A typological framework for categorizing infrastructure vulnerability

Author(s): Tony H. Grubesic and Timothy C. Matisziw

Source: GeoJournal, Vol. 78, No. 2, Special Section on Geographic Aspects of Vulnerability
(2013), pp. 287-301
Published by: Springer
Stable URL: http://www.jstor.org/stable/42006320
Accessed: 04-03-2018 04:21 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

Springer is collaborating with JSTOR to digitize, preserve and extend access to GeoJournal

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
Geojournal (201 3) 78:287-301
DOI 1 0. 1 007/s 1 0708-0 1 1 -94 1 1 -0
A typological framework for categorizing infrastructure
vulnerability
Tony H. Grubesic • Timothy C. Matisziw
Published online: 3 May 201 1
© Springer Science+Business Media B.V. 201 1
Abstract The concept of vulnerability is increas-
ingly important in engineering and the socio-economic
planning sciences, particularly given the enormous
costs associated with addressing it. The ability to
identify and mitigate vulnerabilities is extremely
challenging because it is influenced by a complex
and dynamic set of interacting factors that can
compromise social, economic and infrastructure
systems. Where the latter is concerned, the ability
to assess infrastructure vulnerability involves the
consideration of a range of physical, operational,
geographical and socio-economic characteristics. In
this paper, significant elements of infrastructure
vulnerability are identified and discussed with a focus
on their intrinsic spatial nature and their propensity to
interact across space. Further, the developed typology
of vulnerability outlined in this paper emphasizes the
need to ensure that policy, planning and disaster
T. H. Grubesic (E3)
College of Information Science and Technology, Drexel
University, Philadelphia, PA 19104, USA
e-mail: grubesic@drexel.edu
T. C. Matisziw
Department of Geography, University of Missouri,
Columbia, MO 6521 1, USA
T. C. Matisziw
Department of Civil and Environmental Engineering,
University of Missouri, Columbia, MO 65211, USA
This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
mitigation efforts are strongly integrated at global,
regional and local levels.
Keywords Critical infrastructure • Networks •
Interdependency • Interdiction • Disaster mitigation •
Spatial analysis • Public policy
Introduction
In the United States, there is a growing interest in
both social vulnerability (Cutter and Finch 2008;
Borden et al. 2007) and the vulnerability of infra-
structure systems, both of which are intimately
related (NRC 2006; White House 2003). Moreover,
during the past two decades, both the executive and
legislative branches of the U.S. government have
made concerted efforts at better defining what
systems and assets are most important to the day-
to-day functioning of the nation, giving rise to what is
now termed critical infrastructure (Sec. 1016[e]).
Critical infrastructure and assets include transporta-
tion systems, telecommunication networks, the elec-
trical grid, banking networks, reservoirs, natural gas
distribution systems, and many other interdependent
infrastructures, vital to the day-to-day functioning of
social, economic and physical systems in the United
States (Lewis 2006; Murray and Grubesic 2007).
Although, the importance of such infrastructure is
obvious, how to most appropriately describe sources
of vulnerability is not, particularly given the sheer
Springer
 288 GeoJournal (2013) 78:287-301

variety of critical
an entity assets and
to be vulnerable. In this system
context, the concept
matters is the ofrange and
risk is an important one. Kaplan level
and Garrick o
infrastructure systems
(1981) summarize risk as provide,
a simple function of three the
ces, and factors: (1) threats,
potential what can go wrong; (2) whatallis the of
significantly probability of
across it going wrong;
time and(3) what are the
space
The purpose of this
consequences if it does gopaper
wrong. For example is
typological framework
consider the impacts offor
a hurricane categoriz
making landfall in
vulnerability. a populated area.
While The loss of human life, economic
complimentary
social disruption and
vulnerability, the environmental/physical
which focuses damage o
of populations to hazards
caused and
by high winds, rain, storm surge andtheir
flooding
to/recover are certainly
from the negative outcomes. However, when
impacts of
(Cutter 2006), the analysis
efforts are of
made to mitigate risk, levels infra
of exposure
ability is primarily concerned
may not change, but the degree to which an area, w
operational, andpopulation
geographic charact
or system is vulnerable can be altered. For
structure instance, by
elements, developing and fragility
their implementing a compre- to
in the system(s) hensive
with hurricanewhich
evacuation plan, one does not
they in
the potential decrease the probability that aof
implications hurricane will hit
disrup
many factors influence vulnerab
communities along a coastal region, but it can reduce
primarily focuses
the vulnerability
on of their
populations within
spatial
the region. f
often the most difficult features to characterize. It is There are also elements of uncertainty associated
also important to note that the development of a with vulnerability. In many situations it is difficult to
typology concerning the spatial facets of critical ascertain the specific threats or events which may
infrastructure vulnerability, is largely absent in the generate negative outcomes for a population or
existing literature. While multiple facets of infra- infrastructure system (Sage and White 1980). Again,
structure vulnerability are certainly addressed in knowledge (or lack thereof) of these threats does not
other work (ASCE 2005; Pederson et al. 2006; decrease exposure (Holton 2004), but a systematic
Murray and Grubesic 2007), the failure to provide an understanding of, and planning for potential threats
overarching framework for categorizing these facets can help reduce vulnerability.
is a major oversight. Given the geographic emphasis Although the fundamental concept of vulnerability
of the taxonomy developed in this paper, it is hopedis related to the susceptibility of people, places and
that this will serve as a useful multiscalar (i.e. global, infrastructure to disruptive events, there are elements
regional and local) framework for developing more of uncertainty associated with how, when and where
effective and holistic policy, planning and disaster these events will occur. For example, although the
mitigation efforts. timing of earthquakes has proven exceedingly diffi-
In the next section, a brief review of the vocab- cult to predict, the scientific community has devel-
ulary, terms and literature that examines the spatial oped a relatively good understanding of which areas
and temporal context of infrastructure vulnerability is are vulnerable to major seismic events (USGS 2008).
provided. This is followed by the introduction of a It is also important to acknowledge that vulnera-
typological framework for categorizing infrastructurebility comprises elements of sensitivity and response.
vulnerability. This paper is concluded with a briefFor instance, some infrastructure systems are more
discussion and provide a roadmap for future research.resilient (i.e. able to recover or respond quickly) to
disruptions than others. Portions of this resilience can
often be attributed to careful planning (e.g. organized
Vulnerability vocabulary evacuations, availability of shelter, water, food and
medical supplies) and through good infrastructure
The concept of vulnerability is multifaceted (Adger design (e.g. retrofitted buildings, redundant and
2006). By definition, vulnerability means suscepti- diverse supply and distribution systems, etc.). From
bility to injury or attack (MW 2008a). However, there a social perspective, Cutter and Finch (2008: 2301)
are a number of pre-conditions that are required for note that response is also a multidimensional

