European Journal of Operational Research 230 (2013) 201211

Contents lists available at SciVerse ScienceDirect

European Journal of Operational Research

journal homepage: www.elsevier.com/locate/ejor

Invited Review

Review of recent developments in OR/MS research in disaster

operations management
Gina Galindo a,b, Rajan Batta a,
a
Department of Industrial and Systems Engineering, University at Buffalo (SUNY), Buffalo, NY 14260, USA
b
Fundacin Universidad del Norte, Barranquilla, Colombia

a r t i c l e i n f o a b s t r a c t

Article history: Potential consequences of disasters involve overwhelming economic losses, large affected populations
Received 25 January 2012 and serious environmental damages. Given these devastating effects, there is an increasing interest in
Accepted 22 January 2013 developing measures in order to diminish the possible impact of disasters, which gave rise to the eld
Available online 6 February 2013
of disaster operations management (DOM). In this paper we review recent OR/MS research in DOM.
Our work is a continuation of a previous review from Altay and Green (2006). Our purpose is to evaluate
Keywords: how OR/MS research in DOM has evolved in the last years and to what extent the gaps identied by Altay
Disaster operations management
and Green (2006) have been covered. Our ndings show no drastic changes or developments in the eld
Operations research
Management science
of OR/MS in DOM since the publication of Altay and Green (2006). Additionally to our comparative anal-
ysis, we present an original evaluation about the most common assumptions in recent OR/MS literature
in DOM. Based on our ndings we provide future research directions in order to make improvements in
the areas where lack of research is detected.
2013 Elsevier B.V. All rights reserved.

1. Introduction deal with uncertainties related to disaster locations and demands;

According to records from the International Federation of Red
Cross and Red Crescent Societies (IFRC), 7184 disasters took place
between 2000 and 2009. Some of the most memorable of these
events include the World Trade Center attacks in 2001, the tsunami
in Indonesia in 2004, hurricane Katrina in 2005, and the Haiti
earthquake in 2010. The IFRC estimates 986,691 millions of dollars
in economic damage from these events, 1,105,352 casualties, and
2,550,272,267 of affected people (IFRC, 2010). Moreover, the insur-
ance rm Munich Re reported that in 2010 natural disaster caused
more than 295,000 deaths and more than $130 billion in economic
damage (Our Amazing Planet, 2010). These overwhelming statis-
tics show the need of developing strategies in order to reduce
the impact of disasters for humankind.
Disaster operations, as dened by Altay and Green (2006), rep-
resent the set of activities performed before, during and after a
disaster in order to diminish its impact. Many of these activities
are intrinsically related to traditional OR/MS applications. For in-
stance, location of shelters in preparation for evacuations may be
addressed as a special case of location analysis; evacuation itself
may be better analyzed through the application of transportation
techniques; statistics and probabilistic models may be applied to
Corresponding author. Tel.: +1 7166450972; fax: +1 7166453302.
E-mail addresses: ggalindo@buffalo.edu (G. Galindo),
(R. Batta).
and in general, diverse OR/MS techniques may be applied to the
different stages of disaster operations management (DOM) in order
to provide a scientic approach in the process of decision making.
DOM has become a highly active eld in OR/MS. The amount of
papers in DOM published in OR/MS main stream journals during
1990s was more than twice of the amount published during the
previous decade (Altay and Green, 2006). Such popularity of
DOM in OR/MS has not decreased in subsequent years. On the con-
trary, a great amount of research has been published after the
occurrence of the World Trade Center attack in 2001 and the
2004 tsunami in the Indian Ocean (Natarajarathinam et al.,
2009). As evidence of the popularity of OR/MS in DOM, it is noted
that the INFORMS 2010 conference held 16 tracksaccounting for
a total of 61 presentationsrelated to DOM. The signicance of this
number of tracks dedicated to DM is evident when one compares it
to tracks devoted o traditional research areas. As an example, 19
tracks were dedicated to quality related issues, which is a much
more traditional eld in OR/MS. Another sign of continued activity
are the two special issues produced by the journal Socio-Economic
Planning Sciences on Disaster Planning and Logistics. Both of these
special issues appeared in 2012.
In this paper, we present a literature survey of recent OR/MS re-
search in DOM. Our study covers the timeframe 20052010 and
comprises of 155 papers. The inspiration for our work came from
Altay and Green (2006) who offered a survey of OR/MS papers ap-
batta@buffalo.edu plied to DOM, published between 1980 and 2004. Their study pro-

0377-2217/$ - see front matter 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ejor.2013.01.039
202 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
vides a general picture of the state of the art of OR/MS research in
DOM. In their paper, the authors offer a detailed descriptive anal-
ysis of the papers in their survey, based on a classication scheme.
Altay and Green (2006) revealed several important gaps in the ob-
served literature and suggested valuable future research direc-
tions. We started the effort towards this paper six years since the
period covered by Altay and Green (2006). We believe that the
popularity of OR/MS in DOM in recent years warrants the need
to evaluate if the tendencies identied by Altay and Green (2006)
still persist, and even more relevant, to what extent the gaps iden-
tied by them has been addressed in the recent literature. The key
question here is: are OR/MS researchers developing studies to cope
identied gaps in DOM? Or on the contrary, are they ignoring such
gaps when dening their research subjects?
Furthermore, we identify new gaps, tendencies, challenges and
opportunities that are relevant for OR/MS, in order to derive
thoughts about the most appropriate future research directions.
We believe that by having a clearer and unied picture of the past
and present of OR/MS research in DOM, as well as of its most cru-
cial needs, OR researchers will be able to conduct related future re-
search in a more effective way. This is especially relevant for a eld
like DOM due to the magnitude and nature of implications due to
disasters.
Our analysis is performed by classifying papers following the
classication scheme proposed by Altay and Green (2006),
which is based on six dimensions: authors afliation, disaster
type, solution methodology, operational stage, research contri-
bution, and a nal dimension based on a previous classication
framework designed by Denizel et al. (2003). In addition to per-
forming this six dimensional classication, we present an anal-
ysis about the most common assumptions in the eld. To our
knowledge, such an analysis of assumptions has not been per-
formed before.
Since the year 2006, there have been some recent review papers
in DOM; however, none of them can be considered as an update of
the work given by Altay and Green (2006). Lettieri et al. (2009)
developed a systematic review of disaster management that covers
the period between 1980 and 2006, in which the authors dene the
state of the art of disaster management in general (and not partic-
ularly from the point of view of OR/MS). Wright et al. (2006) offer a
survey on OR research in disaster management whose scope is
dedicated to issues related to the Department of Homeland Secu-
rity (DHS). There is not any specic statement in the paper about
the period surveyed, but the references go from 1969 to 2005. Nat-
arajarathinam et al. (2009) offer a literature review on papers pub-
lished before 2008 about supply chain management (SCM) in times
of crisis. The scope of Natarajarathinam et al. (2009) is not limited
to crisis caused by the occurrence of disaster; it also includes nan-
cial and managerial crises. Finally, in Simpson and Hancock (2009),
the authors cover the period 19652007 with the purpose of ana-
lyzing the evolution of OR in emergency management. This latter
paper includes every type of emergencies, and not only those re-
lated to disasters. Comparing our work to the papers discussed
above, we conclude that our main contribution is the evaluation
of the evolution of OR/MS in DOM, by contrasting the reports from
Altay and Green (2006) with our ndings from the recent litera-
ture. The value of this contribution is that it measures recent pro-
gress of the eld and highlight future needs.
The remainder of the paper is organized as follows: Section 2
discusses the search process and the boundaries for our survey.
Section 3 presents an analysis of the papers in our collection based
on the classication framework proposed in Altay and Green
(2006). Section 4 offers future research directions and discusses
the evolution of recent OR/MS research in DOM in the context of
the gaps proposed in Altay and Green (2006). Section 5 presents
our remarks and conclusions. The list of the references for all the
papers gathered in our survey is available in the online comple-
mentary section for this paper.
