From W3C Wiki
Jump to: navigation, search

Disaster Management

The United Nations defines a disaster as: "A serious disruption of the functioning of a community or a society causing widespread human, material, economic or environmental losses which exceed the ability of the affected community or society to cope using its own resources." (from the UN International Strategy for Disaster Reduction)

A disaster is defined and 'declared' when local resources are overwhelmed; when more resources are required than those immediately available. A disaster simply defines the point of escalation where outside help is required: 9/11 was a disaster for the city of New York within minutes of occurrence. As moving resources to overwhelmed areas is a management problem, the term disaster management has come into use to describe larger scale processes of disaster relief and disaster recovery as opposed to emergency management of the more routine sort (see 'what to call this?' below and Naming Survey). Some disasters, such as a pandemic, may preclude any help arriving "from outside", as there may "be no outside" since many areas are affected at once. For this and other reasons, resilience rather than after-the-fact relief has become the primary goal of many agencies.

Sub Sections

Latest news

2 September 2007: Created sub-sections for W3C WIKI to better structure the capture of knowledge, discussions and decisions in the XG

17 August 2007: Awaiting the inclusion of the new W3C RF policy for XGs to include more sponsors and start the XG officially

27 June 2007: Comments and feedback on the Draft XG Charter: Media:DisasterManagement$eiif-charter-02.pdf

6 June 2007: Contribute to the creation of a Charter for a Proposed W3C Incubator Group to drive the work of the group (Thanks - now closed)

5 June 2007: Join the W3 email discussion list

What W3C can do

Emergency management is a data-intensive activity which requires evaluation of many forms of data, and often consideration of combinations of data which have not been combined before. By definition, disasters involve local emergency response capacity being exceeded and relief and resilience processes called on to provide short- and long-term aid. They also involve or initiate recovery and reconstruction effort which includes the upgrading of local resilience and responses systems to avoid future problems: data gathered in any crisis, whether or not it is a disaster, is critical to avoiding or reducing the next one of its kind.

While many agencies offer standards in this area the W3C's role is pivotal, as semantic web standards (RDF in particular) enables syllogistic reasoning on simple factual constructs - which while less reliable than expert-compiled information has the merit of being instantly available. Since much of the data required in a disaster is about where/when/who, a rigorous ontology dealing well with typical relief tasks and roles would be instantly useful.

This page exists to coordinate and track related efforts, and identify any other ontology and semantic tag/link work taking place in local resilience, emergency management, relief and response, biosecurity, peacekeeping, etc..


presently being debated on the mailing list associated with this incubator.

An ideal emergency management ontology describes the following critical steps:

  • once the emergency is widely anticipated, sharing of data describing response and resource characteristics that are needed in this kind of event, e.g. databases of vulnerable persons and places, ensuring warnings take place
  • as the crisis unfolds, gathering of data on its scope and emerging effects
  • as the response begins, gathering of data on its outages and missing links (including communication and data systems that fail or are incompatible with the standard ontology and must be replaced) and matching with relief capacity
  • as the response by first responders is overwhelmed, sharing relief requests to prioritize relieving the first responders who are most overloaded or tired.
  • as the relief unfolds, gathering and integrating data from all responders to build a common baseline map of the situation and facilitate probes and first attempts at pro-active data gathering (e.g. MSF lead logistician role)
  • characterizing problem states as chaotic (no baseline and no reliable map), complex (changing too fast to identify causes, requires probes) or manageable
  • rapidly deploying compatible information and communication systems to local authorities and institutions capable of dealing with the manageable situations
  • calling for expert review of action proposals to limit/contain chaotic situations, and mass peer review of probes that better define complex ones, with intent to limit the unanticipated side effects of management decisions
  • comparing predicted to measured effects of interventions within 48-72 hours
  • identifying situations which are not improving and calling for more options (in chaotic or complex situations) or more resources (for manageable but large scale problems)
  • helping experienced response teams move on to the more complex situation by facilitating rapid handoff and just-in-time training of those less experienced
  • guiding recovery and reconstruction efforts by identifying those outages or problems that most inhibit the resilience networks and outside relief efforts
  • guiding resilience efforts by identifying which prevention and anticipation options (e.g. evacuation) could have prevented the most morbidity or loss of life-sustaining infrastructure, e.g. weather-resilient homes, levees
  • passing off all data gathered in the disaster to the appropriate authority after the crisis passes, updating databases of vulnerable persons and places

Making it work

While some forms of disaster, such as hurricanes, take a well-known form, and others, such as terrorist attacks, can be very novel an unexpected, the essence of a disaster is its unexpected nature. Dealing with it will involve combining data instantly in unplanned ways, which is a core strength of Semantic Web technology. Local conditions, whether it is weather, chemical, or military; reports from the public and from relief organizations, and government data flows may all provide information which must be instantly integratable for judgments to be made in the field.

