomenici
2019; Dindar et al., 2019; Uzun and Oğlakci, 2019). Rodkin and Pisarenko
(2019) present an examination of methods of statistical analysis of seismic
regime and related damages. Finally, Volterrani (2019) presents and discusses
a draft model for the prevention strategies in relation to the risk of disasters
through an analysis of the Italian campaign “I don’t take risks”.
The result is a book that represents an essential point of reference for those
interested in disaster studies regarding earthquake risk in Europe.
10
References
Braucher C. Giandomenici M., 2019, Risk Mitigation Through Local
Building Knowledge: Turkish Van Region Case Study. In: Farabollini P.,
Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception, communication
and mitigation strategies across Europe, Il Sileno Edizioni, Geographies of
the Anthropocene book series, Rende, 2, 2, 84-103.
Carone T., Burattini G., Marincioni F., Assessing Resilience of Mountain
Communities Hit by the Central Italy Earthquakes of 2016. In: Farabollini
P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception,
communication and mitigation strategies across Europe, Il Sileno Edizioni,
Geographies of the Anthropocene book series, Rende, 2, 2, 285-301.
Cerase A., 2019, “Geoscientists’ voice in the media: framing Earth science
in the aftermath of Emilia 2012 and Amatrice 2016 seismic crises. In:
Farabollini P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception,
communication and mitigation strategies across Europe, Il Sileno Edizioni,
Geographies of the Anthropocene book series, Rende, 2, 2, 123-144.
Dindar A.A, Tüzün C., Akinci A., 2019, Urban Seismic Risk Reduction
and Mitigation Strategies in Turkey. In: Farabollini P., Lugeri F.R.,
Mugnano S. (Eds.), Earthquake risk perception, communication and
mitigation strategies across Europe, Il Sileno Edizioni, Geographies of the
Anthropocene book series, Rende, 2, 2, 19-42.
Farabollini P., 2019, The 2016 Earthquake in Central Italy. The Alphabet
of Reconstruction. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 145-171.
Farabollini P., Angelini S., Fazzini M., Lugeri F.R., Scalella G.,
GeomorphoLab, 2019, Earthquakes and Society: the 2016 central Italy
reverse seismic sequence. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 249-266.
Indirli M., 2019, An historical flight and some open questions towards a
pluralistic but holistic view of resilience. In: Farabollini P., Lugeri F.R.,
Mugnano S. (Eds.), Earthquake risk perception, communication and
mitigation strategies across Europe, Il Sileno Edizioni, Geographies of the
Anthropocene book series, Rende, 2, 2, 194-248.
11
Lucini B., Disaster Resilience from a Sociological Perspective. Exploring
Three Italian Earthquakes as Models for Disaster Resilience Planning,
Humanitarian Solutions in the 21st Century, Springer, Cham, 2014.
Marincioni F., Gioia E., Zoppi M., Vittadini E., 2019, Food management
in disasters: the case study of the earthquakes of 24 august 2016 in Central
Italy. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk
perception, communication and mitigation strategies across Europe, Il Sileno
Edizioni, Geographies of the Anthropocene book series, Rende, 2, 2, 172-
192.
Mugnano S., Carnelli F., Zizzari S., 2019, The recovery strategy of
second homeowners and tourists after a disaster: insights from the 2016
central Italy earthquakes. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 267-284.
Pisarenko V., Rodkin M.V., 2019, A Collection of Statistical Methods for
Analysis of the Disaster Damages and the Seismic Regime. In: Farabollini
P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception,
communication and mitigation strategies across Europe, Il Sileno Edizioni,
Geographies of the Anthropocene book series, Rende, 2, 2, 43-63.
Uzun A., Oğlakci B., 2019, “Turkey’s Earthquake History and Institution
Based Earthquake Reduction Policies and Strategies. In: Farabollini P.,
Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception, communication
and mitigation strategies across Europe, Il Sileno Edizioni, Geographies of
the Anthropocene book series, Rende, 2, 2, 64-83.
Volterrani A., 2019, Communication-Based Prevention Strategies: A
Draft Model Proposal. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 105-122.
12
Introduction
Piero Farabollini1, Francesca Romana Lugeri2, Silvia Mugnano3
Risk and disasters are social constructs deriving from an unsustainable
human-environment interaction. Earthquake hazard doesn’t create damages
and destruction; it is our vulnerability and exposure to such processes that
creates the conditions of risk. There is nothing natural about an earthquake
disaster, yet the common perception is that humans are victims of nature’s
extreme events. Moreover, the ability of a society to respond to earthquakes
does not depend primarily on the emergency conditions created by the impact,
but rather on the pre-disaster settings and circumstances.
From the agricultural revolution onwards, humans have tried to free
themselves from the control of nature by modeling the territory for their
benefit. This, on the one hand, has enabled the social development we enjoy
today, yet, on the other, the interaction with natural processes we do not fully
understand has created problems of exposure and vulnerability. The
consequences went beyond the creation of risk conditions and caused
profound changes in environmental cycles contributing to the current
geographies of the Anthropocene.
Recent earthquakes, including those in Italy, have unequivocally shown
the dominant role of societal vulnerability in creating those disasters. The
Mediterranean region, unceasingly affected by strong earthquakes and almost
all type of known natural hazards, is very representative of these complex and
multi-scale dynamics.
From an examination of the dramatic events that have recently occurred in
the central region of Italy, there emerges the need to provide the general
public with correct and clear information on the complex scenario
characterising this as well as another- country. Experience teaches us that
tackling the subject of the prevention of risk and protection from danger (the
avoidance of exposure) is very difficult. What is needed is a communicative
strategy that informs the public of the characteristics of a territory (understood
1 Extraordinary Government Commissioner for the reconstruction in the earthquakes areas of
the 2016 and 2017; Scuola di Scienze e Tecnologie, Sezione di Geologia, Università degli
Studi di Camerino, Via Gentile da Varano, 1, 62032 Camerino (MC); e-mail:
piero.farabollini@unicam.it.
2 Servizio Geologico d’Italia - ISPRA, Via Vitaliano Brancati, 48, 00144 Rome, Italy, e-mail:
francesca.lugeri@unicam.it.
3 University of Milano-Bicocca, Milan, Italy, e-mail: silvia.mugnano@unimib.it.
13
as a natural and cultural environment) and the relative operative dynamics,
just as one should understand the anatomy and physiology of ones own body
in order to manage and protect it in the best possible way.
Indeed, a disaster is above all a social event (Alexander, 1991; Ligi, 2009;
Pelanda, 1981), in which people are actively involved in the process leading
to the occurrence of the catastrophe. It is not by chance that the social sciences
engaged in the study of disasters in Europe have experienced an important
consolidation in recent years; in Italy, especially since the earthquake of
L'Aquila onwards, the national scientific production has substantially aligned
to the international growth trend. Such vivacity, as Davide Olori (2017)
states, does not correspond to a theoretical reconstruction of the proposals,
which on the contrary have widened the distances between the different
positions, pursuing - mostly - an applied approach. This volume, instead, is
intended to be the first attempt of a proposal that aims to bring together
different approaches and viewpoints of scholars from different disciplines on
the subjects of reduction, mitigation and communication of earthquake risk:
physical and social scientists, physicists, engineers and humanists who
participated in the S41 session of the 36th Assembly of the European
Seismological Commission which took place in Valletta, Malta from 2 to 7
September 2018, coordinated by Elena DellAgnese, Francesco De Pascale,
Piero Farabollini, Francesca Romana Lugeri, Fausto Marincioni, and
Francesco Muto. This session encouraged abstracts discussing the multiple
dimensions of earthquake risk reduction, including, but not limiting to, the
following research lines: risk communication and social perception;
prevention and population preparedness; community-based approach;
adaptive capacity; representation of earthquakes in popular culture; new
technologies for investigations of hazards and risk; vulnerability reduction;
disaster governance. As a result, this volume, has collected several
contributions presented during this session to which other interesting
proposals of scholars presented after the publication of the Call for Book
chapters of the series have been added. Hence, this book is an output of a
rigorous review of those proposals and contributions. The volume is divided
into three sections:
1) Mitigation Strategies of Seismic Risk Communication;
2) Communication and Prevention Strategies of Seismic Risk.
3) Resilience and Post-Disaster Recovery.
In the first section, Mitigation Strategies of Seismic Risk Communication”,
Cüneyt Tüzün, Ahmet Anıl Dindar, Aybige Akıncı (2019) explain one of the
most comprehensive and challenging disaster mitigation strategy being
applied in Turkey based on the real experience since the 1999 earthquakes.