2. Search methodology and scope of the study
In this section we discuss the search methodology and the
boundaries of our survey. Our study focuses on published journal
papers that exhibit the application of OR/MS in DOM. The dat-
abases used in our search are: ISIs Web of Science, Business Source
Complete, Compendex Engineering Village 2, Scirus, Emerald, Jstor, Sci-
tation and Google Scholar. The keyword disaster was searched in any
place of the documents corresponding to journal articles published
in English. Conferences proceedings, book chapters, books, working
papers, and theses were not included in our study. We limited the
period of our search to 20052010, which corresponds to the sub-
sequent years to the period considered in Altay and Green (2006).
The exploration was limited to engineering and business manage-
ment topics. Some databases provided more tools to dene the
boundaries of our search, as in the case of ISIs Web of Science for
which we were able to choose the topic of disaster management
to perform the exploration.
The scope of our survey, as that of Altay and Green (2006) is OR/
MS research in DOM. Therefore, the boundaries of our survey de-
pend on the denitions of OR/MS and of DOM. In Altay and Green
(2006) OR/MS is dened as a scientic approach to support deci-
sion making in complex systems. On the other hand, DOM is de-
ned as the set of activities performed before, during and after a
disaster in order to reduce its impact on the economy and human
casualties, and returning the community to its normal functioning.
Note that the denition of DOM requires us to establish the deni-
tion of disaster. According to the IFRC a disaster is (. . .) a sudden,
calamitous event that seriously disrupts the functioning of a com-
munity or society and causes human, material, and economic or
environmental losses that exceed the communitys or societys
ability to cope using its own resources (from http://www.
ifrc.org/en/what-we-do/disaster-management/about-disasters/
what-is-a-disaster/). We agree with this denition except for the
use of the term sudden which can be understood as if only ra-
pid-onset events can be classied as disasters, leaving aside known
disasters such as epidemics, droughts and complex humanitarian
emergencies. Instead we would prefer the term shocking to re-
ect the emotional component that disasters cause in society.
We complement the denition of disaster from the IFRC with
thoughts from Altay and Green (2006) and with a denition offered
by Salkowe and Chakraborty (2009). In Altay and Green (2006) the
term emergency response is used to refer to catastrophic events,
while excluding daily emergencies and situations that may be typ-
ically managed by a governmental agency through standard proce-
dures. On the other hand, Salkowe and Chakraborty (2009) offer a
denition of disaster from a governmental perspective, according
to which a disaster declaration request occurs when the local or
state government indicates that it has been overwhelmed by the
effects of the disasters. Based on the denitions presented above,
we dene a disaster as a shocking event that seriously disrupt
the functioning of a community or society, by causing human,
material, economic or environmental damage that cannot be han-
dled by local agencies through standard procedures. This deni-
tion of disaster established the scope of our survey.
Having dened the boundaries of our survey, we proceeded to
verify which of the papers identied t into our scope. Our screen-
ing process can be divided into two levels: initially there were re-
sults that could be rapidly eliminated by inspecting the titles of the
papers and their abstracts, since they provided clear evidence that
such papers were not related to OR/MS and/or DOM. We used a
second and nal lter for inspecting papers that passed the rst
 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
screening test. To do this we read the introduction of the paper, as
well as the problem description to determine whether the paper
would be included in our list. Then, we performed a forward refer-
ence search using the papers that had been already selected. We
used the two-lters approach to screen those papers obtained from
the forward reference search.
Unfortunately, some subjectivity is inevitably involved in the
manual scrutiny of the papers, as pointed out in previous reviews
such as in Altay and Green (2006) and in Natarajarathinam et al.
(2009). In our case, the subjectivity arises in cases where it is not
straightforward to determine whether the subject of the paper
should be classied as a disaster. For instance, should an airplane
crash be considered as a disaster? What about a re in a factory?
What about a nancial crisis in a given company? In these situa-
tions we stuck to the denition of disaster previously discussed,
by asking the following questions: (1) Does the event seriously dis-
rupt the functioning of a community? (2) Does the event exceed
the community resources to cope with its consequences? (3) Are
several non-local agencies required to get involved through a
non-daily or non-standard procedure? According to these ques-
tions, the airplane crash would not qualify as a disaster, since it
may be managed by local agencies through standard procedures.
However, if the crash is caused by a terrorist attack, then it would
be considered as a disaster since it would require actions from
multi agencies involving the federal government. Even the ques-
tions stated above leave room for subjectivity since, for instance,
it is not clear what a serious disruption is. These subjectivities
may be eliminated if there were quantitative criteria for dening
an event as a disaster, such as the number of deaths or amount
of economic damage caused. Unfortunately, to our knowledge,
such categorization has not been yet done.
3. Characteristics of the articles and comparative analysis with
respect to Altay and Green (2006)
A total of 155 papers were nally included in our survey. This
collection of papers is by no means an exhaustive bibliography of
OR/MS research in DOM within the period covered by our study,
since it only includes journal papers obtained from the databases
mentioned before. However, it may be considered as a representa-
tive sample of recent literature. Fifty-one of the papers belong to
OR/MS related journals and 37 of these 51 correspond to main
stream OR/MS related journals: Journal of the Operational Research
Society (JORS), European Journal of Operational Research (EJOR), Com-
puters & OR, Management Science (MS), Interfaces, Annals of OR,
Mathematics of OR, OR Spectrum, and Omega. Fig. 1 shows the num-
Fig. 1. Number of DOM articles in main stream OR/MS outlets.
203
ber of papers found in these journals along with related results
from Altay and Green (2006).
From Fig. 1 it is clear that JORS and EJOR dominate the propor-
tion of recent OR/MS publications in DOM. These two journals hap-
pened to be also the most frequently used according to ndings
from Altay and Green (2006). There are in total three journals pres-
ent in Altay and Green (2006) that do not feature in our review:
Operations Research, Operations Research Letters, and RAIRO-Opera-
tions Research. On the other hand, Computers & OR shows an incre-
ment in its participation with respect to Altay and Green (2006),
which may be explained in part by the special issue on disaster
recovery released by this journal in 2009. Other journals that have
also published special issues related to DOM include International
Journal of Production Economics (IJPE) in 2010, Risk Analysis in
2007, International Journal of Services Sciences in 2008 and 2009,
and International Journal of Physical Distribution and Logistics Man-
agement in 2010.
Among the 104 papers that do not belong to OR/MS related
journals, 80 correspond to engineering related journals. The engi-
neering related journals are in the areas of transportation, logistics,
and production. The participation of these types of journals in our
list turns out to be signicant. In fact, the journal with the highest
frequency in our list is Transportation Research Record (TRR). The
articles from this journal included in our list are mainly referred
to evacuation procedures as in Liu et al. (2006), and Tuydes and
Ziliaskopoulos (2006). Finally, 24 out of 155 papers belong to
non-engineering journals such as Advances in Geosciences, Disasters
and The Professional Geographer.
Table 1 offers a comparison between the statistics from our re-
view and those reported by Altay and Green (2006). In the follow-
ing subsections we will discuss each of the categories listed in
Table 1 by carefully comparing our ndings to those reported by
Altay and Green (2006). Additionally to the discussion on these
categories, we also offer an analysis about the appropriateness of
the assumptions most commonly made in the recent literature.