Furthermore, disaster management cannot be an isolated application. It must quickly be aware of information from many sources, and also must feed information externally. The world looks at a disaster, often willing to help, but needing an accurate picture. Possible funders, including, increasingly, the general public, need to see what is necessary. Relief organizations need an accurate on the ground picture so that resources are sent to the right place and are not wasted. Families of those involved need survivor information. For chemical or biological events, scientists across the globe with expertise in a particular field may need to apply themselves to the problem at short notice.

These requirements suggest that humanity's effectiveness will be much increased if relevant data streams are made available. John Hardin of OASIS, a similarly difficult effort to compile electronic business "patterns", says this kind of "information integration won't happen unless:

a) the path to implementation of standards based messaging between partner, or producer/consumer, applications, is very clear

b) all players involved in complex processes can participate

c) the integration effort doesn't involve such a long, drawn out mapping effort everytime (this kills progress on the entire movement)"

What's happening

@@ Missing is an overview of government data integration and standardization efforts at different scales: household, emergency response unit, facility, local and regional authority, national, large-scale bioregional, international.

In the blogs, in 2006-12, Soenke Ziesche wrote of "Semantic Wikis and Disaster Relief Operations", and in 2007-02, Tim O'Reilly suggested Peer to Peer Information During Disasters. The term resilience network has come into use (in US official circles including NDRCI) to describe "intelligent social networks used for providing citizens and their communities with prospective best practices, situational awareness, coordinated focus on mission critical gaps," etc., to address "the needs of vulnerable populations incident command systems often do not address and on situations where command and control systems fail" (quote from Michael McDonald)

The Sahana system is an open source disaster management system. It was produced in response to the 2005 Asian Tsunami, and tracks missing and found people, shelters, relief efforts, missing people, aid, volunteers, between Government groups, the civil society (NGOs) and the victims themselves. Sahana has been further developed and used for several disasters, including the 2006 Indonesian Earthquake. However, each implementation of Sahana is a standalone system and it remains poorly integrated with resilience and relief ontologies.

In a disaster there are likely to be many government and non-government bodies involved, using different implementations of different software based systems. Some disasters, such as an Avian Influenza Pandemic, may occur on a large scale and inhibit movement of experts or even local responders, increasing the pressure for data gathered by one group to be verified by another and used by a third.

It would be valuable to be able, for example, to:

  • Maintain lists of vulnerable persons and places that persist from situation to situation
  • Merge missing and found data from different systems in different countries to get a wide picture, including local police and fire responders and service providers like hospitals
  • Coordinate people missing in one system but found elsewhere
  • Connect resource requirements on the ground to existing resources or potential aid; Send out targetted requests for specific types of aid
  • Generate a single map of outages, unmet needs, missing resources or expertise, that can be used to quickly convince or compel them to be met
  • Understand the disaster better by merging specific emergency management data with other local data of all kinds, such as weather data, population demographics, epidemiological and medical data, etc.
  • Anticipate emerging needs based on trends in the incoming disaster data plus any anticipated risk scenarios such as earthquaker aftershocks, downstream floods after heavy rains, or conflicts between groups of victims
  • Quickly deploy a free software and open content based system to localities that had no such automation before, and rapidly train them to keep it up to date for the current disaster and ongoing comprehensive emergency management

and so on.

Next steps

A possibility is the creation of a common Relief or Disaster or Comprehensive Emergency Management ontology effort or a more general (Crisis, Relief and Resilience) Semantics, Ontology, Interoperability and Integrativity Group (name unclear) to carry common terms from many emergency management and disaster management systems. The first such term is the name of the ontology. See Naming Survey.

What to call this?