<0 Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 Geojournal (2013) 78:287-301 289

construct and certain

is not services (e.g. broadband, electricity,
easily capture rail
variable. As a result,
service, etc.) are spatiallyelements
and temporally heteroge- o
economic class, neous. As a result, the and
gender impacts of infrastructure
housing
ability of socialfailures
groupsand associated lossesto
can varyrespond
dramatically.
extreme events. A similar argume
the economic Spatial
impacts of infrastru
ardson et al. 2006), where many d
be considered. As noted previously, the spatial vulnerability of
Regardless of the type of vulnerability in question infrastructure systems is multifaceted. In a highly
(e.g. social, economic or infrastructural) both the generalized context, the locations where infrastruc-
timing and location of disruptive events can influence ture is placed, along with the relative locations of
their overall impact. For example, while the loss of failures within or between systems, is extremely
electricity for several days or even a week would important when evaluating spatial vulnerability. For
have minimal impact on rural villages in developing example, Grubesic et al. (2008) illustrate that
countries where the power grid is thin, that same loss depending on the spatial configuration of telecom-
would have catastrophic impacts to New York City, munication nodes in network, a variety of scenarios
where electrical power is crucial for the day-to-day of infrastructure damage exist, each having a differ-
operations of transportation systems, telecommuni- ential impact on infrastructure performance - both in
cation networks, water distribution pumps, comput- terms of magnitude and geographic scale. Further, the
ers, etc. In fact, this is exactly what happened in disruptive potential of each scenario can be measured
August 2003, when a series of infrastructure-related in numerous ways, such as the impact to system
failures generated a massive blackout in the North- capacity, cost, connectivity, demand/provision, and
eastern U.S. and portions of Canada (Minkel 2008). redundancy. As a result, decline in system connec-
In this scenario, while the type of failure is identical tivity may not be particularly problematic if connec-
(i.e. loss of electricity), the vulnerability of these two tivity among vital supply and demand nodes is still
locations is dramatically different. available, but the delays or re-routing costs associated
Perhaps the most significant concern relating to the with delivering a good or service may be crippling.
vulnerability of social or infrastructural systems is Again, this hints to the mutuality of infrastructure,
potential for failure. Specifically, where infrastruc- social vulnerability and failure. For example, as noted
ture is concerned, the concept of failure does not by Cutter and Finch (2008), locations increasing in
necessarily imply uniform levels of service disruption social vulnerability between 1960 and 2000 often did
within a system. In many cases, critical infrastruc- so because of extreme population growth and lack of
tures have back-up facilities or services that ensure corresponding infrastructure support. In this case, the
operational continuity. For example, most hospitals authors cite Orange County, California as a location
maintain some type of back-up electrical system if that was determined to be moderately vulnerable in
the main power grid fails. However, vulnerabilities 1960, but due to a nearly 300% growth in population
do increase the risk of failures because systematic (among other things) is now considered one of the
weaknesses increase the likelihood of system disrup- most socially vulnerable location in the United States.
tions when exposed to extreme events. Not surpris- Population growth generates greater demands for all
ingly, disruptions can lead to different types of resources and services. In areas where system capac-
failures given the dependencies among system com- ities and use are near maximum levels, vulnerability
ponents as well as interdependencies between sys- increases and any type of disruption or failure can
tems (Rinaldi et al. 2001; McDaniels et al. 2007). intensify the social or systemic impacts.

Temporal
The spatial and temporal aspects of vulnerability
In addition to the spatial implications of infrastruc-
Given the complex interplay of vulnerability and ture vulnerability and failure, systems and their use
failure, it is also important to note that the demand for are rarely static - requiring the acknowledgement of

â Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 290 Geojournal (2013) 78:287-301

temporal dimensions of
most recent (2009) vulnerabilit
evaluation are sobering. Aviation,
overly dams, drinking
generalized statementwater, energy, hazardous
for waste, spa
provided inland waterways,
previously, the levees, roads,
"when"schools, transit, of
structure and wastewater allis
disruptions received grades of D-f, D or as
equally im
"where". For D- (ASCE 2009). Essentially,
example, if we a "D" quality level in
revisit
electrical disruptions, the
the ASCE report means temporal
that the infrastructures are
ated with not functioning
outages can inbe a safe and reliable manner,
quite pron
the teetering on collapse.
use
electricity of oftenThe question is,spikes
why? More b
winter (heating) and
importantly, how do summer
these conditions contribute to mon
temperate, mid-latitude
infrastructure vulnerability? climates (
loss of electricity during
As discussed these
in the previous sections, peri
vulnerability
problematic manifests
than duringin a variety of ways. Both spatial and spr
milder
months when climate related
temporal aspects of the level of demand for ahealth
service
heat stroke or are of obvious importance.
hypothermia) may be
Switching sectorsHowever,
for othera
dimensions
moment, of vulnerability do tem
can also be examined at a much smaller scale. exist for critical infrastructures. In this section, we
Consider the loss of telecommunications services to a will introduce these additional considerations and
residential household at 3 a.m. versus 7 p.m. The propose a typological framework for better under-
impacts of the 3 a.m. loss are likely less problematic standing their contributions to vulnerability. Figure 1
than the same loss at 7 p.m., when the demand and presents a depiction of our typological framework
use of broadband services is highest. It is also and displays eight distinct facets of infrastructure
important to note that larger temporal windows (e.g. vulnerability, including condition, capacity and use ,
decadal or multi-decadal) also impact vulnerability of obsolescence , location and topology , interdependen-
populations and infrastructure. The duration of a ces , disruptive threats , policy and political environ-
disruption is also relevant. As evident from the 9/1 1 ment , and safeguards. While this is not an exhaustive
attacks, although backup systems (e.g. emergency list, largely ignoring issues related to economic and
power supply) can provide a short-term bridge for social vulnerability, we believe it captures the bulk of
operational continuity, they are not viable long-term those factors underlying the geographic aspects of
substitutes (Grubesic and Murray 2006), frequently infrastructure systems. Individually, each of these
failing prior to initialization or simply running out of
alternative energy supplies (i.e. diesel).
Regardless of how one conceptualizes vulnerabil-
ity in time and space, infrastructures maintain a range
complex relationships (amongst and between each
other) and characteristics that contribute to their
vulnerability. As a result, efforts directed at main-
taining the operational continuity of critical infra-
structure systems must be driven by a broader
understanding of vulnerability. In the next section, a
typological framework for categorizing infrastructure
vulnerability is proposed to address these issues.