3.1. Authors afliation
From Table 1 we note that the trends related to authors aflia-
tion observed in Altay and Green (2006) seem to continue: most of
the papers come from USA, followed by other nations in second
place, and international afliations in the third. Papers with
authors afliations different from USA are mostly from Europe
and Asia, whereas only about 1% had an afliation to Latin Ameri-
can countries and Oceania. We agree with Altay and Green (2006)
in that DOM research can benet from international collaboration
among authors. For instance, researchers from third-world coun-
204 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211

Table 1 An interesting nding is that about 11% of the papers use a com-
Summary of our Statistics and Comparison to those from Altay and Green (2006). bination of analytical techniques such as in Afshar et al. (2009)
Number of articles Altay and Green (2006) Our research which use three methodologies: a statistical approach to estimate
109% 155% risk; a Monte Carlo simulation for verication purposes; and math
programming for their optimization model. Simulation and statis-
Authors nationality
USA 43.1 52.9
tics seem to be the techniques most frequently combined with oth-
Other nations 42.2 28.4 ers. In the case of simulation this may be explained by its
International 14.7 18.7 usefulness for validation purposes, and in the case of statistics,
Methodology by its convenience for modeling uncertainty, both issues very com-
Math Programming 32.1 23.1 mon in OR/MS studies. Nevertheless, suitable methodologies for
Probability and Statistics 19.2 6.4 DOM such as fuzzy sets, systems dynamics and soft OR, have even
Simulation 11.9 9.0
a lower participation in recent research than in the review per-
Decision Theory and MAUT 10.1 9.0
Queuing Theory 9.2 0.6 formed by Altay and Green (2006). These methodologies can be
Fuzzy Sets 5.5 1.9 useful for obtaining insights about the problems (Simpson and
Stochastic Programing 3.7 9.6 Hancock, 2009) and for modeling the complex variables and pro-
Experts Systems and AI 3.7 3.8
cesses involved in disaster problems.
Systems Dynamics 1.8 1.3
Constraint Programing 0.9 0.6
Soft OR 0.9 1.3 3.3. Disaster operations life cycle
Conceptual Analysis NA 16.0
Network opt. NA 4.5 Traditionally, there are four stages dened for DOM: mitigation,
Game Theory NA 1.3
preparedness, response and recovery (McLoughlin, 1985). Mitiga-
Combined Meth. NA 11.6
tion refers to the performance of activities in order to reduce the
Operational Stage
long-term risk of a disaster or to diminish its potential conse-
Mitigation 44.0 23.9
Preparedness 21.1 28.4 quences. The preparedness stage involves all those activities per-
Response 23.9 33.5 formed before the disaster occurs in order to accomplish a more
Recovery 11.0 3.2 efcient response. Response stage involves those activities per-
Multiple stages 0.0 11.0
formed upon the occurrence of the disaster and whose objectives
Disaster type are related to the deployment of vital resources to serve the af-
Natural 28.4 20.0
fected population. And nally, recovery refers to short-term and
Man-made 33.1 5.8
Humanitarian 0.9 2.6
long-term activities performed in order to restitute the normal
All disasters 37.6 71.6 functioning of the community. Further information about the
Research contribution
stages of DOM may be found in Coppola (2011) who offers a de-
Theory 26.6 19.3 tailed explanation of the subject.
Model 57.8 75.5 Under the category of multiple we classify those papers address-
Application 15.6 5.2 ing more than one DOM stage, such as Fiorucci et al. (2005) who
Denizel et al. classication propose a two-stage model where the rst stage is related to the
MS1 19.2 34.8 strategic location of air bases in charge of re emergencies (pre-
MS2 33.9 5.2
paredness), and the second stage refers to the dynamic re-alloca-
ME1 15.6 22.6
ME2 3.7 16.8 tion of resources once an event has occurred (response).
MC1 8.3 12.9 Conceptual papers offering reections about general issues in
MC2 19.3 7.7 DOM are also included in the category multiple.
Our study suggests a shift on the trends related to mitigation re-
search, since according to Altay and Green (2006) this stage was
the most widely studied, whereas it features third in our list. Re-
tries can share their perspective and knowledge about the actual sponse and preparedness appear respectively in the rst and sec-
conditions present in their countries (which happens to be the ond places. On the other hand, our ndings agree with Altay and
ones most susceptible to disasters damages), whereas researchers Green (2006) and other previous review papers in that there is still
from more prosperous countries can share their access to valuable a lack of research in recovery stage (Simpson and Hancock, 2009;
technological resources. Wright et al., 2006; and Natarajarathinam et al., 2009).

3.2. Methodology 3.4. Disaster type

Table 1 shows that mathematical programming is still the most
preferred methodology. Simulation, and Decision Theory and Multi
Attribute Utility Theory seem to maintain their participation as
some of the methodologies most widely used. On the other hand,
Probability and Statistics, and Queuing Theory appear with a rela-
tive lower frequency in our study compared to the results from Al-
tay and Green (2006). This contrasts to the methodology of
Stochastic Programming, which signicantly increased its partici-
pation in the eld. We added a few methods not included in the pa-
per by Altay and Green (2006). Among these, conceptual works are
the most popular ones. In these, the authors offer reections on a
given eld of DOM without applying any analytical technique, as
in Kovcs and Spens (2007) who offer an analysis for a better
understanding of logistic activities under disaster settings.
Our ndings regarding types of disaster show a shift in the ten-
dencies observed in Altay and Green (2006) who found that man-
made disasters were more popular than natural ones. This shift
may be caused by a series of overwhelming natural disasters that
have occurred since 2004, which have captured the interest of pub-
lic audience (such as the tsunami in Indonesia in 2004, hurricane
Katrina in 2005, and the Haiti earthquake in 2010).
We are glad to have found that OR/MS researchers are adopting
an all-hazard approach, as evidenced in Table 1. An important re-
mark is that some of the research developed for general disasters
includes a case study where some specic type of disaster is con-
sidered. For example Yi and zdamar (2007) develop a model for
logistic coordination and evacuation for any general disaster and
they illustrate the application of their model through a case study
 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
that involves an earthquake in Istanbul. We believe that this type
of study is of great value since they adopt an all-hazard approach
while illustrating the applicability of the research for particular
situations.
3.5. Research contribution
Articles may be classied according to the type of contribution
that they propose. Following the classication scheme proposed by
Altay and Green (2006), the three categories for this dimension are
theory, model and application. Articles classied in the rst category
correspond to those that present reections about a particular area,
test hypothesis, describe the behavior of a system, provide a frame-
work, or dene principles or taxonomies. Review and survey pa-
pers, like the one hereby presented, would fall within this
category. The second category groups articles that develop analyt-
ical models, whereas the third category refers to articles whose
outcomes consist in the development of some product such as
computer software or devices. Research papers that describe the
development of decision support systems (DSS) would fall within
this last category.
When ranking categories according to their frequency, we ob-
tain similar results to those observed by Altay and Green (2006).
However, according to our ndings, the proportion of application
studies has signicantly decreased, whereas the opposite have oc-
curred with model papers. The participation of theory articles has
diminished but not as signicantly as that of application papers.
In Fig. 2 we present our ndings regarding the research type
according to DOM stages. The gure also shows the corresponding
ndings from Altay and Green (2006). For all of the stages of DOM,
models are by far the most common type of research contribution.
Theory would be in second place, whereas application is the least
frequently observed. On the other hand, in the case of articles re-
ferred to multiple stages, theory is the most common type of con-
tribution. This may be explained by the fact that many of these
articles present abstractions about general aspects of DOM, which
involves all of its stages. Even though Altay and Green (2006) also
found models as the most common type of contribution, in their
case they found a considerable proportion of application papers
for the response stage, and of theoretical articles for mitigation,
which differs from our ndings.
Although the development of models is valuable in DOM, re-
search in the other two categories, i.e. theory and application,
should not be disregarded. Theory is relevant for a better under-
standing of DOM problems and as a base for developing model re-
search. On the other hand, the importance of application studies is
that they provide a tool for taking theoretical and analytical re-
search into practice. DOM requires tools that support the process
Fig. 2. Distribution of research contribution to stages of DOM.
205
of decision making in a quick and efcient way. Therefore, more re-
search in Decisions Support Systems (DSS) and other applications
would be of value.
3.6. Denizel et al. (2003) classication
An interesting dimension considered by Altay and Green (2006)
is based on a classication framework given by Denizel et al.