14
Mikhail Rodkin and Vladilen Pisarenko (2019) deal with a review of a series
of previous publications by authors about the methods of statistical analysis
of seismic regime and related damages. The work of Alper Uzun and Burak
Oglakci (2019) covers the prevention and risk management studies to be done
before an earthquake occurs, focusing on awareness level and risk
governance. Chiara Braucher and Mattia Giandomenici (2019) would
propose the proactive and participative approach to the Environment
Construction at large, including the “direct intervention from settled
communities - still persistent but in serious decrease all around the world - as
an important strategy for risk mitigation, an alternative to the profit-based
narrations of political decisions”. In the second section “Communication and
Prevention Strategies of Seismic Risk, Volterrani’s chapter (2019) presents
and discusses a draft model for the prevention of communication in relation
to risk of disasters and other types of crisis, starting from the experience of
the Italian campaign “I do not risk”, and, finally, to risk of radicalization of
second young migrant generation. Andrea Cerase’s work (2019) considered
the media coverage of scientific issues during the Emilia 2012 and Amatrice
2016 seismic crisis by the four most circulating Italian national newspapers
within the 31 days following the first earthquake shock, through a
comparative analysis. The contribution of Piero Farabollini (2019) aims to
illustrate, through a sort of alphabet the activity of the commissioner, the
legislative and financial system and the route - with the relative rules to reach
the objectives - necessary to give society the due guarantees. The study of
Fausto Marincioni, Eleonora Gioia, Mirco Zoppi and Elena Vittadini (2019)
investigates, through a questionnaire, food management in the case of the
earthquakes of 24 August 2016 in Central Italy, assessing survivors’ ability
to access food (food security) and the field kitchens practices to ensure
hygiene and avoid food-borne disease outbreak (food safety).
In the third section Resilience and Post-Disaster Recovery”, Maurizio
Indirli’s work (2019) presents an excursus through the ages and a brief (not
exhaustive, of course) state-of-the-art regarding “resilience”, pointing out
some open questions of the current debate among researchers of different
disciplines, working in the fields of hazard mitigation, sustainability, risk
assessment, heritage preservation, and so on.
Piero Farabollini, Francesca Romana Lugeri and other authors (2019) deal
with the case study of the 2016 central Italy, describing the reverse seismic
sequence and the geological effects.
The work of Silvia Mugnano, Fabio Carnelli and Sara Zizzari (2019) aims
to discuss what needs to be tackled by response and recovery disaster
management policies when second homes are involved, by considering also
15
the expectations and intentions of the affected owners with regards to tourists
needs included in the redevelopment plans.
Finally, the chapter of Teresa Carone, Giulio Burattini and Fausto
Marincioni (2019) aims to clarify the influence of territorial bonds on social
resilience of small mountainous communities in the aftermath of the August
24, 2016 central Italy earthquake.
References
Alexander, D., 1991, Natural Disasters: A Framework for Research and
Teaching, Disasters, 15(3): 209-226. DOI: 10.1111/j.1467-
7717.1991.tb00455.x.
Braucher C. Giandomenici M., 2019, “Risk Mitigation Through Local
Building Knowledge: Turkish Van Region Case Study”. In: Farabollini P.,
Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception, communication
and mitigation strategies across Europe, Il Sileno Edizioni, Geographies of
the Anthropocene book series, Rende, 2, 2, 84-103.
Carone T., Burattini G., Marincioni F., “Assessing Resilience of Mountain
Communities Hit by the Central Italy Earthquakes of 2016”. In: Farabollini
P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception,
communication and mitigation strategies across Europe, Il Sileno Edizioni,
Geographies of the Anthropocene book series, Rende, 2, 2, 285-301.
Cerase A., 2019, “Geoscientists’ voice in the media: framing Earth science
in the aftermath of Emilia 2012 and Amatrice 2016 seismic crises”. In:
Farabollini P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception,
communication and mitigation strategies across Europe, Il Sileno Edizioni,
Geographies of the Anthropocene book series, Rende, 2, 2, 123-144.
De Pascale F., Farabollini P., Lugeri F.R. (Eds.), “Comunicare il rischio,
il rischio di comunicare”, Prisma Rivista di Economia Società Lavoro,
3, 2018.
Dindar A.A, Tüzün C., Akinci A., 2019, “Urban Seismic Risk Reduction
and Mitigation Strategies in Turkey”. In: Farabollini P., Lugeri F.R.,
Mugnano S. (Eds.), Earthquake risk perception, communication and
mitigation strategies across Europe, Il Sileno Edizioni, Geographies of the
Anthropocene book series, Rende, 2, 2, 19-42.
Farabollini P., 2019, “The 2016 Earthquake in Central Italy. The Alphabet
of Reconstruction”. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
16
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 145-171.
Farabollini P., Angelini S., Fazzini M., Lugeri F.R., Scalella G.,
GeomorphoLab, 2019, “Earthquakes and Society: the 2016 central Italy
reverse seismic sequence”. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 249-266.
Indirli M., 2019, “An historical flight and some open questions towards a
pluralistic but holistic view of resilience”. In: Farabollini P., Lugeri F.R.,
Mugnano S. (Eds.), Earthquake risk perception, communication and
mitigation strategies across Europe, Il Sileno Edizioni, Geographies of the
Anthropocene book series, Rende, 2, 2, 194-248.
Ligi, G., 2009, Antropologia dei disastri, Laterza, Bari.
Marincioni F., Gioia E., Zoppi M., Vittadini E., 2019, “Food management
in disasters: the case study of the earthquakes of 24 august 2016 in Central
Italy”. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk
perception, communication and mitigation strategies across Europe, Il Sileno
Edizioni, Geographies of the Anthropocene book series, Rende, 2, 2, 172-
192.
Mugnano S., Carnelli F., Zizzari S., 2019, “The recovery strategy of
second homeowners and tourists after a disaster: insights from the 2016
central Italy earthquakes”. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 267-284.
Olori, D., 2017, Per una questione subalterna" dei disastri, in: Mela, A.,
Mugnano, S., Olori, D. (Eds.), Territori vulnerabili. Verso una nuova
sociologia dei disastri italiana, Sociologia Urbana e Rurale, FrancoAngeli,
Milan, 81-86.
Pelanda, C., 1981, “Disastro e vulnerabilità socio sistemica”, Rassegna
italiana di Sociologia, 4, 507-532.
Pisarenko V., Rodkin M.V., 2019, “A Collection of Statistical Methods for
Analysis of the Disaster Damages and the Seismic Regime”. In: Farabollini
P., Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception,
communication and mitigation strategies across Europe, Il Sileno Edizioni,
Geographies of the Anthropocene book series, Rende, 2, 2, 43-63.
Uzun A., Oğlakci B., 2019, “Turkey’s Earthquake History and Institution
Based Earthquake Reduction Policies and Strategies”. In: Farabollini P.,
Lugeri F.R., Mugnano S. (Eds.), Earthquake risk perception, communication
17
and mitigation strategies across Europe, Il Sileno Edizioni, Geographies of
the Anthropocene book series, Rende, 2, 2, 64-83.
Volterrani A., 2019, “Communication-Based Prevention Strategies: A
Draft Model Proposal”. In: Farabollini P., Lugeri F.R., Mugnano S. (Eds.),
Earthquake risk perception, communication and mitigation strategies across
Europe, Il Sileno Edizioni, Geographies of the Anthropocene book series,
Rende, 2, 2, 105-122.
Section I
Mitigation Strategies of Seismic Risk
19
1. Urban Seismic Risk Reduction and Mitigation
Strategies in Turkey
Ahmet Anıl Dindar1, Cüneyt Tüzün2, Aybige Akinci3
Abstract
Since the early ages of humankind, safety and security has been a critical
issue against the forces of nature. However, history has always proven the
power of nature over humankind in certain regions on Earth for centuries.
Indeed, this is a never-ending war between Earth and its inhabitants, namely
us, human beings. Humankind’s organization (cities, roads, lifelines etc.) in
the nature has never been perfect within the view of environmental pollution
and excessive consumption of the resources. Particularly, the quality of civil
engineering design and practice is strongly affected from the social and
economic background of the country. The societies in rapid development
claim excessive demands in terms of housing and transportation. Such
demands may create vulnerable urban areas if the economic and social
conditions are not in balance or harmony. Thus, nature should not be blamed
as the scapegoat in the regions where disasters claim human and economic
losses. In fact, the reason for the losses is nothing else than humankind itself.
A rational question arises then about how to overcome human and economic
loss due to natural disasters. The idea of determining the most vulnerable
items in urban areas and reconstructing with the most reliable equivalents
may seem very challenging. Even though the macroeconomic implications
are very complex, reconstructing the items in densely populated areas is the
most effective mitigation action against disasters in the short term. Having
learnt lessons from the major earthquake disasters in the heart of the industry
and mostly dense urban areas, Turkish government has drawn a long strategic
road map in the risk perception and the disaster mitigation strategy for almost
all the community services and the infrastructure. The development of
awareness against disasters has become part of formal education at all ages.
The National Disaster Management system was reorganized from scratch and
1 Corresponding Author; Gebze Technical University, Department of Civil Engineering,
Turkey, e-mail: adindar@gtu.edu.tr.
2 Gebze Technical University Department of Earthquake Engineering, Turkey. e-mail:
cuneyttuzun@gtu.edu.tr.
3 Istituto Nazionale di Geofisica e Vulcanologia, Rome, Italy, e-mail: aybige.akinci@ingv.it.
20
the capabilities improved by providing additional financial and human
resources. All school and hospital buildings in İstanbul were assessed in terms
of seismic safety. Those found inadequate were demolished and then
reconstructed. In addition, a law on urban renewal of the seismic risk areas
was enacted in 2012 allowing the licensed engineering offices to assess the
seismic risk of residential buildings at the request of the house owners. If the
assessment report is approved by the local municipality, the building is set to
demolish within 60 days following the legal notice to the property owners.