A typology of infrastructure vulnerability

Every 2-3 years, the American Society for Civil

Engineers (ASCE) releases a report card that sum-
marizes U.S. infrastructure on the basis of condition,
performance, capacity and funding. The results of the Fig. 1 Eight facets of critical infrastructure vulnerability

Ô Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 GeoJournal (201 3) 78:287-301 291

eight has aspects

suspension the
system on the potential
Silver Bridge, which linked t
infrastructure Point Pleasant, West Virginia and Gallipolis,
disruptions. It Ohio.is im
The bridge collapse
however, that most killed 46 people and motivated
infrastructure s
taneously influenced
the establishment of theby a multip
National Bridge Inspection
Program, which requires
which can contribute to that every bridge longer
systemic d
over, while we have
than twenty feetendeavored
be inspected every 2 years (Reid to
category is relatively exclusive,
2008). In this context, because decay compromises
unavoidable. the structural integrity of critical infrastructure and
assets, decay contributes to the vulnerability of these
Condition and decay systems. It is also important to note that the "trigger"
for condition-related failures need not be an extreme
By definition, decay refers to a decline from a soundevent as they can also occur under normal use.
or prosperous condition or, destruction by decompo- Finally, condition and decay can also impact
sition (MW 2008b). The sources of infrastructurefederal, state and local highway and road systems in
decay vary widely, but dams provide a particularly more mundane ways. Environmental conditions com-
interesting case-study for considering this process. bined with car and truck traffic eventually degrade
Decay, through erosion, corrosion, weathering or
the quality and condition of both concrete and asphalt
other atmospheric and geologic forces stronglyroad systems. Decay, in the form of potholes or worn
surfaces increases both accident frequency and
contribute to both decomposition and dam failure.
Further, many problems in the condition and struc-
municipal liability (Tighe et al. 2000). Furthermore,
tural integrity of dams can be attributed to the
maintenance needed to mitigate the effects of decay
movement and/or failure of the foundation supportingand maintain suitable infrastructure condition can
the dam and inadequate maintenance and upkeep require significant expenditures, detracting resources
(FEMA 2008). According to the 2009 ASCE report, from other areas of need.

there are 15,237 high hazard dams in the United

States, an increase of nearly 3,300 since 2007. 1Capacity and use
Interestingly, there were 67 dam incidents were
reported between 2003 and 2005. While this figure As alluded to in the previous sections, the availability
may seem relatively low, particularly considering that
of infrastructure is a major concern in many regions.
there are over 80,000 dams in the U.S., these For example, estimates suggest that the world's
incidents only hint to the breadth of the problem. infrastructure stock is valued at approximately
For instance, current estimates suggest more than$15 trillion, but only 13% of this stock is located in
3,500 dams have deficiencies severe enough to leave low income countries (Fay and Yepes 2003). The
them susceptible to failure. Further, it is estimated problem is particularly acute in Africa, where pop-
that $36 billion will be needed over the next decade ulations in many of the least developed countries,
to rehabilitate and repair the most critical structures such as Niger and Chad, lack access to electricity,
(Reid 2008). telephones and other relatively standard physical
Decaying infrastructure also impacts other critical infrastructure systems (Borgatti 2005). In other cases,
assets, particularly bridges. As recounted by Lich-access to infrastructure may not be a problem, but the
tenstein (1993), one of the worst bridge disasters in availability of infrastructure that can handle all of the
U.S. history occurred because of a corroded eyebardemand placed upon it, can be.
The definition of capacity is the maximum amount
or number that can be contained or accommodated
1 According to the ASCE (2009), A dam's "hazard potential"
is classified on the basis of the anticipated consequences of(MW 2008c). For infrastructure systems, this might
failure, not the condition of the dam. The classifications represent bandwidth availability on a fiber-optic
include "high hazard potential" (anticipated loss of life in the backbone, or the volume of car and truck activity a
case of failure), "significant hazard potential" (anticipated
highway can accommodate. Increasingly, infrastruc-
damage to buildings and important infrastructure), and "low
hazard potential" (anticipated loss of the dam or damage to the ture systems in the United States and abroad lack the
floodplain, but no expected loss of life). capacity to meet user demand. For example, consider

Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 292 Geojournal (2013) 78:287-301

the commercial Infrastructure

air obsolescence
traffic system in
In a recent report by the Federal
tration Although there is no universally
(2007), estimates accepted definition
suggest th
seven of infrastructure
metropolitan areas obsolescence, any element
will of a
need
and/or system that suffers from poor
service capacity by design, 2015
is based on to
flowing outdated engineering
smoothly andor design
to components,
meet can no pro
demand. longer meet current
However, safety standards or lacks
simply meetthe
physical ability to handle new applications,
infrastructure needs can be consid-
for t
ered obsolete.
needed improvements will requiThat said, it is important to note the
differences between
investment (FAA 2007). With p obsolescence and infrastructure