(2003), which is in turn based on Corbett and Van Wassenhove
(1993). In Corbett and Van Wassenhove (1993), the authors pro-
vide a classication framework that categorizes research into three
main groups: management science (MS), which contains papers
whose objective is to contribute to the body of knowledge of a gi-
ven research area; management consulting (MC), which covers re-
search where a practical problem is solved by standard methods;
and management engineering (ME), which refers to research that
uses existing methods in a fundamentally novel way to solve prac-
tical problems. Denizel et al. (2003), expands the classication
from Corbett and Van Wassenhove (1993), creating six categories:
MS1, MS2, MC1, MC2, ME1, and ME2 (see Fig. 3). The attributes
considered by Denizel et al. (2003) are: problem setting (real,
hypothetical or none), source of data (real, random or no data), sit-
uation (novel or widely studied in OR), approach (novel or widely
studied in OR), results (specic or general), and further research
implications (existent or inexistent). We refer the reader to Denizel
et al. (2003) for further information.
The process used to apply the Denizel et al. (2003) classication
scheme is as follows: (i) settings and data are considered real only
if the authors make an explicit mention about it; (ii) the novelty of
the situation and approach is dened based on the literature re-
view offered by each article; (iii) results are considered specic if
the authors develop a study for a specic setting, for example a
particular simulation model, and they do not explicitly mention
the possible extension of their outcomes to other scenarios; and
(iv) further research implications are considered existent only if
they are explicitly mentioned by the authors. Review and survey
papers are treated as special cases and their classication is per-
formed following the procedure proposed by Denizel et al.
(2003). According to this, if the paper reviews the state of the art
of a particular area, it is coded as MS2. If it summarizes and states
the relevant issues from previous work, it is classied as MC2. Fi-
nally if the review paper proposes future research implications
based on its observations, it is categorized as ME2.
The most frequent category observed in Altay and Green (2006)
is MS2, followed by MS1, MC2, ME1, MC1, and ME2 in that order. In
our case, MS1 turns to be the most popular category, whereas MS2
(which groups articles that use no data) went from the rst to the
last position in the list. In fact, we are pleased to have found that
60% of the papers in our collection use real data for their case stud-
ies, whereas 35% used random data. Altay and Green (2006) had
pointed out their concern about the lack of ME2 articles in OR re-
search applied to DOM. We have gladly found that in our review
ME2 went from the last place to the second in the list. This result,
combined to the fact that ME1 ranked in third place in our review,
shows current trends of adapting existing OR/MS tools or using
them in a novel way to solve real problems in DOM.
In Fig. 4, we present our ndings regarding the research type
according to DOM stages. The gure also shows the corresponding
ndings from Altay and Green (2006). The articles related to MC
have a relatively low participation for all the stages of DOM. On
the other hand, ME has a strong participation at every stage, except
for the recovery stage which shows a larger contribution from both
MC and MS. Finally, articles where multiple stages are considered
are mostly classied as ME, with an approximately equal participa-
tion of the other two types of studies. These ndings are quite dif-
ferent from Altay and Green (2006) who found a large proportion
206 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
Fig. 3. Classication Scheme for Research Type. Source: Denizel et al. (2003).
Fig. 4. Distribution of Denizel et al. (2003) classication to stages of DOM.
of MS articles for every stage of DOM, except for the preparedness
stage, which had a greater proportion of MC works. ME had the
smallest proportion of papers in all of the stages. This implies that
for every stage of DOM, research is not as focused as before in just
developing results to contribute to the body of knowledge of the
eld, but it is becoming more focused on the applicability of such
results, and on exploring novel problems and methods to solve
practical problems.
3.7. Research assumptions
In Table 2 we present a list of the most frequent assumptions
contained within the papers that were gathered. We evaluated
the appropriateness of each assumption considering its applica-
bility to DOM. According to our ndings, assumptions may be
classied into three categories: reasonable, limited, and unrealis-
tic. Reasonable assumptions are those that do not compromise
the applicability of the study, even though they might leave
room for improvement. The assumption that a set of candidate
locations is given for locationallocation problems would fall in
this category. Limited assumptions include those that apply to
some problems, but that are impractical to others. Therefore,
researchers should carefully evaluate the appropriateness of such
assumptions for every particular case. For instance, statistical
independence is generally assumed in DOM models to avoid
the complications arising from considering statistical dependence
in its place. However, the suitability of this assumption is some-
times questionable, as in Lodree and Taskin (2009) who assumed
sequential observations of the wind speed for a given hurricane
as independent events. Finally, unrealistic assumptions refer to
those that are not actually suitable for general disaster settings
and that compromise the applicability of the research, e.g.
assuming known deterministic post-disaster demand values.
Unrealistic assumptions help to reduce the complexity of re-
search, especially when developing new approaches or facing
problems that have not been widely studied. However, since
they seriously jeopardize the usefulness of the study, OR/MS
researchers should nd ways to substitute them by more reason-
able assumptions. The second column of Table 2 shows the clas-
sication group for each assumption in our list.
4. Future research directions
For dening our future research directions, we focused on what
we consider as the main components for addressing a DOM prob-
lem. Based reading of the papers in our survey, these components
can be dened as: actors, technologies and the DOM problem itself.
The latter component can be dened as a set of steps: (1) identify-
ing and dening the problem, (2) building a set of assumptions, (3)
 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
Table 2
Research assumptions.
ID Assumption
1 Immediate availability of supplies and other resources after the
occurrence of the disaster (Grmez et al., 2011; Zaric et al., 2008)
2 Deterministic and given post-disaster travel times, costs and demand. (a
few papers assume statistical distribution for demand) (Fiorucci et al.,
2005; Minciardi et al., 2009). If scenarios are used, these parameters are
assumed deterministic and given within each scenario (Beamon and
Kotleba, 2006; Yi and Kumar, 2007)
3 A given set of scenarios whose probabilities and behavior are based on
experts opinions and historic data (Chang et al., 2007; Rawls and
Turnquist, 2010)
4 Static parameters (demand, travel time, costs, etc.) (a few consider
dynamic parameters) (Chang et al., 2007; Nagurney et al., 2011)
5 Resistant buildings and network infrastructures (Stepanov and Smith,
2009; Sbayti and Mahmassani, 2006)
6 For distribution of relief goods and humanitarian supply chains, total
resource availability generally considered enough to cope total demand
(otherwise, a penalty for unmet demand is considered as an input)
(Nagurney et al., 2011; Rawls and Turnquist, 2010)
7 Network topology is given for transportation problems (Chiu and
Mirchandani, 2008; Opasanon and Miller-Hooks, 2009)
8 Set of candidate locations considered as an input for location-allocation
problems (Rawls and Turnquist (2010), Jia et al., 2007)
9 Statistical independence of events (Taskin and Lodree, 2010; Matisziw
and Murray, 2009)
10 Information about disaster impact (demand, network status, etc.)
available immediately after the disaster (Georgiadou et al., 2007; Chen
and Chou, 2009)
11 Perfect information of evacuees about road network and trafc
conditions (Kongsomsaksakul et al., 2005; Chen and Zhan, 2008)
12 No time-windows delimiting delivery times in relief goods distribution
(Rawls and Turnquist, 2010; Campbell et al., 2008)
13 Population divided into categories according to their urgency of
attention (Wang et al., 2010; Childers et al., 2009)
14 Post-disaster coverage models based on pre-disaster computed
distances regardless infrastructure vulnerability (Paul and Batta, 2008;
Grmez et al., 2011)
15 Evacuation area divided into zones, where each zone has a given
destination (Chen and Zhan, 2008; Abdelgawad and Abdulhai, 2010)
16 In staged evacuations, evacuees from the same area evacuate at the same
time (Chen and Zhan, 2008; Chiu et al., 2008)
17 Poisson distribution for describing the process of disaster occurrence
through time (Bogen et al. (2007), Doerner et al., 2009)
18 Capacitated roads or links (Xie and Turnquist, 2009; Chiu and Zheng,
2007)
19 Evacuation guidance performed ideally (Chiu et al., 2007; Sbayti and
Mahmassani, 2006)
20 Evacuees free-will for selecting which shelter to evacuate and which
route to take (Chiu and Zheng, 2007; Zou et al., 2005)
selecting a methodology for solving the problem, and (4) formaliz-
ing the solution through its application. We took into consideration
these four components of DOM when evaluating the papers in our
list in order to identify gaps and related future research directions.