Disagreeing owners have the right to get the assessment re-evaluated by the
independent peer reviewers. In the case of demolition, the house owners are
eligible to receive 12 months of rental support from the government. During
the time period 2012 to 2019, more than 120 000 buildings were assessed and
74% of them were demolished, the majority of the latter were in İstanbul area
where a major earthquake is expected within the following decades. This
chapter is intended to explain one of the most comprehensive and challenging
disaster mitigation strategies being applied in Turkey based on experience
since the 1999 earthquakes.
Keywords: buildings, disasters, mitigation strategies, Turkey, urban
seismic risk reduction,
1. Introduction
Because of the real earthquake threat in Turkey, due to the country’s
geological and tectonic structure characteristics, the need for seismic hazard
studies has become progressively more important for engineering
applications, mitigation and reduction of earthquake risk particularly after the
two recent earthquakes; İzmit-August 17, 1999, M7.4 and Düzce -November
12, 1999, M7.2 (Erdik et al., 1999). According to statistical results, natural
disasters in Turkey from 1900 to 2011 are dominated by earthquakes, and
earthquakes are a synonym with the concept of disaster in Turkey (Sonmez
Saner, T. 2015; Ergunay, 2007). The United Nations Development Program
(2004) and the Global Assessment Report on Disaster Risk Reduction (2009)
reported that Turkey ranks high among countries according to mortality risk
and significant losses of property due to earthquakes. For example, 1939
Erzincan, M7.9 and 1999 İzmit M7.4 earthquakes caused almost 32,000 and
17,000 fatalities and left more than half a million people homeless. Economic
losses caused by larger earthquakes have often exceeded $5 billion (US$) and
21
have reached $23 billion and $10 billion for the 1939 Erzincan and 1999 İzmit
earthquakes, respectively.
These major earthquakes have also revealed that buildings are quite
vulnerable in the country. A majority of the population is living in
earthquake-prone areas where there are also the industrial facilities producing
75% of the nation’s economic income (e.g. Marmara and western Anatolian
region). The rapid migration from rural to urbanized areas since the 1950s in
Turkey caused severe circumstances in terms of vulnerability in every aspect
of life. Only 25% of the entire population was living in urban areas in 1950,
but this ratio reached 75% in 2017. The new residents of the metropolitan
cities demanded to a house in very large numbers. Due to the steadily
increasing population, with improper land-use planning, inappropriate
construction techniques and inadequate infrastructure systems, associated
with existing high hazard level, many major cities in Turkey (e.g. İstanbul,
Bursa) have become some of the most risky cities in Europe and the
Mediterranean region (Ansal et al., 2010).
The constant and persistent risk of being hit by a devastating earthquake
has become a crucial social and financial issue for the Turkish government.
The earthquakes of 1999 generated a strong national determination in Turkey
to devise new and effective methods of tackling disasters. A number of risk
assessment studies have been carried out in Turkey both at national and local
levels since 1999. These studies that related to settlement level risks are the
Earthquake Master Plan of İstanbul (EMPI) imposed by the Metropolitan
Municipality of İstanbul and carried out by four universities in 2003 (ITU,
METU, BU and Yıldız Tech. Un.). EMPI developed a comprehensive
framework for the determination of urban risks and methods of reducing
them. One of the important national projects was the İstanbul Seismic Risk
Mitigation and Emergency Preparedness (ISMEP) Project initiated by the
Turkish Government, financed by a World Bank loan and carried out by the
İstanbul Special Provincial Administration (ISMEP, 2010; www.ipkb.gov.tr).
Its objective was to transform densely-populated İstanbul (hosting over 14.6
million people is approximately one-fifth of Turkey’s population) into a city
resilient to a major earthquake by strengthening the emergency management
capacity, enhancing emergency preparedness, activating the seismic risk
mitigation actions for priority public buildings and the enforcement of
Building Codes. However, all those efforts have not been specified in a
particular policy or action plan. In 2009 the Disaster and Emergency
Management Presidency of Turkey (AFAD) was also established effective
emergency management and civil protection issues on a nationwide scale.
22
As a part of the Declaration of the National Earthquake Strategy Plan for
Turkey (NESAP-2023) between 2012 and 2023, the government decided to
implement a very strict policy for action through a law called
“Transformation of Areas under the Disaster Risks (No: 6306)” legislated in
May 2012. This policy calls for the demolition of risky and illegal buildings
and the renewal of those based on some rules and procedures. The cost of
urban transformation is roughly estimated to be at $500 billion and the
timeframe for completion is, ambitiously, 20 years (Güneş, 2015). By now it
has been in use almost in every town in Turkey, however, numerous
discussions and allegations have made by academic and non-governmental
organizations due to the application procedures This paper deals with the
rationale of Law No: 6306, its scope, its procedures, and explains the current
situation in its application. Having been enforced for 7 years, there have been
many lessons learned from the application of the law and its social and
economic effect on society.
2. Seismotectonic Setting and Seismic Activity in Turkey
Turkey is located on the Alpine-Himalayan Seismic Belt which is one of
the most seismically active regions in the world. Recently, the compiled
historical catalog lists or identifies 2247 events for the time period from 2000
BC to 1900 AD with 212 earthquakes with an intensity (Io) of nine (IX)
greater during the last 4000 years (Soysal et al., 1981; Ambraseys 2009;
Albini et al., 2013) Figure 1. During the last century and in the instrumental
catalog (1900-2012) 203 events are registered with a magnitude of 6.0 and
greater in Anatolia and the surrounding region (Kalafat et al., 2011;
Kadirioglu et al., 2016; Duman et al., 2018) (Figure 2).
23
Figure 1 - Primary, active faults (Emre et al., 2013) and the distribution of historical
earthquakes (BC 2000-AD 1900) in Turkey and surrounding areas (Modified from Duman
et al., 2018).
Figure 2 - Primary, active faults (from Emre et al., 2013) and the instrumental seismicity
(1900 -2012) for earthquakes M>4.0 in Turkey and surrounding areas (Modified from
Duman et al., 2018).
24
Epicenters of the major earthquakes are particularly concentrated in the
segment boundaries of the main active faults. Major structures related to
strike-slip tectonic regime are the dextral (right strike-slip)North Anatolian
and sinisterly (left strike-slip) East Anatolian Fault systems, along with the
interim Anatolian plate has been slipped in WSW direction onto easily
deductible oceanic lithosphere of the Eastern Mediterranean Sea since the late
Early Pliocene (Reilinger et al., 2006, 2010; McKenzie 1978; Le Pichon and
Angelier 1979; McClusky et al., 2000; engr et al., 1984, 1985; Kocyigit et
al., 1999).
The majority of the seismic activity is concentrated along the North
Anatolian Fault (NAF) and the East Anatolian Fault (EAF) zones resulting
from the westward movement of the Anatolian plate due to the collision of
the Arabian and Eurasian plates (engr et al., 1984, 1985). The NAF is a
large right-lateral strike-slip fault which is continuing roughly 1200 km from
the Karliova junction in the east and to the northern Aegean Sea in the west
in Turkey (Barka 1992). A sequence of devastating earthquakes occurred on
this fault from east to west, starting with the Erzincan earthquake 1939 and
followed by seven damaging earthquakes larger than M>7.0; 1942 Erbaa-
Niksar, 1943 Tosya, 1944 Bolu-Gerede, 1957 Abant, 1967 Mudurnu and
finally 1999 İzmit and Düzce in the 20th century. İstanbul, situated 20 km
from the NAFZ in the Marmara Sea, is the largest city in Turkey; the area has
experienced high levels of earthquake ground motion since the beginning of
human history. Within the past centuries, four earthquakes of M7.6 (1509,
1719, and 1766) and M7.0 (1894) situated in the Marmara Sea have generated
intensities up to ten to eleven (X-XI) in the city (Ambraseys, 1971, 2002).
Recent studies have shown that the probability of having an earthquake
(M≥7.0) close to İstanbul rises from a Poisson estimate of 35% to values of
47% under the time-dependent interaction model during the 30 years starting
from 2014 (Murru et al., 2016).
The Eastern North Anatolian fault is also capable of producing large
magnitude earthquakes and has experienced a sequence of damaging events
including 1949 Karliova M6.8, 1971 and 2003 Bingol M6.9, M6.4, 2010
Elazı-Karakoan M6.1 earthquake (Saroglu et al., 1992; Nalbant et al., 2002;
rgülü et al., 2003; engr et al., 2005; Bulut et al., 2012). Compression
deformation in Eastern Anatolia has resulted in thickening of the crust and
includes dominantly reverse faults. The area was exposed to major damaging
historical earthquakes in 1111, 1648, 1715, 1881. In 1976, M7.3 an
earthquake located near the town of Caldıran, 20 km northeast of Muradiye,
caused severe damage in the Van Province killing around 3840 people and
25
leaving around 51,000 homeless (Copley and Jackson 2006; Reillinger et al.,
2006). Recently in 2011, M7.1 an earthquake occurred close to the city of
Van, killing around 604 people and once again leaving thousands homeless
(AFAD, 2011; Akinci and Antonioli, 2012).