projected to condition/decay.
exceed 1 Specifically,
billion while obsolete infra-
by 20
sufficient structure canexists
capacity be in poor condition inor a decayed
this s
state, this is not
According to the always the case. Simply
ASCE put, brand
(2005), f
"significant new components can be obsolete, particularly inve
infrastructure if
delays are they do not meet
expected to cost th current application standards.3
$170 billion Obviously, things 2000
between get more complexandwhen engi-2012
In this type ofneered systems and critical infrastructure
operating are exam-
environ
induced ined for both obsolescence
vulnerability and condition. As noted
manifests in
instance, if we by the ASCE (2009), while
revisit the many infrastructure
capacity
ciated with systems are in extremely
airports, many poor condition,
major it is
capacity extremely difficult
to handle to say if one or more parts
traffic are
durin
According to obsolete. Analysis
Forrey at this minute scale
(2007), is time
preside
Air Traffic consuming, costly and
Controllers Associatio requires a significant invest-
were scheduled ment of
tohuman resources.
depart from
Unfortunately, it is all toobetween
International Airport frequent that one of the 9
on September parts
5, of an infrastructure system
2007, only is linked to a45
accommodated.catastrophic
Two system malfunction.
days For example,
later, one of a
occurred in Chicago at O'Hare
the most high-profile infrastructure failures in the last I
lack of capacity25 years was the collapse of airports
makes the Interstate 35 W
Bridge in Minneapolis,
significant ground control Minnesota indelays
August 2007. p
cient Killing 13 of
functioning people andboth
injuring nearly 100 more, the
airlines
for passengers. Nationalother
In Transportation and Safety Board (NTSB
infrastruct
of capacity has 2008) investigation revealed
similar several problems asso-
impacts. Fo
disrupted ciated with
systems arebridge design
unableand its structuralto
condi- rer
a lack tions. Post-collapse
capacity, of the analysis indicates that the bridge
ability for a
was constructed
and readjust after a disruption with gusset plates that were too thin i
for connecting steel beams
telecommunications in the truss bridge. In fact,
systems, th
the gusset plates were
congestion collapse is aboutparticular
50% too thin for this
particular application, resulting in sixteen
(Nagle 1984; Albuquerque et platesal. 2
with the loss of high bandwidth l
width connections are forced t
packets. If the remaining
Footnote 2 continued lower ba
overloaded with transmission
little throughput is available in the system. As a result, high r
increases in the system and con
levels of latency, packet delay and loss emerges (Johari and
Tan 2001).
becomes a possibility (Johari an
A good analogy here is the eight-track audio cartridge. While
querque et al. (2004).2
it is possible to design, produce and use a brand new eight-
track cartridge today, the technology is obsolete and the sound
2 Congestion collapse occurs when an overloaded network has quality, relative to digital audio technologies (e.g. compact
settled into a stable state, where traffic demand is high, but disc) is poor.

^ Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 GeoJournal (201 3) 78:287-301 293

being fractured Interdependencies

prior to the collapse (
addition, because the bridge was
fabricated in the 1960s,
A fourth it is included
aspect to vulnerability the degree to which
an infrastructure
"fracture critical" design system is reliant or dependent upon
component
failure of a fracture critical
another infrastructure memb
system for operation. While
this seems
expected to result in to bethe
a relativelycollapse
simple concept, it is of
(TRB 2005). important to note that there is a difference between
One of the majordependency
problemsand interdependency. Rinaldi et al.FC
with
the lack of (2001, 14) definefor
redundancy dependency
theas, "a linkage
struct or
structures. In connection
this between two infrastructures,
context, redundanc through
"the which the state of one
quality
bridge infrastructure
that of influences or is
enables a it
design function correlated
in to the state ofdamaged
the another". This suggests a sta
LRFD 2004). In unidirectional
a recent report
relationship between systems, where re
one relies on the other for functionality. Conversely,
inspection and maintenance of FCM
they define interdependency
Transportation Research Board (200 as "a bidirectional
relationship between two
during the late 1970s, infrastructures through
materials, de
which the state of
fabrication of steel bridge compone each infrastructure influences or

dramatically, is correlated to the

resulting instate of the other". In this
fewer instan
and fracture. However, while
instance, the infrastructures these
are co-dependent on each
improved othererected
bridges for operation. Pederson et al. (2006) formalize
after the
were infrastructure
implemented, 76% interdependency
of the asbridges
a network as
the United States were fabricated before 1978. In follows:

fact, national statistics reveal that approximately 11%

1. An infrastructure network, /, is a set of nodes
of all steel bridges in the U.S. have FCMs (TRB
related to each other by a common function. The
2005).
network can be connected or disjoint. It may be
Obviously, there are other examples of obsolescent
directional, bi-directional or have elements of
infrastructure and related failures, such as the steam
both. Internal relationships/dependencies within
pipe explosion in New York City during July 2007
I are represented by edge {a, b) with a, b e I.
(Barron 2007) and problems with water storage tanks
2. Given /f and Ij are infrastructure networks, i ^ j,
and low-pressure water distribution lines in Pitts-
a e Ii and b e Ij, an interdependency is defined as
burgh, Pennsylvania (PIIA, 2007). Many of these
a relationship between infrastructures and repre-
older pipeline infrastructures still use cast iron
sented as the edge ( a , b) which implies that node
components. Therefore, in addition to rust and decay
b is dependent upon node a. The nature of
(i.e. condition), cast iron systems are far last resistant
this relationship may also be reflexive in that
to geological events such as earthquakes - particu-
(a, b) -► C b , a).
larly when compared to newer steel or polymer
structures. Taxonomies for interdependencies have also been
Clearly, in all of the examples illustrated above, developed. For example, Rinaldi et al. (2001) differ-
infrastructure longevity, system age, design and the entiate between four basic types of interdependen-
diminished ability of these infrastructures to copecies. The first, physical , is the reliance on material
flow from one infrastructure to another. Cyber
with increased demands make them highly vulnera-
ble. As noted by Marland and Weinberg (1988),interdependencies are largely related to computer
control systems or a reliance on computerized
infrastructure systems do have limited life-spans,
even if the exact duration of their functionality isinformation transfer between infrastructure systems.
If changes in the local spatial environment can affect
difficult to precisely predict. Simply put, devices and
components across multiple infrastructure systems
structures wear out, operation and maintenance can
become too expensive and competing systems and due to physical proximity, a degree of geographic
interdependence is in place. Finally, logical interde-
technologies become available - offering improved
functionality at less cost. pendencies reflect linkages in human decision

â Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 294 GeoJournal (2013) 78:287-301