Note that the categories in the classication scheme discussed so
207
Type Comments on classication
Limited For unpredictable disaster, this assumption is not reasonable. Also, it
would be of value to take into consideration restrictions concerning
production lead times and transportation delays when planning the
distribution of relief supplies
Unrealistic Several factors involved in a typical disaster setting introduce
uncertainty into parameters such as demand, costs, and travel times.
Therefore, it is important to model the uncertainty of such parameters.
The use of scenarios might help, but some uncertainty might need to be
considered within each scenario, as well
Limited For predictable disasters it would be more reasonable to consider
disaster characteristics, e.g. for hurricanes consider path and wind speed
as in Taskin and Lodree (2010). When the information about scenarios
characteristics is based on historical data, an important consideration is
that the scarcity of previous data and the low frequency of previous
disasters may lead to highly inaccurate predictions
Unrealistic Costs, demand, travel times, etc. may change as disaster relief evolves.
Also, if using penalties for unmet demand, these should be time
dependent
Unrealistic Disaster typically affect infrastructures which may signicantly affect
travel times and costs
Limited Not reasonable for unpredictable disasters, since there is no warning for
gathering all the required resources. Also, it does not account for
possible damage to inventories. In the case in which penalties are used
for quantifying the cost of unmet demand, such penalties need to be
carefully analyzed since unmet demand of vital assets may be a matter
of death or life
Reasonable It is reasonable to assume that there is some basic knowledge regarding
the transportation network in disaster relief, based on the existing
network infrastructure
Reasonable Consider the problem of prepositioning supplies. It would be reasonable
to have a given set of candidate locations, instead of assuming that we
could virtually preposition in any random point
Limited Statistical independence is a strong assumption and it should be veried,
e.g.in Taskin and Lodree (2010), the authors assume that successive
forecasts of hurricanes characteristics are statistical independent, but no
supports are given for such an assumption
Unrealistic During disasters, chaos and communication failures are not uncommon,
which may lead to disruptions and delays regarding the transmission of
the disasters information
Unrealistic In disasters, we would expect that it might take some time to state trafc
conditions. Also, this assumption does not leave room to population
without knowledge about road networks, such as tourists
Unrealistic In disaster, demand (especially of vital items) need to be satised within
a given period. After a certain time, it may be too late to serve demand.
This must be considered in research related to distribution of relief
goods
Reasonable This is a reasonable assumption since it is more realistic to divide the
population according to the urgency of their needs, instead of assuming
that every person has the same level of urgency
Limited It does not account for road and building failures caused by the disaster
Reasonable We could reasonably expect that the destinations for evacuees are
specied with respect to their original location. Therefore, it is
reasonable to assume that each zone has a given destination
Reasonable In evacuation procedures, it is reasonable to assume that evacuation
advisories are issued for all the people in a given area
Limited Assumes independence of events (see assumption 9)
Reasonable It is realistic to assume that roads or links do not have innite capacity
Unrealistic It does not leave room for alterations due to driver behaviors and other
contingencies
Limited Limitations due to closed roads and ofcial evacuation plan might apply
far, (with the exception of authors nationality) can be linked
to the process of solving DOM problems. In our analysis, we also
used the gaps and research directions suggested in Altay and Green
(2006) to evaluate to what extent they had been addressed in re-
cent literature.
208 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
4.1. Actors
One of the research directions suggested in Altay and Green
(2006) is the development of studies that address the denition
of organizational and network structures, in order to facilitate
communication and coordination among the different agencies in-
volved in the resolution of a disaster. In our survey, we found some
very interesting articles such as Balcik et al. (2010) who review the
challenges for coordination in humanitarian supply chain and eval-
uate the adaptation of coordination mechanism of commercial
supply chains to humanitarian ones. Other related papers are the
ones proposed by Schulz and Blecken (2010), Gonzalez (2010),
and Adivar and Mert (2010). But this is a very small proportion
of the articles reviewed and the fact is that this topic is still in its
infancy in OR/MS research. An interesting analysis on multi-agency
coordination is offered by Christopher and Tatham (2011), who
discuss the evolution of the initiative United Nations (UN) Cluster
Approach (CA). The CA was born in order to improve multi-agency
coordination and building system capacity. The basic idea behind
the CA is to create clusters of organizations where each cluster
has a leader agency. Even though the CA has found important chal-
lenges and there is still much work to do, it seems that it is on the
right track for coordination in humanitarian response (Christopher
and Tatham, 2011). We believe it would be of value for OR/MS
researchers to incorporate coordination initiatives such as the CA
into their models. This implies considering additional assumptions
that allow modeling coordinated operations, such as sharing
capacity and information. OR/MS research can be conducted with
the goal of addressing logistic challenges in coordinating initiatives
such as the CA.
We would like to point out the lack of research developed in
conjunction with humanitarian organizations. This is an evident
gap between OR/MS practitioners and people in charge of perform-
ing relief activities. Even though our survey includes a number of
papers with cases studies, most of them were not conducted in
conjunction with humanitarian organizations. We need to get clo-
ser to humanitarian agencies in order to know their need, the most
urgent problems from their perspective, and the actual budget con-
straints, among other issues, in order to formulate models that use
realistic assumptions and conduct more applicable research.
4.2. Technology
Regarding technologies, in Altay and Green (2006), the authors
state that there is a gap between what we need and what we have
available. They also state the need for developing new technologies.
During the Health and Humanitarian Logistics Conference, 2011, Jar-
rod Goentzel from the MIT Center for Transportation & Logistics
mentioned that the needs of technology in DOM involve, among oth-
ers: (i) turning data into demand and supply information, (ii) evalu-
ation of alternatives, and (iii) planning and budgeting
implementations (from http://www.scl.gatech.edu/humlog2011/
program/). The development of such technologies could be ad-
dressed through application studies, which turn out to be very
scarce in our review. New technologies arising from other areas
can be useful in OR/MS research. In this respect, we would like to
highlight a proliferation of work using geographic information sys-
tems (GIS) such as Chang et al. (2007), Saadatseresht et al. (2009),
and Jotshi et al. (2009). GIS can be helpful for developing location
analysis and obtaining more accurate data about spatial elements
(road networks, geographic obstacles, etc.). In fact, the methodology
HAZUS, which was created by Federal Emergency Management
Agency (FEMA) for estimating disaster losses, uses GIS in order to
perform its estimations (from http://www.fema.gov/plan/prevent/
hazus/). However, this tool is only directed at geographical modeling
and more applications are needed for other tasks.
4.3. Solving DOM problems
4.3.1. Identication and denition of DOM problems
As discussed in Section 3.3, there is still a lack of research in
problems related to the recovery phase of DOM, a trend that was
previously observed in Altay and Green (2006) and in other review
papers (Natarajarathinam et al., 2009; Lettieri et al., 2009). The
experience of Haiti underscores the importance of performing ef-
cient recovery efforts; even two years after the 2010 earthquake
this country had not returned to normal functioning. OR/MS re-
search can be applied to model the logistics involved in recovery
activities, such as debris removal (Altay and Green, 2006), recon-
struction of roads and buildings, and sustained attention to af-
fected population.