Moreover, subduction of the African plate beneath the Aegean plate
alongside the Hellenic trench has generated a back-arc N-S directed
extensional regime and associated normal faulting in the Western Anatolia
(Jackson and McKenzie, 1984; Westaway 1990). In the past fifty-year major
earthquakes caused extensive damages and destruction in the zone. For
example: 1949 Edremit-Ayvalik M7.0 destroyed nearly 5000 buildings; 1953
Yenice-Gonen M7.4 destroyed 1800 buildings; the 1969 Alasehir M6.9,
damaged 3700 buildings and 1970 Gediz M7.3, destroyed 9500 buildings
and killed the total 1400 people (Akinci et al., 2013 ).
Therefore, an understanding of earthquake structure is an important and
unique way to assess and evaluate the earthquake hazard estimation and
mitigate losses due to earthquake in Turkey.
3. Building Inventory in Turkey
Seismic risk reduction efforts and strategies require gathering detailed
information on the building inventory as well as the seismic hazard level in a
country. Distribution of the population in a country affects urbanization and
eventually the building inventory. The migration of people from rural areas
into cities has always created demand in the construction of residential
buildings. Depending on the numbers of people, the construction progress and
quality can get out of control. In addition to migration, the population growth
rate is another factor for the building inventory. Turkey has been a steadily
growing country. The population increased from 40 million to 85 million in
40 years period between the 1970s and the 2010s. Thus, the building
inventory in Turkey has been affected from both migration from rural areas
to cities and the excessive population growth.
The recent building inventory can be divided into two main classifications;
construction materials and height (Crowley et al 2012). The construction
material is a key parameter in understanding rapid housing. If the demand of
housing is huge, it is inevitable that the cheap and widely available materials
are preferred. Concrete, particularly reinforced concrete, is a good example
of this statement. Combining the cement, limestone, aggregate and water with
reinforcing rebar is a relative new technique in building construction.
26
Compared to the traditional timber and stone masonry buildings those have
been around for centuries, reinforced concrete has been used in buildings
since the 1940s and its use has growth proportionally since then. From recent
research (Demirciolu, 2009) the total number of buildings in Turkey is
7,513,380 in which 51% of the buildings are made of reinforced concrete,
Figure 3a.
(a)
(b)
Figure 3 - Number of buildings with respect to (a) construction materials and (b) stories
(URL1).
The cultural characteristics of societies influence social life. In countries
like Turkey, it is a common convention to own a strong and durable house for
a long time. Hence, people invest on the properties that they feel would last
for a very long time and would protect them from all kind of natural threats.
Based on this convention, reinforced concrete is the commonly preferred
building material due to its cost-effective production, widespread availability
and cheap labor cost in countries suffering from natural hazards. Moreover,
27
the weight and the toughness of concrete contribute in making people feel
psychologically safe and comfortable.
The other classification in describing the building inventory is the building
height in terms of number of stories. There is a strong relation between the
number of stories of a building and its location. Since the area needed for
buildings are expensive in the vicinity of the city centers, the mid-rise
buildings are generally preferred rather than low-rise buildings. It is also true
that public services such as transportation, electricity and water procurement,
sewage etc. is broadly provided to high-populated districts. Thus, the number
of stories in relation to the population is a valuable parameter in
understanding the structural risk in the country. The distribution of the
building height, named in low, mid and high-rise is given in Figure 3b. The
number of stories is considered as a realistic value in the definition of the
building height.
The number of high-rise buildings is significantly lower than low and mid-
rise buildings. Therefore, the spatial distribution of low and mid-rise
buildings provides a better understanding in the description of the inventory
within the perspective of urban renewal. The building density distribution for
all Turkey for low and mid-rise buildings is given in Figure 4a and b,
respectively. It is apparent that building density is great in the major cities
such as İstanbul, Ankara, İzmir, Antalya, Bursa.
(a)
28
(b) Figure 4 - Distribution of (a) low and (b) mid-rise buildings in Turkey.
Comparing the building inventory distribution maps (Figure 4a and b) with
the seismic hazard map including most active fault lines (Figure 2) distinctly
displays the most vulnerable areas in Turkey. Keeping in mind the huge and
rapid increase of the building inventory and the seismic hazard, it is evident
that major risk mitigation strategies are essential.
4. Urban Renewal Law in Turkey
Starting from the early 1970s, the population in urban areas increased
rapidly and consequently serious social and economic arose. The major
problem that the big cities faced was the need for accommodation and
infrastructure for the new residents. Thus, the construction industry had a
huge opportunity to meet the high demand in housing in urban areas all
around Turkey, especially in İstanbul. However, the opportunity came with
severe problems both in design and construction terms. The main problems
can be listed as;
1. Huge demand for reconstruction in a very short time,
2. Lack of modern seismic design codes for professional design
engineers,
3. Lack of a peer review process in seismic design of buildings,
4. Inadequate quality control in construction progress,
5. Low quality of workmanship.
29
In addition to the above-mentioned issues, urban planning strategies and
regulations were not compatible with proper seismic risk mitigation
principles (Özdemir and Yılmaz, 2011). This situation has lasted for more
than three decades resulting with a very huge vulnerable and seismically risky
building stock all around Turkey (Green, 2008).
The year of 1999 can be named as the “turning point” in Turkish
earthquake history. Two major earthquakes in the north western part of
Turkey, 1999 Kocaeli and Düzce earthquakes hit the most urbanized and
industrialized cities of İstanbul, Kocaeli, Düzce and Yalova. The results of
these earthquakes were catastrophic for Turkey both on a social as well as an
economic level. The country suffered a lot from the damages and losses
(Durukal and Erdik, 2008). Immediately in the following months, strict
measures and actions in the education, construction, legislations and design
codes were planned for a resilient society. The planned actions are
chronologically listed in the Table 1.
Table 1 - The major actions in disaster resilience
Year
Action
1999
Marmara Earthquakes (M7.4 on 1999-08-17 and M7.2 on 1999-11-12)
2000
Establishment of Turkish Natural Catastrophe Insurance Pool
2004
Rehabilitation of the public schools in İstanbul
2006
Initiation of the İstanbul Seismic Risk Mitigation and Emergency
Preparedness Program
2007
Revision of Turkish Earthquake Code
2008
Rehabilitation of the highway and road bridges
2009
Establishment of Disaster and Emergency Management Directorate
2010
Rehabilitation of the public schools in İzmit
2011
Declaration of the National Earthquake Strategy Plan until 2023
2012
The Law of Transformation of Areas under the Disaster Risks (No: 6306)
2013
Guidelines for the use of seismic isolations in City Hospitals and Seismic
Risk Assessment Code for the Buildings
2014
Project for updated Turkish Earthquake Risk Map
2015
Initiation of the National Disaster Response Plan
30
2016
Detailed revision of the Turkish Seismic Design code Draft
2016
Revision of the Design Code of the Steel Structures
2018
New Turkish Seismic Hazard Map & revision of Turkish Building Seismic
Code
2019
Revision of the Seismic Risk Assessment Code for the Buildings
Among these actions, the Law of Transformation of Areas under the
Disaster Risks (No: 6306) has been most effective in terms of practicality and
applicability. The law is applied in three phases. It starts with the assessment
of the building and ends with re-construction of the new building with
reduced bureaucratic procedures. The phases are summarized in Table 2 and
visualized in Figure 5.
Table 2 - Urban Renewal Application in Turkey can be divided into three phases.
Phases
Steps
Phase 1
Assessment of the
building
In this phase, relevant official documents of the building are
collected, and licensed engineering firms perform engineering
inspections and calculations in order to prepare an assessment
report that involves seismic safety of the building. As the last
step of this phase, the assessment report is delivered to the local
authority.
Phase 2
Seismic safety
assessment approval of
the building by the
municipality
Local authority accepts the evaluation report and informs
the property registration office. Property owners receive a
warrant from the local authority for demolishing or retrofitting
options. Once two thirds of the owners agree on the retrofit
option, the municipality is informed accordingly. Otherwise,
the municipality will have the right to cancel essential services
such as electricity, gas and water. Following these measures,
the property owners are expected to evacuate the building to be
demolished within two months. In case of no evacuation, the
owners are forced to leave the property under the control of the
police officers.
Phase 3
Demolishing and
rebuilding the new
property
Demolishing the building is arranged by the owner or
his/her representative. Government provides nonrefundable
financial support to the rent cost up to 18 months. During this
period, the building owners are strictly supposed to either
retrofit or rebuild the new building. General practice is to agree
with a contractor to get this engineering services.
31
The steps of the Law of Transformation of Areas under the Disaster Risks
are illustrated in Figure 5.
Figure 5 - Steps of the Urban Renewal Law in Turkey.
The law legislated by the ministry of Environment and Urbanization
delegates the municipalities for the implementation. Initially, the cities of
İstanbul, Bursa and İzmir were selected as the areas for preliminary
implementation of the law. Since 2012, many cities have benefitted from the
law. In early 2019, the ministry requested that all municipalities establish
their own urban renewal strategies in their most vulnerable zones. This
request was intended to extend the application of the law to almost every part
of Turkey rather than major cities to compliment the national mitigation
action.
5. Process and Lessons Learned from the Urban Renewal Law in
Turkey
As of 2019, a large number of citizens have benefitted from the urban
renewal law. Too many lessons learned within that 7 years of application.