making, are bidirectional and d

lifeblood of many information intensive businesses
depend on any throughout the United States spatial
physical, (Mack and Grubesic or
While space 2009), the geographic footprint
limitations prevent of most backbone us
of the nuances associated
systems includes extremelywith
large continentalinfras
tran-
pendences, sects through the most sparsely
particularly those populated regions in
detail
as complex adaptive
the United States (Fig.systems
2). The same can be said for (A
1999), most details
additional large networked infrastructures
can inbe the U.S., fou
provided by including gas pipelines
Rinaldi et (Fig.al.
3). This (2001).
makes the
numerous geographic location
examples in ofrecent
network components and their
years
infrastructure ambient environment important factors when evalu-
interdependencies e
bilities. Obviously, the
ating the vulnerability example
of infrastructure systems.
regarding the Consider,
August for instance, the differential
2003 in time and
blac
excellent case study
effort associated within how
transporting equipment andinter
crews for repairing
structures fail after an supervisory
initial control and shock
data
(i.e. electrical). acquisition
For (SCADA)
more devices or other network
details,
Murray (2006), elements
Minkelto geographically remote
(2008)or inaccessible or
When one locations examining
begins versus more central, urbanized locales.
the For in
between more example, Gothenburg,
than two Nebraska is located along sys
or three
matrix of Interstate 80, a major corridor
interactions for fiber optic back-
becomes ex
Complicating matters is isthe
bones in the United States. Gothenburg also located no
between at least 250
systems. miles away from any major
Tightly city (e.g.
coupled
slack buffer Denver or Omaha)
or
in and most likely
their lacks the
operatio
resources, equipment
They are time dependent, goods and technical expertise to se
mount an effective local response
mation are continually to heavy fiber
moving,
backbone damage. Conversely,
process are invariant, and support crews and
reactions
equipment located
system are almost in Chicago could respond to a
instantaneou
Conversely, major disruption
loosely on a fiber backbonesystem
coupled (or any
ant of infrastructure-related
processing delays, problem) insequencin
the city within
alternative minutes.
methods may be ava
Similarly, some areas of to
necessary, and reactions the United States are
changes
not immediate. more
For prone to more details,
natural disasters, such as wildfires, s
floods, hurricanes, tornadoes, etc. As a result, the
Location and ambient local environment often influences the
topology
overall vulnerability of infrastructure systems, par-
As noted in the ticularly if they are exposed to extremesecti
introductory events.
location of Consider, for example,
critical infrastructure the geographic distribution of
components interstate
plays an gas pipelines displayed in Fig. 3. Extremely
important rol
system vulnerability. Regardles
high densities of pipelines exist along the Gulf Coast
region (e.g.
systems maintain someLouisiana) anddegree
portions of Oklahoma.of
Both locales,
Again, while there is incidentally,
uncertaintyare subject to extreme a
specifics of weather, withwhen
where, hurricanes along the
andGulf of Mexico
why
occur, the and coastal Louisiana andlocations
geographic tornadoes in Oklahoma.4 o
critical Another aspect to systems
infrastructure consider is the spatial scale
do
vulnerability. associated with geographic vulnerability. As noted
It is important to note that because critical
infrastructure varies in form and function, as does
the demand for services provided by these systems, 4 While a significant portion of pipeline infrastructure is
their locational attributes are also varied. For exam- subterranean, a tornado recently hit a natural gas pumping
ple, while telecommunications backbones are the station in Tennessee (AP 2008), generating a massive fire.

Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
Geojournal (2013) 78:287-301 295

Fig. 2 a Qwest nationwide converged back

previously, the Typically,

form the most
and function of
the United using
States varya tremendou
directed li
critical topology
systems have are the
geographic foom
multiple many
continents, instances,
others the
are limi
cities or even a link
single
in building. A
these system
considering network. For
vulnerability, more
geograph
Finally, the systemstructure
topological topology of
an
structure Grubesic
systems (2007).
is an importan

Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 296 Geojournal (2013) 78:287-301

Fig. 3 United States natural gas pipeli

Disruptive threats
communications
tion to the loss of
The set of region
potential (Louisiana
threats to cri
highly diverse, eight 911
ranging call na
from ce
ical disasters to sabotage.
many of Consid
the mos
overall spatial were lost
distribution to
of flo
na
United States. House 2006).
Schmidtlein Iro
et
clusters of environmental
systems threat
that man
Plains (tornadoes and flooding
problems. In mo
(hurricanes and tropical
lines were storms)
overw
fornia (wildfiresnating and earthquake
from the
that other Combine these telecommunication
regions are losses with the
devoid of
1965 and 2004 natural disasters
damage done to transportation infrastructure, levees, w
the previously identified
the electrical areas
system, hospitals and most govern-
2008). mental buildings in New Orleans, the severity of this
The now classic example
natural disaster at the local and regionalof
level is a
disaster, Hurricane
apparent. Katrina, ex
ronmental threats
While Katrina wasand
an extreme eventthe
that garnered inter
and storm surge wreaked
national attention, there are also many smaller-scale hav
structure systems atthreats
environmental and technological local
that occur on and
example, in the
a weekly,report A
monthly or yearly basis in the Failure
United States
by the U.S. House of
that do not generate significant, Represe
long-term attention
brief inventory of
although they damage
can be similarly damaging to specific to th

Ö Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 GeoJournal (2013) 78:287-301 297

infrastructure sectors. Winter storm related losses in Sarewitz et al. (2003) explore six different "asser-
the Northeastern U.S. exceeded $45 million between tions" associated with both vulnerability and risk in the
1971 and 2007 Changnon (2008). Similarly, techno- context of policy, ranging from the need for acquiring
logical threats, such as the Baltimore Freight Rail accurate probabilistic information about extreme
Crash of 2001 which generated a chemical fire that events to understanding that such events are context
burned for nearly a week, ruptured a water main that driven. Consider, for example, the impacts of political
caused streets to flood, disrupted East Coast rail and economic rows between countries (or corpora-
service and slowed the Internet are also extremely tions) on the operational continuity of critical infra-
disruptive to a wide variety of sectors (NTSB 2004). structure systems. In 2008, Gazprom, headquartered in
While natural and technological disasters are unin- Russia and one of the largest oil and natural gas
tended, Sabotage and terrorist attacks are acts of companies in the world claimed that Ukrainian held
subversion with the direct intent to inflict physical and gas company, Naftogaz Ukrainy, owed it more than
emotional damage to people, property and critical $2 billion in missed payments and fines. Naftogaz
infrastructure. Although the terrorist attacks of Sep- Urkainy countered that they had paid the bill and that
tember 1 1, 2001 provide the most horrific example of Gazprom was trying to force a new price for gas which
such events, there are many others. For example, it could not afford. In a response to this disagreement,
during the initial stages of the U.S. war in Iraq, acts of Gazprom completely shut down its supply pipelines to
sabotage crippled critical infrastructure systems the Ukraine (Kramer 2009). While this may appear to
throughout the country. Glanz (2004) notes that over be a relatively isolated geographic incident, it is
100 electrical backbone lines were cut between 2003 important to note that Austria, Turkey, the Czech
and 2004, with 1,200 transmission towers toppled.Republic, Germany and Greece were impacted by the
Even after President Bush declared the end of major shutdown, forcing these nations to seek alternate
hostilities in Iraq (April 2003), over 200 oil pipeline supply sources for nearly a week. As a result of this
attacks occurred (April 2003 through December 2004), crisis, the Nabucco Pipeline project was developed,
primarily along the 600-mile, 40-inch Kirkuk-Ceyhan seeking to route alternative supplies of natural gas to
pipeline (Luft 2005). Europe via Turkey (Lyons 2010) - lessening Europe's
Interestingly, sabotage is not always committed by dependence on Russian supplies and its politically
outside threats such as terrorists or guerilla armies. Invulnerable distribution system.
many instances, acts of sabotage are committed by In a similar vein, geopolitical context also impacts
"insiders". As noted by Keeney (2005), these are homeland security policy in the United States. With the
typically individuals who were authorized to userecent failures of command and control infrastructures
infrastructure or its associated support systems (e.g.in the U.S. due to extreme events such as the
SCADA) that eventually leveraged this access to September 1 1th attacks and Hurricane Katrina, home
perpetrate a destructive act. Results of this detailed land security policy and vulnerability planning
study also reveal that 59% of the saboteurs were slowly moving to a more regionalized structure. For
former employees or contractors while 41% wereexample, Caruson and MacManus (2007) argue tha
currently engaged with the victimized company. The strongly integrated regional systems help overcome
vast majority of the insiders were employed in the multiplicity of state agencies and local govern-
technical positions and 96% were male. In additionments during disasters, facilitating stronger vertical
to citing odd pre-attack behavior around the office by (e.g. federal-state-local) and horizontal (local-local
the saboteurs, it was determined that the majority ofnetworks. In essence, these stronger and better-
perpetrated attacks were accomplished using com-integrated networks allow state and local government
pany computer equipment (Keeney 2005). to "harness the collective benefits of shared resources
and information: (Caruson and MacManus 2007, 1)".
Policy and political environment
Safeguards
From a political and policy perspective, vulnerability
is somewhat more difficult to define than many of the A final influence on infrastructure vulnerability is the
other facets of previously discussed. For example,presence and effectiveness of safeguards. In general,

Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 298 Geojournal (2013) 78:287-301

safeguards refer to any

When evaluating actions
the vulnerabilities of different
detect/identify and
critical plan
infrastructure for
systems, one can moreposimp
represent these complexities
infrastructure operation. as a function,
Such eff usin
tracking, various combinations of the
identification of outlined facets. Furthe
vulner
ment of if feasible,
disaster one can also assign either positive (+) o
response/restorat
lance of infrastructure,
negative (- ) factors to protection,
each facet. In many instance
hardening (Church and
the positive Scaparra
and negative factors for each facet20
ar
2007;Matisziw mutable.
et For al. 2008,
example, 2010)
newer systems that are in go
condition and exhibit
better implementation ofno decay would receive
safeguar
result in enhanced proactive
positive factor andofr
for ô. Conversely, if elements t
response, and system are decayed orof
reduction in poor condition, ô mig
vulnera
ple, given the receive a negative factor. Consider,
application of varioufor instance, t
methodologies, worst-case
vulnerability scen
of telecommunications infrastructur
(Vf), which is a function
damage (and associated of capacity and u
componen
fied. Based on obsolescence,
these insights,
interdependencies, vu
location and topol
components can ogy,
thendisruptive be
threats more
and safeguards. effec
More simply
protection, thereby
Vj =/(/+,a", reducing/elim
r,¿+, v",ç+). In this instance, ne
vulnerabilities (Matisziw et
ative factors are assigned al. 200
to obsolescence, interd
Given that resources for
pendencies and disruptive applying
threats. However, if the
safeguards are are
limited, it
no safeguards in place, is capacities
network esse a
accurate characterization
running at a maximumof exactly
level and h
the network displa
of safeguards may result in vulnera
a sparse typology, it is possible to assign negati
course, the ability
factors to
to /, / accomplish th
and ç. The reverse, for obsolescenc
well other elements of
interdependencies andthe
disruptive vulner
threats may also
understood. For true.
instance, developin
Given the facets of vulnerability outlined abov
for one the exact contributionmight
infrastructure of each of these variables to
require
pendences and hardening componen
characterizations of vulnerability will be context
dependent, varying between infrastructure types and
locations.
Discussion and conclusion
Table 1 provides a snapshot of how the different
facets of this typology may interact or display some
Given the preceding narrative on the different facets
level of simultaneity. It is important to note that this
of vulnerability, it is possible to more concretely
matrix does not represent a definitive categorization
define a matrix of vulnerability for individual infra-
of interactions between facets. Instead, it represents a
structure systems. Specifically, the vulnerabilities of a
very broad, context-dependent, view of how they
particular infrastructure, Vh can be thought of as a
might interact simultaneously across systems. What
function of a complex and diverse set of features,
holds true for telecommunications systems may not
roughly corresponding to the outlined facets. Con-
be applicable to the national air transportation
sider the following notation, where each variable
system. That said, it is clear that these facets are
represents a range of potential vulnerabilities:
not mutually exclusive. For example, as noted
ô condition and decay previously, although condition and obsolescence are
X capacity and use unique, both facets have the potential to interact
a obsolescence within a single system (e.g. the I-35W bridge).
i interdependencies Finally, given this paper's focus on the geographic
À location and topology nature of infrastructure vulnerability, it is important
v disruptive threats to highlight some potential geographic strategies to
7i policy and political environment reduce infrastructure vulnerability. While it is clear
ç safeguards that infrastructure systems must be designed to meet

Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 GeoJournal (201 3) 78:287-301 299

Table 1 Vulnerability matrix

Condition Capacity Obsolescen

and use and threats political
typology environment

Condition - A A

Capacity and use A - A A A A A

Obsolescence A A -

Location and typology A - A A A A

Interdependencies A A - A A
Disruptive threats A A - A A
Policy and political A A A -
environment