Altay and Green (2006) also point out the lack of answers of
how ethical factors should be included in modeling service alloca-
tion to disaster victims. There is still lack of work in this respect,
except for a few papers such as Balcik et al. (2008) where equity
is considered in performing last mile relief distribution. Altay and
Green (2006) also point to the lack of research in the areas of Busi-
ness Continuity and Infrastructure Design. In the rst case, the
authors suggest future research addressing innovative nancial
and insurance tools. For the second case, the authors note the need
for a better understanding of connectivity between critical infra-
structures and for ways to incorporate survivability into building
maintenance and design. In our ndings, there were virtually no
articles related to business continuity or infrastructure design. In
the case of business continuity, researchers seem to be more inter-
ested in developing models for helping affected population in gen-
eral and little attention is paid to business disruptions. To the
challenges proposed by Altay and Green (2006) we would like to
add the following: How would a disruption of one of the organiza-
tions in a supply chain affect its partners? For organizations in
charge of producing vital goods, what are the timeframes allowed
for recovery considering the impact to the community? For the
case of infrastructure design, we would like to highlight the need
for research in order to better understand the infrastructure design
of hazardous facilities. This should consider the impact that a fail-
ure in such infrastructures would have in the surrounding commu-
nity and beyond.
Business Continuity and Infrastructure Design are not the only
areas that show lack of research by OR/MS researchers in DOM.
In general, OR/MS research in DOM is mostly concentrated in few
areas such as location-allocation, humanitarian logistics and sup-
ply chain management, evacuation, and risk assessment, whereas
other areas that are relevant to DOM and that could highly benet
from OR/MS research have been understudied. Another relevant
understudied problem in DOM is related to searching for an entity
(such as an injured person) after the occurrence of a disaster.
Improving the efciency of entity search could help to save lives
or, at least, to decrease the time during which a person requiring
attention remains unattended. Research in entity search can be ap-
plied to diverse types of disasters that involve damages to infra-
structures, such as hurricanes, oods, and terrorist attacks among
others. However, only two papers in our collection address the
problem of entity search (Jotshi and Batta, 2008; Jotshi and Batta,
2009). This subject could considerably benet from OR/MS re-
search through the use of techniques such as simulation, network
optimization, and probabilistic analysis. Since both papers deal
only with immobile entities, a possible extension of the problem
can include searching for entities with limited motion capabilities.
Another example of an important DOM area that could benet
from OR/MS research is post-disaster housing. According to El-An-
war et al. (2010a) there is a lack of studies that minimize the socio-
economic disruption of displaced families while minimizing the
associated costs for temporary post-disaster housing. An efcient
 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
process of post-disaster housing can help the community to re-
cover sooner from the shocking impact that a disaster can cause.
Some of the specic problems involved in post-disaster housing,
include delays on the deployment of resources, and location of
houses considering natural and man-made hazards (El-Anwar
et al., 2010a). OR/MS techniques, such as location analysis, risk
assessment and resources scheduling, can be used to cope with
these problems. Despite the suitability of OR/MS research for deal-
ing with these issues, in our survey we identify only two papers
focusing on the subject of post-disaster housing (El-Anwar et al.,
2010a; El-Anwar et al., 2010b).
It is important to realize that when a disaster occurs there are
numerous problems that must be solved in order to reestablish
the normal functioning of the community and to attend to the af-
fected population. Such problems are not likely to be isolated but
interrelated. Our professional duty is to use OR/MS research in
the most efcient way to continue our work on well-studied prob-
lems in DOM while we explore the understudied aspects of it. Our
ultimate task is to address and connect every piece of DOM from
the OR/MS perspective in order to improve DOM as a whole.
4.3.2. Validity of assumptions
As discussed in Section 3.7, some of the assumptions made in
OR/MS models in DOM are limited or unrealistic. The comments
from Altay and Green (2006) with respect to the validity of
assumptions relate to a claim for a better understanding of the in-
puts of DOM models, especially in the case of specic events. In
this respect, our ndings show that researchers often adopt unre-
alistic assumptions about the inputs for such models, as suggested
by Table 2, e.g. deterministic data and statistical independence. Lit-
tle efforts have been done to understand the nature and uncertain-
ties of input data or assumed parameters. In this respect, it would
be of value to have papers like that offered by Barker and Haimes
(2009), in which the authors develop a framework for analyzing
the sensitivity of extreme event consequences to uncertainties on
the parameters used on underlying distributions. Unlike daily
emergencies and business operations, the frequency of disaster
occurrences is relatively low. Therefore, the availability of reliable
input data is an important challenge for DOM researchers, as stated
in Albores and Shaw (2008). There is a lack of robust models that
examine the impact of data uncertainty on the estimation of disas-
ter impacts, such as in Barker and Haimes (2009).
One of the most important challenges detected from our survey
is the necessity of conducting formal analysis (with solid statistical
basis) that can help build realistic assumptions about stochastic
behavior of problems in DOM. For instance, research addressing
the location of facilities for coverage models does not commonly
consider the potential damage to roads. In papers such as Paul
and Batta (2008) and Jia et al. (2007) a given affected site is as-
sumed covered if a response facility is located within a certain dis-
tance, where such a distance is computed based on pre-disaster
conditions. It would be of value to incorporate coverage probabil-
ities based on infrastructure reliabilities.
For dealing with stochastic behavior, commonly authors use a
scenario-based approach; e.g. see the papers by Rawls and Turn-
quist (2010), Chang et al. (2007), and Snediker et al. (2008). A ma-
jor drawback of scenarios is that they rarely comprehend all
possible outcomes. Moreover, researchers generally assume the
set of scenarios as a given input and they do not typically offer
an efcient, systematic, and reliable way to dene them. A more
appropriate path would be to study the potential outcomes of a
disaster through a solid probabilistic analysis. If possible, such an
analysis should account for observable disaster characteristics.
For instance, the path and wind forces of hurricanes can be used
to estimate the potential impact of a predicted hurricane, as in Lod-
ree and Taskin (2009). When performing the probabilistic study, it
209
is important to evaluate the statistical dependence among events,
which is generally disregarded without a proper analysis that sup-
ports such an assumption, as mentioned in Section 3.7.
Probability and statistics can also be used to overcome the com-
mon assumption that all the required information is available
whenever required. This approach does not take advantage of the
benets from additional information as the disaster evolves. In this
respect, there is a lack of research that accounts for information
updates and dynamic parameters. A Bayesian approach can be suit-
able for incorporating additional observations such as in Lodree
and Taskin (2009).
Finally, we would like to note the lack of papers in the recent
OR/MS literature that focus on specic conditions in developing
countries. According to reports from the British network BBC (from
http://www.bbc.co.uk/news/science-environment-15592761),
among the 2.7 billion of people that resulted affected from natural
disasters in the last decade, only 1% corresponded to people living
in richer countries. Some of the special challenges for DOM in
developing countries include poor conditions of road networks
(limited coverage, high vulnerability, and restricted links capacity);
internal conicts; and limited technological resources. Addition-
ally, it should be expected that the social behavior of the popula-
tion differs from that found in developed countries, which
impose an additional consideration, as suggested by Peacock
et al. (1997) who discuss how ethical and socioeconomic contexts
may affect the response of population to evacuation warnings.
4.3.3. Methodologies
Regarding the methodologies used to solve DOM problems,
Altay and Green (2006) remark the suitability of the eld for
using soft OR techniques, which is also pointed out by Simpson
and Hancock (2009). The authors further claim that there is a
need for interdisciplinary methodologies and new methods.
Additionally, the authors suggest dening a way for selecting
methods to be used to address specic class of disaster prob-
lems. We found that Soft OR is still an underused method in
OR/MS in DOM, despite its suitability to the eld. Also, new
methodologies which apply techniques and concepts from other
disciplines are still lacking. However, it was interesting to nd
some papers that combine different methodologies in their ap-
proach as evidenced by Afshar et al. (2009). Also, there is de-
nitely still a lack of work to dene rules for deciding what
methods to use in specic situations.