Based on the official statistics, 174,661 buildings have been assessed by
licensed engineering firms. Among these buildings, only 1% was found to be
safe in terms of seismic risk. The majority of the assessed buildings are
Reinforced Concrete and Masonry type buildings, both 39% (Figure 6).
32
Figure 6 - Distribution of the construction material in the assessed buildings all over
Turkey (URL1).
Considering the distribution of the assessed buildings in the city, İstanbul
is significantly leading with 60%, in the application of the law, Figure 7.
Figure 7 - Distribution of the assessed buildings in major cities (URL1).
The age of the assessed building is significant; the most vulnerable
buildings were constructed between 1970s to 1990s where the huge demand
occurred, Figure 8a.
33
(a)
(b)
(c)
Figure 8 - Construction year of the assessed buildings (a) all, (b) Reinforced Concrete,
(c) Masonry (URL1).
The distribution of construction years of the RC buildings given in Figure
8b, has great similarity with the overall distribution in Figure 8a. This
34
indication can be evaluated as the proof of the low quality RC building
construction in the period. However, the trend has not been observed in the
masonry building type, Figure 8c. The number of the assessed buildings
represents 2% of the entire building inventory in Turkey. Whereas, 131,715
buildings, which is 76% of all assessed buildings, have been demolished. The
values reveal that more time and effort are needed to reach the ultimate
resilient society.
5.1. Examples for good practice
İstanbul is the largest city where the urban renewal cases occurred.
Perhaps, the population and the low-quality residential buildings are the main
reasons. The Ministry of Environment and Urbanization with the
collaboration of municipalities have declared 40 different zones in 16 districts
for the preliminary areas for the renewal (Figure 9). The total area of the
selected zones is about 11 million m2. The largest two zones are 1,582,476
m2 and 1,341,759m2 on European and Asian sides of the city, respectively.
Figure 9 - Urban Renewal Areas declared in İstanbul (URL2).
The zone in the Asian side is called the Fikirtepe of Kadıky district. The
building stock in the zone is composed of low-rise buildings that were
35
constructed in the mid 1980s. The strategy followed in Fikirtepe relies on the
demolishing the 1,500 small buildings to clear the area, and then the
construction of high-rise buildings according to the most recent design codes
and engineering practices, Figure 10. The total budget of the renewal was
predicted as €4 billion. The huge budget is supposed to be funded by private
investors rather than government budgets. The private investors are supposed
to prepare the new building design projects and conclude agreements with the
property owners. Most of the property owners had legal agreements with the
investors either for payment or for ownership of the new buildings.
Figure 10 - Conceptual view of Fikirtepe district after urban renewal.
The view of Fikirtepe has significantly changed from a poor environment,
Figure 11, to a modern environment, Figure 12.
36
Figure 11 - The satellite views of Fikirtepe district in 2007 and 2019 (URL3).
As of the current situation in 2019, work in Fikirtepe has not been
completed, but the progress made has been an example of the Law for the
zones, Figure 14.
Figure 12 - Actual view of Fikirtepe District (as of May 2019)
5.2. Examples for bad practice
Even though the zones for large-scale renewal were declared, the law has
been applied to single buildings. This application has been both positive and
negative consequences. The positive side is that the individual buildings in
the renewal zones have benefitted from the law for demolishment.
Demolishing the detached buildings with seismic risk is a common practice.
However, demolishment of a single building in the non-detached buildings
does not make a sound in the seismic risk reduction in the urban areas. For
example, the building in the middle of Figure 13 demolished within the
regulations given in the law. This action is not a real renewal success within
the perspective of reducing the risk for an area but only for a single property.
37
The remaining buildings continue to carry the seismic risk potential for
themselves or their surroundings.
Figure 13 - Bad practice for urban renewal.
6. Conclusion
As one of the major actions to mitigate the vulnerability of the building
stock in Turkey the Turkish Government has issued “The Law of
Transformation of Areas under the Disaster Risks (Law No. 6306)”, which
came into law in 2012 (published in the Official Gazette of 31.5.2012,
no.28309). The scope of the law is to determine the procedures and principles
regarding the rehabilitation, clearance, and renovations of areas and buildings
at disaster risks in accordance with relevant standards with a view to create
healthier and safer living environments in urbanized areas. The number of the
expected building renewal is 6.5 million all over Turkey. This is the largest
housing project in the world as a part of seismic hazard mitigation.
After seven years of the law, a huge and valuable experience has
accumulated through these processes, which could act as a useful example for
countries with similar seismic risk. The main points learned from the short
history of the urban renewal law can be concluded as:
- Strategically individual assessment preferred to large areas was
unsuccessful,
- New constructions are not satisfactory/appealing due to the smaller
room size,
38
- Economic loss of the property owners due to illegally constructed
stories,
- Application is more focused on areas where apartment prices are
higher than the others,
- Application in the zones like Fikirtepe provided not only seismic
safety but also improved infrastructure to the region,
- The law has provided a permanent plan for construction industry
causing large economic benefits and increase in real estate values.
In order to provide sustainable urban renewal process for the coming years,
possible actions can be recommended as;
- Increasing the economic support ( rent for other building during
construction) of the building owners by the central government,
- Modifications and update in application process of urban renewal in
terms of bureaucracy and regulations,
- Providing benefits for the applicants in terms of land use in
suburb/rural areas such as extra stories, larger building base area, exemptions
in disaster insurance premiums in the cases of innovative technologies (base
isolation, damper, etc.) are used in seismic design or seismic design
performance level is taken higher than the seismic code requirement.
The experiences with good and bad examples of economic and engineering
approaches applied in the last seven years are invaluable resources for
countries suffering from similar hazardous risks for possible adaptations into
their own risk mitigation strategies.
References
AFAD, 2011. “Updates, Reports, Damage Statistics Van Earthquake”,
various reports and data updates from 23/10 to 2/11, available from
URL:http://www.afad.gov.tr.
Akinci, A., S. D’Amico, L. Malagnini, and A. Mercuri, 2013, Scaling
Earthquake Ground Motions in Western Anatolia, Turkey, Physics and
Chemistry of the Earth 63, 124135.
Albini, P., R.M.W. Musson, A.A. Gomez Capera, M. Locati, A. Rovida,
M. Stucchi, and D. Viganò, 2013, Global Historical Earthquake Archive and
Catalogue (1000-1903), GEM Technical Report 2013-01 V1.0.0, 202pp.,
GEM Foundation, Pavia, Italy.
39
Ambraseys, N. 2002, The seismic activity of the Marmara Sea region over
the last 2000 years, Bull. Seismol. Soc. Am., 92, 1 18.
Ambraseys, N. N., 1971, Value of historical records of earthquakes,
Nature, 232, 375379.
Ambraseys, N. N., 1988, “Engineering seismology: part I”, Earthquake
engineering & structural dynamics, 17(1), 1-50.
Ansal, A., Akinci, A., Cultrera, G., Erdik, M., Pessina, V., Tuzce, G. and
Ameri, G., 2010 Loss estimation in İstanbul based on deterministic
earthquake scenarios in the Marmara Sea region, Turkey”, Soil Dynamics
and Earth. Eng., 29.4,699-709.
Baris Binici, Ahmet Yakut, Guney Ozcebe, and Atila Erenler, 2015,
Provisions for the Seismic Risk Evaluation of Existing Reinforced Concrete
Buildings in Turkey under the Urban Renewal Law”, Earthquake Spectra,
Vol. 31, No. 3, pp. 1353-1370.
Barka, A. A., 1992, The North Anatolian fault zone, Ann. Tecton. 6,
164195.
Bulut, F., M. Bohnhoff, T. Eken, C. Janssen, T. Kılı., and G. Dresen, 2012,
The East Anatolian Fault Zone: Seismotectonic setting and spatiotemporal
characteristics of seismicity based on precise earthquake locations, J.
Geophys. Res., doi:10.1029/2011JB008966.
Copley, A., and Jackson, J., 2006, Active tectonics of the Turkish-Iranian
Plateau”, Tectonics, 25(6).
Crowley, H., Özcebe, S., Spence, R., Foulser-Piggott, R., Erdik, M., &
Alten, K., 2012, Development of a European building inventory database.
15th World Conference on Earthquake Engineering, 24-28.
Demircioglu, M. B., 2015, Earthquake hazard and risk assessment for
Turkey, PhD. thesis. Bogazici University, Turkey.
Duman TY, Çan T, Emre Ö, Kadiriolu FT, Başarır Baştürk N, Kılı T,
Arslan S, Özalp S, Kartal RF, Kalafat D, Karakaya F, Erolu Azak T, Özel
NM, Ergintav S, Akkar S, Altınok Y, Tekin S, Cingöz A, Kurt A., 2018,
“Seismotectonic database of Turkey”, Bulletin of Earthquake Engineering,
16(8), 3277-3316.
Durukal, E., & Erdik, M., 2008, Physical and economic losses sustained
by the industry in the 1999 Kocaeli, Turkey earthquake”, Natural Hazards,
46(2), 153-178.
Green, R. A., 2008, Unauthorised development and seismic hazard
vulnerability: a study of squatters and engineers in İstanbul, Turkey”,
Disasters, 32: 358-376.