Safeguards A AAA

A Facets are interrelated or displ

user demand, this can be a difficult task when operational strategies for critical infrastructure sys-
simultaneously attempting to reduce vulnerabilities. tems with more locational flexibility such as critical
Users must be able to access the system somewhere, stockpiles of emergency supplies (Church and Scap-
even if this access increases systematic exposure to arra 2007). These can be moved, albeit with some
risk. Not surprisingly, the need to serve infrastructure effort, to both minimize vulnerability and maximize
demand often results in relatively "problematic" accessibility.
clusters or agglomerations of critical infrastructure in In sum, regardless of the functional representation
certain locations (Parfomak 2005). For example, of vulnerability, or the selection of strategies for
considering that over 45% of U.S. securities are vulnerability mitigation, it is clear that the acquisi-
traded in the American Stock Exchange and New tion, analysis and synthesis of data from a wide
York Stock Exchange in lower Manhattan, A targeted variety of sources is needed to determine the specific
strike to this area, much like the events of September nature of vulnerability for infrastructure systems.
11th, 2001, could yield massive disruptions to theMore importantly, understanding both where and
global marketplace. The question is, how can suchwhen systems may be vulnerable to disruption is
vulnerabilities be mitigated? important for developing disaster mitigation plans
Again, while there is no single 'best' strategy, aand policies structured to minimize systemic weak-
variety of approaches have been recommended. For nesses. Further, as outlined by Murray et al. (2008),
example, during the Cold War era, the Long Linesthe ability to utilize multiple methods for better
Division of AT&T was particularly concerned withidentification and understanding vulnerability is
the vulnerability of critical telecommunications essential - ranging from strategy-specific, simula-
infrastructure in both the military and civilian sectorstion-based and mathematical modeling assessments.
to nuclear attack. In an effort to mitigate the In conclusion, while the presented typological
vulnerability of these systems, AT&T recommended framework for categorizing infrastructure vulnerabil-
a series of geographic strategies, including network ity is both complex and multifaceted, significant work
diversification, separation, avoidance and hardening is still required to capture the many nuances of
to minimize the impacts of a nuclear detonation on vulnerability through space and time and across and
their telecommunication equipment. For more details between systems. As suggested by Sarewitz et al.
on these strategies, see Grubesic and Murray (2005). (2003), extreme events are created by context - and
Obviously, these types of strategies are not always context is highly dynamic. So, while it may be nearly
possible; particularly if there are geographic con- impossible to concretely identify every conceivable
straints (e.g. immobility) associated with critical vulnerability within an infrastructure system, this
infrastructure systems (e.g. oil refining capacity does not absolve policy makers and analysts of their
clustered along the Gulf Coast). However, there are responsibility to explore the complex mesh of

Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 300 GeoJournal (2013) 78:287-301

potential Fay, M., & Yepes, T. (2003). Investing nor

vulnerabilities, in infrastructure: What
their
is needed from 2000 to 2010? World Bank. URL: http://
lating mitigation strategies and
papers.ssrn.com/sol3/papers.cfm?abstract_id=636464.
policies for minimizing the
Federal Aviation Administration impacts
(FAA) (2007). Capacity needs
to critical systems and
in the National society.
Airspace System. URL: http://tinyurl. As
com/24attx.
this paper highlights, these are no
Federal Emergency Management Administration (FEMA)
they are clearly essential ones.
(2008). Why dams fail. URL: http://www.fema.gov/
hazard/damfailure/why.shtm.
Forrey, P. (2007). Airline delays and consumer issues. Testi-
mony before the house transportation and infrastructure
subcommittee on aviation. URL: http://tinyurl.com/5jftzn.
References Glanz, J. (2004). Sabotage taking toll on Iraq's power lines.
New York Times. June 1 1 .

Adger, N. (2006). Vulnerability. Global Environmental Grubesic, T. H., Matisziw, T. C., Murray, A. T., & Snedicker,
Change , 16( 3), 268-281. D. (2008). Comparative approaches for assessing network
Albuquerque, C, Vickers, B. J., & Suda, T. S. (2004). Network vulnerability. International Regional Science Review,
border patrol: Preventing congestion collapse and pro- 31('), 88-112.
moting fairness in the Internet. IEEE/ACM Transactions Grubesic, T. H., & Murray, A. T. (2005). Spatial-historical
on Networking, 72(1), 173-186. landscapes of telecommunication network survivability.
American Association of State and Highway Transportation Telecommunications Policy, 29(11), 801-820.
Officials (AASHTO) (2004). LRFD-US-3. Bridge design Grubesic, T. H., & Murray, A. T. (2006). Vital nodes, inter-
specifications. connected infrastructures and the geographies of network
American Society for Civil Engineers (ASCE) (2005). Report survivability. Annals of the Association of American
card for America's infrastructure. URL: http://www.asce. Geographers, 96(1), 64-83.
org/reportcard/2005/index.cfm. Grubesic, T. H., Murray, A. T., & Mefford, J. N. (2007).
American Society for Civil Engineers (ASCE) (2009). Report Continuity in critical network infrastructures: Accounting
card for America's Infrastructure. URL: http://www.mfra for nodal disruptions. In A. T. Murray & T. H. Grubesic
structurereportcard.org/sites/default/files/RC2009_full_ (Eds.), Critical infrastructure : Reliability and vulnera-
report.pdf. bility. Berlin: Springer.
Associated Press (AP) (2008). Firefighters contain massive gas Holton, G. A. (2004). Defining risk. Financial Analysts Jour-
fire in Tenn. URL: http://www.msnbc.msn.com/id/2302 nal, 60(6), 19-25.
3302/. Johari, R., & Tan, D. K. H. (2001). End-to-end congestion
Axelrod, R., & Cohen, M. D. (1999). Harnessing complexity: control for the Internet: Delays and stability. IEEE/ACM
Organization implications of a scientific frontier. New Transactions on Networking, 9(6), 818-832.
York: Free Press. Kaplan, S., & Garrick, B. J. (1981). On the quantitative defi-
Barron, J. (2007). Steam blast Jolts Midtown, killing one. New nition of risk. Risk Analysis, 1( 1), 11-27.
York Times. URL: http://www.nytimes.com/2007/07/ Keeney, M. (2005). Computer system sabotage in critical
1 9/nyregion/ 1 9explode .html . infrastructure sectors. U.S. Secret Service and CERT
Borden, K. A., Schmidtlein, M. C., Emrich, C. T., Piegorsch, Coordination Center/SEA, Washington, D.C.
W. W., & Cutter, S. L. (2007). Vulnerability of U.S. citiesKramer, A. E. (2009). Gazprom shuts off gas links to Ukraine.
to environmental hazards. Journal of Homeland Security New York Times. URL: http://tinyurl.com/3mb24nh.
and Emergency Management , 4(2). Lewis, T. G. (2006). Critical infrastructure protection in
Borgatti, L. (2005). Status of infrastructure in the LDCs: A homeland security. New York: Wiley.
cluster analysis. Background paper prepared for the least Lichtenstein, A. G. (1993). The Silver Bridge collapse
developed countries report 2006, UNCTAD, Geneva. recounted. Journal of Performance of Constructed
Caruson, K., & MacManus, S. A. (2007). Designing homeland Facilities, 7(4), 249-261.
security polity within a regional structure: A needs Luft, G. (2005). Pipeline sabotage is terrorist's weapon of
assessment of local security concerns. Journal of Home- choice. Energy Security. URL: http://www.iags.org/n0328
land Security and Emergency Managementy 4(2), 1. 051.htm.