4.3.4. Applying the solution
In Altay and Green (2006), the authors claim for more research
in ME that allow carrying theory into practice. This is perhaps the
only gap from those identied in Altay and Green (2006) that has
shown considerable improvements in the recent literature. It
seems that researchers are interested in narrowing the gap be-
tween theory and application through the development of case
studies. Nevertheless, such case studies are generally simplied
by using limited and unrealistic assumptions as discussed in Sec-
tion 4.3.2. Finally, we would like to remark that there is still a lack
of research in efciency indicators. We found little research in this
subject, even though it has been consistently identied as a gap in
previous review papers (Altay and Green, 2006; Simpson and Han-
cock, 2009; Lettieri et al., 2009).
We must realize that the three components discussed from Sec-
tions 4.2 through 4.4 are intrinsically interrelated. Therefore, when
developing research in any of these components, we should facili-
tate and promote the interconnection of our research with that
belonging to the other components of DOM.
210 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
5. Conclusions and remarks
In this paper we have presented a review in recent OR/MS liter-
ature in DOM. Our work is a continuation of a previous review from
Altay and Green (2006). We present a classication scheme based
on the one used by Altay and Green (2006). Our ndings were com-
pared to those reported by Altay and Green (2006) in order to ana-
lyze how the eld has evolved since their review was published.
We found that most trends have remained, e.g., lack of research
for recovery activities; but others have experienced shifts, e.g., pro-
portion of ME articles in OR/MS research in DOM. Our review con-
cludes that there has not been drastic changes or developments in
the eld of OR/MS in DOM since the publication of Altay and Green
(2006). Our ndings also show that most of the gaps observed by
Altay and Green (2006) still remain. This suggests that there is
not a clear initiative in OR/MS research to explicitly ll such gaps.
We have provided a set of future research directions based on
the gaps that still persist since Altay and Green (2006) and the
new gaps identied in our survey. We believe it would be profes-
sionally benecial for the OR/MS community to focus on lling
these gaps. This will provide a solid basis for OR/MS research in
DOM, and more importantly, allow the eld of OR/MS to make
effective contributions in actual disaster relief efforts. Suggested
future research directions include: (i) improvement of the coordi-
nation among DOM actors; (ii) introduction of new technologies
through more application studies; (iii) study of DOM problems as
a whole by exploring well-studied as well as understudied areas
that can benet from OR/MS, using formal statistical analysis to
establish realistic assumptions in DOM models that reect the sto-
chastic nature of DOM; (iv) in-depth exploration of methodologies
such as Soft OR and interdisciplinary techniques that are suitable
to DOM; and (v) measurement of the effectiveness of adopted
strategies through the use of performance indicators.
We believe that our results give an accurate perspective of re-
cent OR/MS literature in DOM. Despite this assertion, our list of ref-
erences cannot be considered an exhaustive bibliography of the
eld. It is our sincere hope that our ndings and recommendations
will contribute to the improvement of DOM operations and will in-
spire other researchers to develop work in order to ll the gaps
identied.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ejor.2013.01.039.
References
Abdelgawad, H., Abdulhai, B., 2010. Managing large-scale multimodal emergency
evacuations. Journal of Transportation Safety & Security 2 (2), 122151.
Adivar, B., Mert, A., 2010. International disaster relief planning with fuzzy
credibility. Fuzzy Optimization and Decision Making 9 (4), 413433.
Afshar, A., Rasekh, A., Afshar, M.H., 2009. Risk-based optimization of large ood-
diversion systems using genetic algorithms. Engineering Optimization 41 (3),
259273.
Albores, P., Shaw, D.A., 2008. Government preparedness: using simulation to
prepare for a terrorist attack. Computers & Operations Research 35 (6), 1924
1943.
Altay, N., Green, W.G., 2006. OR/MS research in disaster operations management.
European Journal of Operational Research 175 (1), 475493.
Balcik, B., Beamon, B.M., Smilowitz, K., 2008. Last mile distribution in humanitarian
relief. Journal of Intelligent Transportation Systems 12 (2), 5163.
Balcik, B., Beamon, B.M., Krejci, C.C., Muramatsu, K.M., Ramirez, M., 2010.
Coordination in humanitarian relief chains: practices, challenges and
opportunities. International Journal of Production Economics 126 (1), 2234.
Barker, K., Haimes, Y.Y., 2009. Assessing uncertainty in extreme events: applications
to risk-based decision making in interdependent infrastructure sectors.
Reliability Engineering System Safety 94 (4), 819829.
BBC News, n.d. Retrieved from <http://www.bbc.co.uk/news/science-environment-
15592761>.
Beamon, B., Kotleba, S., 2006. Inventory modelling for complex emergencies in
humanitarian relief operations. International Journal of Logistics: Research and
Applications, 118, 9:1.
Bogen, K.T., Jones, E.D., Fischer, L.E., 2007. Hurricane destructive power predictions
based on historical storm and sea surface temperature data. Risk Analysis 27
(6), 14971517.
Campbell, A.M., Vandenbussche, D., Hermann, W., 2008. Routing for relief efforts.
Transportation Science 42 (2), 127145.
Chang, M.-S., Tseng, Y.-L., Chen, J.-W., 2007. A scenario planning approach for the
ood emergency logistics preparation problem under uncertainty.
Transportation Research Part E: Logistics and Transportation Review 43 (6),
737754.
Chen, C.C., Chou, C.-S., 2009. Modeling and performance assessment of a transit-
based evacuation plan within a contraow simulation environment.
Transportation Research Board: Journal of the Transportation Research Board
2091, 4050.
Chen, X., Zhan, F.B., 2008. Agent-based modelling and simulation of urban
evacuation: relative effectiveness of simultaneous and staged evacuation
strategies. Journal of the Operational Research Society 59 (1), 2533.
Childers, A.K., Visagamurthy, G., Taaffe, K., 2009. Prioritizing patients for evacuation
from a health-care facility. Transportation Research Board: Journal of the
Transportation Research Board 2137, 3845.
Chiu, Y.-C., Mirchandani, P.B., 2008. Online behavior-robust feedback information
routing strategy for mass evacuation. IEEE Transactions on Intelligent
Transportation Systems 9 (2), 264274.
Chiu, Y.-C., Zheng, H., 2007. Real-time mobilization decisions for multi-priority
emergency response resources and evacuation groups: model formulation and
solution. Transportation Research Part E 43 (6), 710736.
Chiu, Y.-C., Zheng, H., Villalobos, J., Gautam, B., 2007. Modeling no-notice mass
evacuation using a dynamic trafc ow optimization model. IIE Transactions 39
(1), 8394.
Chiu, Y.-C., Zheng, H., Villalobos, J.A., Peacock, W., Henk, R., 2008. Evaluating
regional contra-ow and phased evacuation strategies for Texas using a large-
scale dynamic trafc simulation and assignment approach. Journal of Homeland
Security and Emergency Management 5 (1) (Article 34).
Christopher, M., Tatham, P., 2011. Humanitarian Logistics: Meeting the Challenge of
Preparing for and Responding to Disasters. Kogan Page, pp. 8569.
Coppola, D.P., 2011. Introduction to International Disaster Management, 2 ed.
Butterworth-Heinemann, 696p.
Corbett, C.J., Van Wassenhove, L.N., 1993. The natural drift: what happened to
operations research? Operations Research 41 (4), 625640.
Denizel, M., Usdiken, B., Tuncalp, D., 2003. Drift or shift? Continuity, change, and
international variation in knowledge production in OR/MS. Operations Research
51 (5), 711720.
Doerner, K.F., Gutjahr, W.J., Nolz, P.C., 2009. Multi-criteria location planning for
public facilities in tsunami-prone coastal areas. OR Spectrum 3 (31), 651678.
El-Anwar, O., El-Rayes, K., Elnashai, A., 2010a. Minimization of socioeconomic
disruption for displaced populations following disasters. Disasters 34 (3), 865
883.