40
Gunes, O., 2015, Turkey's grand challenge: Disaster-proof building
inventory within 20 years”, Case Studies in Construction Materials, ISSN:
2214-5095, Vol: 2, Page: 18-34.
Jackson J., and D.P. McKenzie, 1984, Active tectonics of the Alpine
Himalayan Belt between western Turkey and Pakistan, Geophys. J. R.
Astron. Soc., 77, 185-246.
Kadiriolu, F.T., Kartal, R.F., Kılı, T. et al, 2018, An improved
earthquake catalogue (M 4.0) for Turkey and its near vicinity (19002012)”,
Bull Earthquake Eng (2018) 16: 3317.
Kalafat D, Güneş Y, Kekovalı K, Kara M, Deniz P, Yılmazer M., 2011, A
revised and extended earthquake catalogue for Turkey since 1900 (M 4.0),
Boazii University, Kandilli Observatory and Earthquake Research
Institute, İstanbul.
Kocyigit A., H. Yusufoglu, E. Bozkurt, 1999, Evidence from the Gediz
graben for episodic two-stage extension in western Turkey, Journal of the
Geological Society, London, 156, pp. 605-616
Le Pichon, X., & Angelier, J., 1979, The Hellenic arc and trench system:
a key to the neotectonic evolution of the eastern Mediterranean area”,
Tectonophysics, 60(1-2), 1-42.
McClusky, S., S. Balassanian, A. Barka, C. Demir, and S. Ergintav, 2000,
Global positioning system constraints on the plate kinematics and
dynamics, J. Geophys. Res., 105, 56955791.
McKenzie, D., 1978, “Active tectonics of the AlpineHimalayan belt: the
Aegean Sea and surrounding regions”, Geophysical Journal International,
55(1), 217-254.
İstanbul Metropolitan Municipality, 2003, İstanbul Earthquake Master
Plan, İstanbul Metropolitan Municipality, Department of Earthquake Risk
Management and Urban Improvement.
Murru, M., Akinci, A., Falcone, G., Pucci, S., Console, R., and Parsons, T.
2016, “M≥ 7 earthquake rupture forecast and timedependent probability for
the Sea of Marmara region, Turkey”, Journal of Geophysical Research: Solid
Earth, 121(4), 2679-2707.
N. Ambraseys 2009, Earthquakes in the Mediterranean and Middle East:
a multidisciplinary study of seismicity up to 1900, Cambridge University
Press. (ISBN 9780521872928).
Nalbant, S. S., J. McCloskey, S. Steacy, and A. A. Barka, 2002, Stress
accumulation and increased seismic risk in eastern Turkey, Earth Planet.
Sci. Lett., 195, 291298.
41
O. Ergunay, 2007, Turkey's disaster profile, TMMOB Afet Sempozyumu,
Turkish Chamber of Civil Engineers, Ankara, Turkey [in Turkish].
rgülü, G., M. Aktar, N. Turkelli, E. Sandvol, and M. Barazangi, 2003,
Contribution to the seismotectonics of Eastern Turkey from moderate and
small size events, Geophys. Res. Lett., 30(24), 8040.
Ozdemir, O., & Yilmaz, C., 2011, Factors affecting risk mitigation
revisited: the case of earthquake in Turkey, Journal of Risk Research, 14(1),
17-46.
Reilinger, R., McClusky, S., Paradissis, D., Ergintav, S., and Vernant, P.,
2010, Geodetic constraints on the tectonic evolution of the Aegean region
and strain accumulation along the Hellenic subduction
zone, Tectonophysics, 488(1-4), 22-30.
Reilinger, R., McClusky, S., Vernant, P., Lawrence, S., Ergintav, S.,
Cakmak, R., and Nadariya, M., 2006, GPS constraints on continental
deformation in the AfricaArabiaEurasia continental collision zone and
implications for the dynamics of plate interactions”, Journal of Geophysical
Research: Solid Earth, 111(B5).
Sengör, A. M. C., N. Görür, and F. arolu, 1985, Strike-slip faulting and
related basin formation in zones of tectonic escape: Turkey as a case study,
in Strike-slip Deformation, Basin Formation, and Sedimentation, Society of
Economics Paleontologist and Mineralogist, Special Publication, Vol. 37 (in
honor of J. C. Crowell), 227264.
Sonmez Saner, T., 2015, Seismic vulnerabilities and risks for urban
mitigation planning in Turkey, Nat Hazards, 78, 1387.
Soysal H., Sipahioglu S., Kolcak D. and Altinok Y., 1981, Historical
earthquake catalog for Turkey and surrounding region. Unpublished report
No. TBAG-341 submitted to the Scientific and Technical research Council of
Turkey (in Turkish).
arolu, F., Emre, Ö., & Kuşu, I., 1992, The east Anatolian fault zone
of Turkey”, Annales Tectonicae, Special Issue-Supplement to Volume VI, 99-
125.
engr, A. M. C., O. Tüysüz, C. İmren, M. Sakın, E. Eyidoan, N. Grür,
X. Le Pichon, and C. Rangin, 2005, The North Anatolian Fault: A new look,
Annu. Rev. Earth Planet. Sci., 33, 37112.
URL1: Ministry of Environment and Urbanization, www.csb.gov.tr,
Accessed on January 2019.
URL2: https://İstanbulakdm.csb.gov.tr/İstanbul-da-ilan-edilen-riskli-
alanlarin-dagilim-haritasi-i-3750, Accessed on May 2019.
URL3: Google Earth http://earth.google.com, Accessed in May 2019.
42
Westaway, R., 1990, Block rotation in western Turkey: 1. Observational
evidence”, J. Geophys. Res., 95(B12), 19857 19884,
doi:10.1029/JB095iB12p19857.
43
2. A Collection of Statistical Methods for Analysis of the
Disaster Damages and the Seismic Regime
Vladilen F. Pisarenko1, Mikhail V. Rodkin2
Abstract
In this paper, we present a collection of statistical methods addressing
heavy tailed distributions. The empirical distributions of damage from natural
disasters, both in terms of material losses and fatalities, are often modelled by
theoretical distributions with a heavy power-law tail. The distribution of
earthquake energy (seismic moments) is another example of such a heavy
tailed distribution. The statistical methods that we discuss here allow to
perform an analysis of empirical distributions at different levels depending on
the amount of available data. We perform a detailed analysis of heavy tailed
distribution using the theory of extreme values, and discuss the related
examples. The presented methods of analysis of heavy tailed distributions
constitute a toolbox, which can be useful in a number of practical
applications.
Keywords: disaster related damage; power-law distribution, heavy tailed
distribution; theory of extreme values; seismic regime.
Introduction
Most of the damages produced by natural disasters, such as earthquakes or
typhoons, are caused by rare and strong events, which release large amounts
of energy. Many authors emphasize the universality of the power law
distribution when characterizing natural hazards [Turcotte & Malamud, 2004;
Newman, 2005; Sornette, 2000; Kijko, 2011; O’Brien et al., 2012; Liu et al.,
2014; Kousky, 2014; Smit et al., 2017; Rougier et al.,2018]. Power law
distributions are observed for damages caused both by natural and by man-
1 Corresponding Author; Institute of Earthquake Prediction Theory and Mathematical
Geophysics, Russian Academy of Sciences, Profsoyuznaya 84/32, Moscow 117997, Russia.
E-mail: pisarenko@yasenevo.ru.
2 Institute of Earthquake Prediction Theory and Mathematical Geophysics, Russian Academy
of Sciences, Profsoyuznaya 84/32, Moscow 117997, Russia. E-mail: rodkin@mitp.ru.
44
made disasters, as well as for energy values and for seismic moments of
earthquakes [Sornette, 2000; Clauset et al., 2009]. The main contribution to
the damage caused by disasters with a heavy-tailed distribution is due to rare
strong events. Therefore, the estimation of magnitude distribution in the
uppermost range plays a key role in the problems related to natural disasters
hazard assessment and mitigation. The statistical theory of extreme values
provides a solid mathematical base for such estimations. In this work, we
mainly focus on the problem of seismic risk assessment, since the
corresponding data sets are more abundant and well suited for the
demonstration of different statistical methods of heavy tailed distribution
analysis.
The empirical distribution of damage produced by various types of
catastrophes can often be described by the Pareto power-law:
F (x) = 1 (a / x)β, x ≥ a. (1)
Such power-law distribution with the exponent β ≤ 1 has an infinite
mathematical expectation E(X):
E(X) =  
  
.
Sample estimates of parameter β in the range β 1 are readily obtained from
empirical data sets. Of course, all real losses and earthquake energies are
finite. Thus, the true law of distribution in the range of rare strong events
should deviate from the Pareto law (1). The finiteness of a random value can
be modelled by a distribution with a maximum possible value [Cosentino et
al. 1977; Kijko and Sellevoll, 1989; 1992; Pisarenko et al., 1996; Ward, 1997;
Burroughs & Tebbens, 2001; Kijko and Singh, 2011; Vermeulen & Kijko,
2017] or by a distribution with special rapidly decaying factor in the extreme
range [Kagan, 1999]. We use below the finite distributions resulting from the
extreme value theory: the General Extreme Value distribution (GEV) and the
Generalized Pareto Distribution (GPD) with negative form parameter (see
below). They naturally appear as the limit distributions for maxima of
observed random values [Gumbel, 1958].