The term Disaster Management is contentious as it implies that disaster and relief efforts can be viewed as simply an extension of local emergency management. That is, that something called a "disaster" is "manageable" at all. This is explicitly contested by some efforts like the US ABIDE framework. The terms "crisis" and "disaster" are not well defined and may be useful for this reason, but also may be avoided. "Relief", "response", "resilience" "emergency management" are less contentious.

Core data

Different types of disaster, and systems used by different types of organization, will of course use different sorts of data, but a common core set may be shared throughout. The default configuration of the Sahana system has been proposed as a seed ontology, as it stems from a practical common system which has been deployed and is using a relatively open mediawiki for specification and requirements analysis. However, it is extremely poorly integrated with existing W3 standards, has and should probably undergo an exhaustive review first to find differences in geospatial and personal data descriptions and choose authorities to defer to. For instance, the Interpol specification for a disaster victim provides a template for local and national authorities to apply to other victim situations.

The UN cluster approach

The essential questions to get an overview of a disaster response (and potential gaps and overlaps) are typically "Who is doing what where?" and "What is needed where?". The UN Office for the Coordination of Humanitarian Affairs has developed a system which comprises for who/what/where the following sub-items, which could be incorporated in an ontology:


  • Organization Name
  • Implementing Partner
  • Donor
  • Contact Person
  • Phone no
  • Email


  • Cluster
  • Project ID
  • Project Title
  • Project Description
  • Project Type
  • Project Status
  • Project Start Date
  • Project End Date
  • Primary Beneficiary Type
  • No of Primary Beneficiary
  • Secondary Beneficiary Type
  • No of Secondary Beneficiary
  • No of Projects
  • Funding Amount (1,000$)
  • Funding Status


  • State
  • State- P Code
  • Locality/County
  • Locality/County- P Code
  • Place Name
  • Place Name P Code

The items cluster (under "what") and P codes (under "where") need further elaboration:

The cluster approach has been established by the UN Inter-Agency Standing Committee in order to "to improve the predictability, timeliness, and effectiveness of humanitarian response, and pave the way for recovery". The following nine clusters have been introduced:

  • Camp Coordination and Management
  • Early Recovery
  • Emergency Shelter
  • Emergency Telecommunications
  • Health
  • Logistics
  • Nutrition
  • Protection
  • Water and Sanitation


GPS is not universally used by all participants in a response. For locations the UN recommends the P-code system, which is similar to a common zip code system and is usually introduced by UN Humanitarian Information Centres in disaster areas, see here for Pakistan. P-codes provide unique reference codes to locations while the use of the names of places as identifiers can easily lead to confusion over spelling, different languages or scripts as well as duplication.

Obviously, GPS coordinates that define the borders and approximate centres of the P-codes are available and given GPS data it is relatively easy to discover which P-code that coordinate pair is in, which direction to move in which to get closer to some known P-code, etc. The functioning roads, depots, and utility features of an area can be mapped as required and become part of a permanent record of the features of the P-codes.

Mapping to codes or addresses or other systems used by the authorities can also be made automatic, if these are well-documented.

W3C Incubator

A W3C Incubator Group is now forming to pursue coordinated transparent research and development, which would get it persistent ontology storage, mail list archives, and moderation standards fit for experienced and focused architects (who expect RFC2119 language to be followed and subject lines to match the subject, and change when the subject changes) rather than social discourse among programmers who resent standardization or architecture.

Initial steps towards such an Incubator have been taken. Contact Ivan Herman or, for the administrative issues on setting up an Incubator Group, Steve Bratt

Other resources

  • U. N. ICLEI Resilient Cities and Communities Initiative which various UN agencies support
  • *a) up-to-date assessments of the overall needs for human, financial, and institutional capacity;
  • *b) reviews of currently available capacities and means for their use;
  • *c) links with other clusters, including preparedness and long-term planning, standards, best practice, advocacy, and resource mobilisation;
  • *d) taking action to ensure that required capacities and mechanisms exist, including rosters for surge capacity and stockpiles; and
  • *e) training and system development at the local, national, regional, and international levels. Designated Global Cluster Leads are accountable to the Emergency Relief Coordinator (ERC) for ensuring predictable and effective inter-agency preparedness and response within the concerned sectors or areas of activity." They act as provider of last resort for the life-critical services within that cluster. Each appoints a field-level cluster lead to supervise provision of those services for a given area or camp.
  • U. N. summary of 2006 achievements of the cluster approach including many detailed pointers to further work and many related standards
  • U. K. UK resilience publications (official) and [http:www.hertsdirect.org/infobase/docs/pdfstore/HRERPpub.pdf examples of published emergency management or "resilience" plans]
  • N. Z. From response to resilience: emergency management reform in New Zealand and a related UN paper on "Comprehensive Emergency Management (CEM) [which]covers reduction,readiness, response and recovery" and aims to create resilience:
  • * "Emergency management has the capability to create resilience in communities through its recognition of the risks in the wider hazardscape, and the essential resources required to reduce their consequences. Resilience is also brought about through emergency managementhelping the community choose a level of risk appropriate to the circumstances of the community. In this way, emergency management helps communities manage as sustainable entities, knowing that simply reducing losses from future disasters is too narrow a goal.