El-Anwar, O., El-Rayes, K., Elnashai, A.S., 2010b. Maximizing the sustainability of
integrated housing recovery efforts. Journal of Construction Engineering and
Management 136 (7), 794802.
Federal Emergency Management Agency, n.d. Retrieved from <http://
www.fema.gov/plan/prevent/hazus/>.
Fiorucci, P., Gaetani, F., Minciardi, R., Trasforini, E., 2005. Natural risk assessment
and decision planning for disaster mitigation. Advances in Geosciences 2, 161
165.
Georgiadou, P.S., Papazoglou, I.A., Kiranoudis, C.T., Markatos, N.C., 2007. Modeling
emergency evacuation for major hazard industrial sites. Reliability Engineering
& System Safety 92 (10), 13881402.
Gonzalez, R.A., 2010. Developing a multi-agent system of a crisis response
organization. Business Process Management Journal 6 (5), 847870.
Grmez, N., Kksalan, M., Salman, F.S., 2011. Locating disaster response facilities in
Istanbul. Journal of the Operational Research Society 62, 114. http://dx.doi.org/
10.1057/jors.2010.67 (Published online in 2010).
Health and Humanitarian Logistic Conference, 2011. Retrieved from <http://
www.scl.gatech. edu/humlog2011/program/>.
IFRC, 2010. World Disasters Reports 2010: Focus on Urban risk. Retrieved from
<http://www.ifrc.org/en/publications-and-reports/general-publications/#>.
Jia, H., Ordez, F., Dessouky, M.M., 2007. Solution approaches for facility location
of medical supplies for large-scale emergencies. Computers & Industrial
Engineering 52 (2), 257276.
Jotshi, A., Batta, R., 2008. Search for an immobile entity on a network. European
Journal of Operational Research 191 (2), 347359.
Jotshi, A., Batta, R., 2009. Investigating the benets of re-optimisation while
searching for two immobile entities on a network. International Journal of
Mathematics in Operational Research 1 (1/2), 3775.
Jotshi, A., Gong, Q., Batta, R., 2009. Dispatching and routing of emergency vehicles in
disaster mitigation using data fusion. Socio-Economic Planning Sciences 43 (1),
124.
Kongsomsaksakul, S., Yang, C., Chen, A., 2005. Shelter locationallocation model for
ood evacuation planning. Journal of the Eastern Asia Society for Transportation
Studies 6, 42374252.
Kovcs, G., Spens, K., 2007. Humanitarian logistics in disaster relief operations.
International Journal of Physical Distribution & Logistics Management 37 (2),
99114.
 G. Galindo, R. Batta / European Journal of Operational Research 230 (2013) 201211
Lettieri, E., Masella, C., Radaelli, G., 2009. Disaster management: ndings from a
systematic review. Disaster Prevention and Management 18 (2), 117136.
Liu, Y., Lai, X., Chang, G.-L., 2006. Cell-based network optimization model for staged
evacuation planning under emergencies. Transportation Research Record:
Journal of the Transportation Research Board 1964, 127135.
Lodree Jr., E.J., Taskin, S., 2009. Supply chain planning for hurricane response with
wind speed information updates. Computers & Operations Research 36 (1), 2
15.
Matisziw, T.C., Murray, A.T., 2009. Modeling st path availability to support disaster
vulnerability assessment of network infrastructure. Computers & Operations
Research 36 (1), 1626.
McLoughlin, D., 1985. A framework for integrated emergency management. Public
Administration Review 45, 165172.
Minciardi, R., Sacile, R., Trasforini, E., 2009. Resource allocation in integrated
preoperational and operational management of natural hazards. Risk Analysis
29 (1), 6275.
Nagurney, A., Yu, M., Qiang, Q., 2011. Supply chain network design for critical needs
with outsourcing. Papers in Regional Science 90 (1), 123142 (Published online
in 2010).
Natarajarathinam, M., Capar, I., Narayanan, A., 2009. Managing supply chains in
times of crisis: a review of literature and insights. International Journal of
Physical Distribution and Logistics Management 39 (7), 535573.
Opasanon, S., Miller-Hooks, E., 2009. The safest escape problem. Journal of the
Operational Research Society 60, 17491758.
Our Amazing Planet, 2010. Natural Disaster Deadly, Costly in 2010. Retrieved from
<http://www.ouramazingplanet.com/natural-disaster-deaths-110103-0921/>.
Paul, J.A., Batta, R., 2008. Models for hospital location and capacity allocation for an
area prone to natural disasters. International Journal of Operational Research 3
(5), 473496.
Peacock, W.G., Morrow, B.H., Gladwin, H., 1997. Hurricane Andrew: Ethnicity,
Gender and the Sociology of Disasters. Routledge, New York, 278pp.
Rawls, C.G., Turnquist, M.A., 2010. Pre-positioning of emergency supplies for
disaster response. Transportation Research Part B: Methodological 44 (4), 521
534.
Saadatseresht, M., Mansourian, A., Taleai, M., 2009. Evacuation planning using
multiobjective evolutionary optimization approach. European Journal of
Operational Research 198 (1), 305314.
Salkowe, R.S., Chakraborty, J., 2009. Federal Disaster Relief in the US: The Role of
Political Partisanship and Preference in Presidential Disaster Declarations and
Turndowns. Journal of Homeland Security and Emergency Management 6 (1)
(article 28).
211
Sbayti, H., Mahmassani, H.S., 2006. Optimal scheduling of evacuation operations.
Transportation Research Record: Journal of the Transportation Research Board
1964, 238246.
Schulz, S.F., Blecken, A., 2010. Horizontal cooperation in disaster relief logistics:
benets and impediments. International Journal of Physical Distribution &
Logistics Management 40 (8/9), 636656.
Simpson, N.C., Hancock, P.G., 2009. Fifty years of operational research and
emergency response. Journal of the Operational Research Society 60 (1),
S126S139.
Snediker, D.E., Murray, A.T., Matisziw, T.C., 2008. Decision support for network
disruption mitigation. Decision Support System 44 (4), 954969.
Stepanov, A., Smith, J.M., 2009. Multi-objective evacuation routing in transportation
networks. European Journal of Operational Research 198 (2), 435446.
Taskin, S., Lodree, E.J., 2010. Inventory decisions for emergency supplies based on
hurricane count predictions. International Journal of Production Economics 126
(1), 6675.
Tuydes, H., Ziliaskopoulos, A., 2006. Tabu-based heuristic approach for optimization
of network evacuation contraow. Transportation Research Record: Journal of
the Transportation Research Board 1964, 157168.
Wang, J.W., Ip, W.H., Zhang, W.J., 2010. An integrated road construction and
resource planning approach to the evacuation of victims from single source to
multiple destinations. IEEE Transactions on Intelligent Transportation Systems
11 (2), 277289.
Wright, P.D., Liberatore, M.J., Nydick, R.L., 2006. A survey operations research
models and applications in homeland security. Interfaces 36 (6), 514529.
Xie, C., Turnquist, M.A., 2009. Integrated evacuation network optimization and
emergency vehicle assignment. Transportation Research Record: Journal of the
Transportation Research Board 2091, 7990.
Yi, W., Kumar, A., 2007. Ant colony optimization for disaster relief operations.
Transportation Research Part E: Logistics and Transportation Review 46 (6),
660672.
Yi, W., zdamar, L., 2007. A dynamic logistics coordination model for evacuation
and support in disaster response activities. European Journal of Operational
Research 179 (3), 11771193.
Zaric, G.S., Bravata, D.M., Holty, J.-E.C., McDonald, K.M., Owens, D.K., Brandeau, M.L.,
2008. Modeling the logistics of response to anthrax bioterrorism. Medical
Decision Making 28 (3), 332350.
Zou, N., Yeh, S.-T., Chang, G.-L., Marquess, A., Zezeski, M., 2005. Simulation-based
emergency evacuation system for ocean city, Maryland, during hurricanes.
Transportation Research Record: Journal of the Transportation Research Board
1922, 138148.