This article presents an overview of the results published in a series of
works by the authors [Pisarenko & Rodkin 2010; 2014; 2015; 2017;
Pisarenko et al., 2014; 2017] and it contains as well several new results.
45
1. A faster-than-linear growth of cumulative damage with time and
its possible incorrect interpretation.
Cumulative damage from natural disasters frequently exhibits a nonlinear
(faster than linear) growth with time. This growth is observed for both the
number of disasters and for the associated losses (examples are shown in Figs.
1, 2).
Figure 1 - The annual number of world natural disasters as a function of time during the
period 19752005 based on EMDAT, [Pisarenko & Rodkin, 2010]; the definition of a
country-level disaster is given in EM-DAT Glossary and means a disaster that has affected
a particular country, if several countries were affected the disaster is indicated several times.
46
Figure 2 - Annual losses from natural disasters (including individual most damaging
events) during the period 19752005 based on EMDAT, [Pisarenko & Rodkin, 2010].
Whereas the increase of the number of documented catastrophes may be
attributed, at least partly, to the development of registration systems and to a
greater availability of the corresponding information, the non-linear growth
of cumulative damage requires a separate explanation. Some authors attribute
such faster-than-linear growth to such factors as worsening ecological
conditions, urbanization, population growth, and climate change [Berz, 1992;
Osipov, 1995; 2002; Seneviratne et al., 2012]. However, such growth can as
well be observed in a stationary situation, provided the distribution of damage
values has a heavy tail [Pisarenko, 1998; Pisarenko and Rodkin, 2010; 2014].
Let us analyze this case in more detail.
As noted above, the empirical distributions of damage produced by
different types of natural disasters can be approximated by the Pareto power-
law distribution (1). Examples of distribution of damage values from floods,
hurricanes and earthquakes in the USA are shown on Fig. 3. The estimates of
the exponent β obtained by the maximum likelihood method are less than one,
thus we are dealing with heavy tailed distributions.
47
Figure 3 - Mean numbers N of events per year with economic losses greater than L ($
USA): floods 1986-1992 (1); earthquakes 1900-1989 (2); hurricanes 1986-1989 (3).
The fitted power-law complementary distributions 1 F(x) are shown by lines, the values
of exponents are: βn = 0.74 (floods): βn = 0.41 (earthquakes), and βn = 0.98 (hurricanes)
[Pisarenko & Rodkin, 2010].
The maximum likelihood estimate of parameter β equals
βn = n/( Σ ln(xi/a) ), (2)
where the sum is taken over all xi ≥ a, i=1, …n.
Now let us assume that the number of events n is a random value obeying
the Poisson law
Pr{n=k}= (λT)k/ k! ; k = 0,1, 2, 3,…
with parameter λT (λ is the intensity of corresponding Poisson process, T is
time of observation). Let us consider the median µ(T) of the largest event
that occurred within time interval [0 T]. The median µ(T) of the maximum
event over time T for the Pareto law equals [Pisarenko & Rodkin, 2010]
µ(T) = a (λT/ln(2))1/β . (3)
Let us denote by Σ(T) the sum of damages and by R(λT, β) the ratio
Σ(T)/µ(T):
Σ() = R(λT, β) × µ(T). (4)
48
It can be shown that R(λT, β) remains limited as T tends to infinity [Feller,
1966]. This property can be expressed in a different way: for distributions of
type (1) with β < 1, the one strongest event is of the same order as the
cumulative sum of all other events. It follows from the above formulas that
the mean value E[Σ())] increases linearly with T for β > 1 and proportional
to T1/β for β ≤ 1 :
E[Σ())] = C(β,T)·T max(1, 1/β) ; C(β,T) is limited. (5)
As it can be seen from (3) - (5), both the single maximum event and the
cumulative sum of all the events increase with time in a nonlinear manner,
that is proportionally to T1/β , if β < 1; and that is true even for a stationary
process.
Of course, one should distinguish the statistical effect of a non-linear
growth due to a heavy tail, from the effect of a real non-stationarity of the
regime of disasters, caused for instance by climate change, by an increased
vulnerability of the technological infrastructure to disasters, or by other
factors.
As we noted above, a distribution with the exponent β<1 is not applicable
to physical quantities, but for relatively small data sample sizes one can obtain
sample estimates βn <1 (see Fig.3). This may occur because small data
samples with β exceeding unity are hardly distinguishable from samples with
β below unity. This effect can be clearly illustrated if we consider the
Gutenberg-Richter (G-R) and the truncated G-R distributions, FGR(x) and
FTGR(x) respectively. FTGR(x) is bounded by some Mmax value. One can
achieve an arbitrarily small departure | FGR(x) - FTGR(x) | within an arbitrarily
wide interval |x| ≤ A by choosing a sufficiently large value of Mmax. This
means it is practically impossible to distinguish these two distributions based
on small data samples. Thus, one can expect that certain statistical properties
of the heavy-tailed distributions cannot be reliably estimated from small or
intermediate data samples. It is very difficult to justify the choice of Mmax for
certain types of damages: the associated costs can be extremely high in
today’s world. But still, one may reasonably assume some kind of saturation
of the growth of damages, at least for some types of damages. This saturation
effect can reveal itself through a precise shape of the distribution’s tail (a bend
down of the tail), whereas a non-saturated sample distribution may include a
huge event, such as the mega-earthquake Tohoku in Japan (2011, M = 9.08).
Such events occur very rarely and can produce effects typical for the heavy-
49
tailed distributions: in particular, they may result in a non-linear growth of
total losses over a certain time interval.
2. Description of the bend down of heavy tail distributions an
approximated technique
This section provides a simple method for description of the bend down in
the distribution tail, suitable for the cases of small and moderate sample sizes;
see [Pisarenko and Rodkin, 2010] for details.
As discussed above, a characteristic feature of the power-law distributions
with β < 1 is a nonlinear initial growth of total damage with time. A similar
nonlinear growth is observed for the total released seismic energy (5):
E[Σ())] = C·T 1/β ; β < 1. (6)
However, this non-linear growth must slow down with time and eventually
become linear with respect to time, which is typical of stationary processes
with a finite average value:
E[Σ())] = C·T ; β ≥ 1. (7)
We define the transition time Ttp as the time moment when the nonlinear
cumulative growth becomes linear. Below we describe how one can evaluate
Ttp. Let us define the characteristic damage value Dtp, with recurrence time
Ttp corresponding to the transition point. In practice, this transition point can
be estimated with a large uncertainty since the available time series of disaster
related damage or those of released seismic energy are extremely variable. In
order to reduce this uncertainty, we suggest the following bootstrap procedure
(see for details [Pisarenko & Rodkin, 2010). We numerically simulate a
number of damage curves S(t), using a randomly shuffled original data sample
(say, 1000 samples). Then for each time t, we take the median MS(t) of the
bootstrap curves S(t) :
MS(t) = median < S(t) >.
The median MS(t) can be approximated by a simple regression relation
lg(MS(t)) = ao + a1 lg(t) + a2 lg2(t), (8)
50
where coefficients ao, a1, and a2 are estimated by the standard least squares
method.
For stationary time series with a finite average C we obtain
MS(t) = Ct, and lg(S(t)) = lg(t) + lg(C).
Thus,
d( lg(MS(t)) / d( lg(t) ) = 1. (9)
So, we can take this condition for the determination of the transition point
Ttp exploiting equations (6) - (7) as an assumed behavior of damage curves.
The transition point Ttp is determined as the smallest t value, for which the
equation (6) sill keeps satisfied. Thus,
a1 + 2·a2 lg(Ttp) = 1 ; lg(Ttp) = (1 - a1) /(2·a2). (10)
Using Ttp we determine the corresponding damage value Dtp:
Dtp = S(Ttp).
The value Dtp can be regarded as the characteristic damage typical of tail
events. The cumulative damage S(t) behaves approximatively linearly with t
for t > Ttp.
The procedure described above is applicable not only to the damage data,
but also to the disaster related death tolls and to other data. Results of
calculation of Ttp and Dtp for several available data sets are given in Table 1.
It can be seen that the characteristic Dtp of victims of natural disasters tends
to decrease with time, which can probably be attributed to the global
improvement of the technological infrastructures (better quality of civil
construction, better hazard mitigation policies etc.).
51
Table 1 - Characteristic values Ttp and Dtp for earthquakes and floods for different world
regions and time intervals.
Disaste
r type
Region,
country
Recur-
rence
time
Ttp,
years
Characte
-ristic
event,
Dtp,
number
of
fatalities
Maximum
event,
number of
fatalities
Data source,
data base
Earth-
quakes
Developed countries
Significant
earthquakes, NOAA,
https://www.ngdc.no
aa.gov/nndc/struts/for
m
1900-1959
33
95000
110000
1960-1999
30
24000
17000
Developing countries
1900-1959
40
270000
200000
1960-1999
60
260000
240000
Floods
North America and European Union
Em-dat,
The International
disaster database,
https://www.emdat.b
e
1950-1980
15
1500
650
1980-2005
10
500
200
SE and S
Asia
1984-2006
20
10000
6000
3. Description of the bend down of heavy tail distribution detailed
examination
A detailed statistical description of heavy tail distributions can be obtained
on the basis of the theory of extreme values. This approach needs, however,
substantially larger data samples than the simple method discussed above
[Pisarenko, Rodkin, 2010; Pisarenko et al., 2010; Pisarenko, Rodkin, 2014;
2015; 2017 et al.].