Changnon, S. A. (2008). Losses from sleet storms in the UnitedLyons, W. (2010). Nabucco at center of gas politics. Wall
States. Natural Hazards , 47 , 465-470. Street Journal. URL: http://tinyurl.com/yehbd9g.
Church, R. L., & Scaparra, M. P. (2007). Protecting criticalMack, E. A., & Grubesic, T. H. (2009). Broadband provision
assets: The r-interdiction median problem with fortifica- and firm location in Ohio: An exploratory spatial analysis.
tion. Geographical Analysis, 39(2), 129-149. Tijdschrift voor Economische en Sociale Geografie,
Cutter, S. L. (2006). Hazards , vulnerability and environmental 100(3), 298-315.
justice. London: Earthscan Publications. Marland, G., & Weinberg, A. M. (1988). Longevity of infra-
Cutter, S. L., & Finch, C. (2008). Temporal and spatial changes structure. In J. H. Ausubel & R. Herman (Eds.), Cities and
in social vulnerability to natural hazards. Proceedings of their vital systems. Washington, DC: National Academy
the National Academy of Sciences, 108(1), 2301-2306. Press.

Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms
 GeoJournal (2013) 78:287-301 301

Matisziw, T. C., Murray,

Pederson, P., Dudenhoeffer, D., A. T.,
Hartley, S., &
& Permann, M. Gr
Exploring the vulnerability of network
(2006). Critical infrastructure interdependency modeling:
disruption. Annals
A surveyof Regional
of U.S. and Scie
international research. Idaho
s00168-008-0235-x. National Laboratory. URL: http://tinyurl.com/6fzweo.
Matisziw, T. C., Murray, A. T., & Grubesic, T. H. (2010).
Perrow, C. (1999). Normal accidents: Living with high-risk
Strategic network restoration. Networks and Spatial Eco- technologies. Princeton, NJ: Princeton University Press.
nomics, 70(3), 345-361. Reid, R. L. (2008). The infrastructure crisis. Civil Engineering.
McDaniels, T., Chang, S., Peterson, K., Mikawoz, J., & Reed, January.
Richardson, H. W., Gordon, P., & Moore II, J. E. (2006). The
D. (2007). Empirical framework for characterizing infra-
structure failure interdependencies. Journal of Infra- economic impacts of terrorist attacks. Northampton, MA:
structure Systems , 13(3), 175-184. Edward Elgar.
Merriam-Webster Dictionary (MW). (2008a). Vulnerability. Rinaldi, S. M., Peerenboom, J. P., & Kelly, T. K. (2001).
URL: http://www.merriam-webster.com/. Identifying, understanding and analyzing critical infra-
Merriam-Webster Dictionary (MW). (2008b). Decay. URL: structure interdependencies. IEEE Control Systems Mag-
http://www.merriam-webster.com/. azine. December 11-25.
Merriam-Webster Dictionary (MW). (2008c). Capacity. URL: Sage, A. P., & White, E. B. (1980). Methodologies for risk and
http://www.merriam-webster.com/. hazard assessment: A survey and status report. IEEE
Minkel, J. R. (2008). The 2003 Northeast blackout - Five years Transactions on Systems, Man and Cybernetics, 70(8),
later. Scientific American . URL: http://www. scientific 425-446.
american.com/article.cfm?id=2003-blackout-five-years- Sarewitz, D., Pielke, R., & Keykhah, M. (2003). Vulnerability
later. and risk: Some thoughts from a political and policy per-
Murray, A. T., & Grubesic, T. H. (2007). Critical infrastruc- spective. Risk Analysis, 23(4), 805-810.
ture: Reliability and vulnerability. Berlin: Springer. Schmidtlein, M. C., Deutsch, R. C., Piegorsch, W. W., &
Murray, A. T., Matisziw, T. C., & Grubesic, T. H. (2007). Cutter, S. L. (2008). A sensitivity analysis of the social
Critical network infrastructure analysis: Interdiction and vulnerability index. Risk Analysis, 28(4), 1099-1114.
system flow. Journal of Geographical Systems, 9(2), Tighe, S., Li, N., Falls, L. C., & Haas, R. (2000). Incorporating
103-117. road safety into pavement management. Transportation
Murray, A. T., Matisziw, T. C., & Grubesic, T. H. (2008). A Research Record, 1699 , 1-10. doi: 10.3 141/1699-01 .
methodological overview of network vulnerability analy- Transportation Research Board. (2005). Inspection and man-
sis. Growth and Change, 39( 4), 573-592. agement of bridges with fracture-critical details. URL:
Murray, C. J. (2007). Fatigue could loom large in bridge col- http://onlinepubs.trb.org/Onlinepubs/nchrp/nchrp_syn_354.
lapse. Design News. URL: http://tinyurl.com/5jclc8. pdf.
Nagle, J. (1984). Congestion control in IP/TCP internetworks.
U.S.-Canada Power System Outage Task Force (USCPSOTF).
Computer Communication Review, 14(4), 61-65. (2004). URL: https://reports.energy.gov/.
National Research Council. (2006). Terrorism and the chemi-United States Geological Survey (USGS). (2008). URL:
cal infrastructure : Protecting people and reducing vul- http://earthquake.usgs.gov/.
nerabilities. Washington, DC: National Academies Press.United States House of Representatives. (2006). A failure of
National Transportation Safety Board. (2004). CSX freight initiative: The final report of the select Bipartisan Com-
train derailment and subsequent fire in the Howard street mittee to investigate the preparation for and response to
tunnel in Baltimore, Maryland, on July 18, 2001. URL: Hurricane Katrina. URL: http://katrina.house.gov/full_
http://www.ntsb.gov/publictn/2004/RAB0408.htm. katrina_report.htm.
National Transportation Safety Board. (2008). Transportation
White House. (2003). The national strategy for the physical
for tomorrow: Report of the National Surface Transpor- protection of critical infrastructures and key assets. URL:
tation Policy and Revenue Study Commission. URL: http://www.whitehouse.gov/pcipb/physical.html.
http://www.transportationfortomorrow.org/final_report/. Willis, H. L. (2002). Spatial electric load forecasting. New
Parfomak, P. W. (2005). Vulnerability of concentrated infra- York: CRC Press.
structure: Background and policy options. CRS Report for
Congress. URL: http://www.hsdl.org/?view&doc=55205
&coll=limited.

Ô Springer

This content downloaded from 14.139.58.194 on Sun, 04 Mar 2018 04:21:59 UTC
All use subject to http://about.jstor.org/terms