Emergency management should also encourage holistic and integrated community management within a wide partnership programme. Few communities, tie hazard reductionand/or emergency management plans into comprehensive community strategic plans.Business-as-usual approaches to risk reduction have focused on saving lives and, to a lesser extent, on disaster recovery. The limited attention paid to risk reduction has encouraged piecemeal resistance to hazards."

  • Federation of Canadian Municipalities policies including Emergency Management: "As first-line responders, municipal governments play a critical role in public safety and emergency preparedness... emergency management systems must be appropriately funded and coordinated across jurisdictions [to deal with] unique hazards associated with sensitive infrastructure, such as airports, ports, borders and embassies [and] distinct threats posed by the new security environment andthe load this places on municipal services, including police, fire, paramedic and public health agencies. Protection of critical infrastructure, such as airports, ports, borders or embassies [require municipalities to] review their emergency response plans and adapt local infrastructure to deal with severe weather, as climate change is widely expected to bring more frequent and more severe extreme weather events, such as floods, hail or wind storms...public health authorities must play an integral role in emergency management... roles and responsibilities of each order of government in national security and emergency management planning should be re-examined."
  • U. S. National Information Exchange Model (NIEM) is actually targeted at emergency and disaster response, it seems. However there are several competing US models including the NDRCI, which focuses on "Psychosocial Dimensions of Biosecurity Preparedness and Response", and on "Resilience Networks", and is studying another model called ABIDE which emphasizes the dangers of "management" in chaotic or complex situations. These exist alongside another credential-focused regime maintained by FEMA, the routine-orienetd Federal Enterprise Arcnitecture (FEA) and the new-biological-threat-oriented LAOS ontology from LANL. The military also maintains its own facility management standards which provide some guidance on resilient, threat-resistant facilities and language. The Journal of Homeland Security is increasingly focused on resilience in like with the UN efforts.
  • OASIS Emergency Management TC Robin Cover's overview; Emergency Data Exchange Language (EXDL XML schema apparently covers inter-agency messaging, geographical region definition, and incident definition for alert distribution. OASIS maintains a pattern catalog that "take into account work done in several standards development organizations including... OASIS...W3, ISO (International Standards Organization, UN/CEFACT (United Nations Centre for Facilitation of Commerce and Trade)" and should be of particular use for the [supply chain] issues explored in Schuldt, e.g. "Natural disaster response team shows up lacking batteries to operate GPS system and walkie-talkie for 200 search and rescue workers – need four hundred 9-volt batteries to even begin the search and rescue effort" which implies finding the 400 nearest such batteries.
  • TOWARDS AN OPEN ONTOLOGY FOR OPEN SOURCE EMERGENCY RESPONSE SYSTEM, paper by Paula di Maio, Mae Fa Luang University, Chiang Rai, Thailand.
  • The Status of Theory in Emergency Management, 2004
  • Semantic Web Technology for Public Health Situation Awareness at the University of Texas, with Oracle and TopQuadrant.
  • ICOR and other business continuity or organizational resilience literature contains a vast array of definitions and tactics and distinctions useful to any institution, such as a hospital or government office, that seeks to continue to operate in and through a crisis or disaster situation. Arguably, any disaster management standard is futile if it doesn't cleanly mesh with the local authorities and institutions that continue to operate before, during and after the external interventions. To the degree that such authorities and institutions rely on resilience approaches and standards oriented towards private sector and one-institution "business continuity" needs, these need to be properly integrated.
  • Soenke's article on semantic web use in emergencies

See also the list of Resources and Emergency Management Definitions which elaborate on the above.