There are two limit distributions in the extreme value theory: the
Generalized Pareto Distribution (GPD) and the General Extreme Value
distribution (GEV). Each of these distributions depends on three parameters
and they are closely interconnected (see for details [Pisarenko and Rodkin
2010]). We use below the Generalized Pareto Distribution (GPD). The GPD
appears as the limit distribution of scaled excesses over sufficiently high
threshold values.
The GPD distribution has three parameters: the form parameter ξ, - <
ξ < + ; the threshold parameter h, - < h < + ; and the scale parameter
s, 0 < s < + .
The GPD distribution function has the form:
52
GPDh( x |
, s) = 1 [1 +(
/s)
(x h)] 1/
, ξ 0;
(11)
GPDh( x |0, s) = 1 exp(-(x –h)/s), ξ = 0;
There is a close connection between GPD and GEV distribution laws. A
Poisson flow of events obeying a GPD law is distributed in accordance with
the GEV distribution law with the same form parameter ξ. There are simple
formulas connecting other parameters of GEV and GPD (see [Embrechts et
al. 1997; Pisarenko & Rodkin, 2010]).
Using GPD distribution as limit distribution for a particular data sample
involves the estimation of three parameters (11). After this estimation is done,
the calculation of any statistical characteristic for the maximum event in any
future time interval is a routine procedure. However, the practical use of GPD
is often limited by a deficit of available data: our experience shows that a
reliable estimation typically requires at least 30 strong events in the
uppermost range.
Table 2 presents several estimates of GPD-parameters for a number of
disasters (see [Pisarenko & Rodkin, 2014] for details). The value (-1/ξ 1)
characterizes for negative ξ the rate of decay of the density to zero in vicinity
of the upper limit bound Mmax :
f(x) ~ 1/(Mmax -x) -1/ξ – 1, x Mmax ; -1 <ξ < 0.
The faster decay rates of the tail to the zero value (Table 2) are observed
for the economic losses produced by floods and hurricanes, whereas the
corresponding fatality and the injured/affected distributions have, as a rule,
smaller | ξ |-values, which corresponds to a slower decay of the tail.
The maximum Mmax of the GPD distribution with negative form parameter
ξ equals
Mmax = h s/ξ, ξ < 0. (12)
Thus, the lesser | ξ | the larger Mmax . Factually it means that in the case of
small | ξ | values, the Mmax is highly unstable and its estimate is not robust.
Instead of using unstable Mmax, we introduced a more stable estimate that
characterizes the uppermost range of the distribution, namely, the quantile of
the GPD-distribution. The quantile Q(q) of probability level q for the
distribution function F(x) is defined by the relation:
53
F(Q) = q. (13)
Thus, the quantile Qq is in fact the inverse function with respect to the
distribution function F(x). The continuous distribution function and the
quantile function are uniquely related. The distribution function is an integral
characteristic of random value (in contrast to the local characteristic the
probability density). So, we can consider the quantile as an integral
characteristic of the distribution’s tail. That is why the quantile gives more
stable and robust characterization of the tail than the point estimation Mmax
eq.(12), see discussion [Pisarenko & Rodkin, 2010; 2014]. Besides, one may
interpret the quantile Qq as an upper confidence bound of level q for
corresponding random value x:
Pr{ x < Qq } = q. (14)
For GPD-distribution the quantile Qq has the form:
Qq = h + (s/ξ)·[ (1-q)-ξ 1]. (15)
We calculated such quantiles Qq(T) for maximum event size in future time
interval T for a number of disasters [Pisarenko and Rodkin, 2014]. Table 2
presents Qq(T) with confidence level q=0.95 and time interval 10 years
Q0.95(10). The intensity of the seismic flow (number of events per unit time)
is designated as λ.
Looking at these estimates, one may conclude that economic losses are
strongly influenced by a rapid global development of the technological
infrastructure and by the population growth. For that reason, a reliable
forecast of such characteristics over long time spans is quite problematic. The
quantiles are more robust with respect to such uncertainties.
The last two rows in Table 2 summarize the results of the analysis of
annualized data. The aggregation of event sizes over one-year intervals
represents in essence a linear filtration (smoothing) of the corresponding time
series. That is why the tails of annualized distributions are as a rule less heavy
compared to the tails of original distributions. This fact can explain the trend
for higher absolute values of the form parameter | ξ | of annualized
distributions in Table 2, compared to the corresponding form parameters for
individual events.
54
Table 2 - Characteristics of disasters and form parameter of fitted GPD-law.
Lowe threshold
m0, Sample size n,
Intensity λ,
(1/year)
Form
parameter
ξ
Maximum
observed
value
Quantile
Q0.95 (10)
Seismic moment
Mw,CMT
catalog,
1976-2012
m0=6.8
n=324
λ=8.8
-
0.16±0.08
9.1
9.1
Earthquake
fatalities, Japan,
1900-2011
m0=3 persons
n=44
λ=0.339
-
0.26±0.11
142 807
persons
58
thousand
persons
Earthquake
injured, Japan,
1900-2011
m0=3
n=99
λ=0.884
-
0.37±0.06
103733
persons
75
thousand
persons
USA, fatalities
from floods,
1995-2011
m0=3
n=41
λ=1.11
-10-9
35
persons
53
persons
Affected in
floods, USA,
1995-2011
m0=500
n=52
λ=3.06
-
0.18±0.11
11 million
persons
17 million
persons
USA, fatalities
from tornadoes,
1953-2012
m0=20
n=53
λ=0.88
-10-9
1200
persons
1500
persons
Annual economic
losses from floods
in USA, 1940-
2011
m0=2.5
n=48
-
0.35±0.09
51.3,
109 $
46,
109 $
Annual economic
losses from
hurricanes in
USA, 1940-2011
m0=32
n=64
-
0.64±0.05
141,
109 $
123,
109 $
4. A two-branch model for distribution of earthquake magnitudes
As mentioned above, the practical use of GPD approach is limited by a
deficit of available data needed for reliable estimation of unknown
parameters. The two-branch model that we introduce below, allows us to
partially lift this limitation. The magnitude distribution of moderate size
earthquakes is well known to obey the normalized Gutenberg-Richter (G-R)
distribution law:
F(m) = 1 exp[-b·(m m0)]; m0 ≤ m
In terms of seismic moment values M0
lg(M0) =1.5m + 16.1 {dine·cm}
55
This law represents a power-law distribution:
Pr{M0 < z } = 1 C/z2b/3ln(10) ; C = const. (16)
The exponent in (16) is typically less than unity; therefore, this is a
distribution with a heavy tail. We discussed above the physical inconsistence
of infinite models with β≤ 1. The family of GPD-distributions includes
infinite distributions with heavy tails when ξ 0. But for ξ < 0 the GPD-
distribution is finite. We propose a model with a distribution that coincides
with the Gutenberg-Richter model in the lower and intermediate range, and
follows the GPD-distribution with ξ < 0 in the large event range [Pisarenko
and Rodkin 2020].
These two laws are smoothly attached to each other at some point h, so
that the overall distribution function F(m) of the two-branch model is:
C1{1 exp[-b·(m m0)]}; m0 ≤ m ≤ h ;
F(m)= (17)
C1{1exp[-b·(h m0)]}+C2{1–[1 + (ξ/s)·(m – h)]-1/ξ }; h m ≤ Mmax , ξ < 0.
In (17) the first branch corresponds to the G-R law, and the second one is
the GPD law with a negative form parameter ξ < 0. Here we a priori consider
only negative values ξ < 0, which corresponds to the finite distribution of
magnitudes, in contrast to models sometimes used for the magnitude
distribution. One can note that most of the estimates of ξ from real data sets
turned to be negative (see Table 2 and [Pisarenko, Rodkin, 2010; 2014]).
Model (17) contains 5 unknown parameters. The threshold h separates 2
branches of the model; b, m0, s, ξ are the model parameters; C1, C2 are
constants that depend on the above parameters and should ensure the
normalization of the distribution function F(m) and its continuity:
C1 = 1/{1 + bs·exp[-b(h-m0)] - exp[-b(h-m0)] },
C2= 1 - C1{1-exp[-b(h-m0)]}. (18)
Moreover, we impose that the distribution density function f(m) = F’ (m)
be continuous at the branches junction point m=h. From this condition we
get:
s = (1+ξ)/b. (19)
Finally, we obtain the following two-branch model:
56
C1{1 exp[-b·(m m0)] }; m0 ≤ m ≤ h ;
F(m) = (20)
3+C2{1[1+ 
(m h)]-1/ξ }; h ≤ m ≤ h - 
 , -1 < ξ < 0;
C1 = 1/(1+ξ exp[-b·(h m0)] );
C2 = (1 + ξ) exp[-b·(h m0)] /(1+ξ exp[-b·(h m0)] );
C3 = {1 exp[-b·(h m0