This Alerts, Warnings, and Notifications (AWN) description is intended to inform and connect three various sets of stakeholders: Public safety officials (e.g., emergency managers, alerting authorities/originators); Hazard modeling agencies, whose technologies are sources of critical information; and the Geospatial community, the experts in making available information that is specific for targeted geographic locations.
Alerts, Warnings and Notifications (AWN, pronounced “a-win”) are a critical source of near real-time information that helps first responders secure and protect people and property before rapidly approaching hazards, during actual emergency incidents, or in their aftermath. AWNs are often dispatched with follow-on instructions, which can provide relevant information for the disaster’s response and recovery efforts. AWNs are being employed more and more frequently by Federal agencies, State and Local agencies, territorial and tribal communities (as well as universities, national laboratories, transit and transportation departments, etc.) to provide key information to people in an efficient and timely manner. AWNs include geospatial information, which provides geographic context as well as streamlined delivery to a targeted and at-risk population1.
The objective of this context is to encourage conversation and enhance cooperation among stakeholders, so that more accurate information about a variety of hazards may be contained within AWNs, beyond the well-established weather related events or AMBER Alerts2. The public would benefit from AWNs that relay information about hazardous material incidents, landslides and wildfires. For example, research has shown that for short, 90-character WEA messages, highly informative maps improve response outcomes3. With a greater demand for spatial context information, the geospatial community will be able to expand its mission, portfolios, and overall reach, and deepen its understanding of how the impact of emergency incidents (real or anticipated) is communicated to the public.
Public alerting is an imprecise science. The issuance of AWNs requires vigilance, modeling expertise, familiarity with dissemination mechanisms and alerting capabilities, and actual alerting authorization or mandates (which vary among alert originators). With the exception of the National Weather Service, few organizations mature enough in these capability areas to tailor AWNs to a broader suite of hazards. Fortunately, we can extrapolate the key factors and components that will allow AWNs to be dispatched across more hazards by looking at specific models used in the Early Warning community.
The Early Warning community is comprised of a variety of stakeholders, including science-focused observer organizations like the U. S. Geological Survey, which has networks of sensors such as stream gauges and seismometers) as well as other organizations that not only detect the early signs of a hazardous condition, but also model the potential consequences of a disaster if the condition develops. When information about various types of hazardous risks are more accurately defined, first responders and government decision makers can more easily provide guidance to the public—through alerts and warnings—that contain accurate information, enabling them to better prepare and avoid harm’s way to a greater extent.
Alerts, Warnings, and Notifications (AWN) typically contain information about a particular hazard such as its intensity, a timeframe for when the hazardous condition is expected to occur (and diminish), the locations likely to feel the impacts of the hazard; sometimes, a recommended action is included that recommends that the recipient take action to protect his or her life and/or property. Most people are familiar with AWNs issued for potential severe weather and can likely recall receiving an alert from the National Weather Service for a tornado or flash flood on their mobile phones or, in years past, on the television or by radio. In addition to Federal, State and Local government entities, many corporations, school systems and universities, and other commercial entities, employ mass notification systems to rapidly convey information on a range of issues such as weather, street closures, neighborhood watch information (e.g., crime), and any event that might disrupt normal schedules.
Within this text, Alerts are defined as a message to the recipient that something significant has happened or may happen. It is intended to make the recipient aware of an increased risk and to monitor the situation. Warnings typically follow alerts, providing more detailed information describing an event, and include some indication of any protective action that should be taken by the recipient. Notifications often follow the occurrence of an event and provide clarity to the situation, but are of lesser urgency than alerts or warnings4.
AWNs may be communicated through a range of technologies (sometimes called dissemination paths) including, but not limited to, emails, text messages, cellular telephone networks, paging systems, radio, TV, social media, sirens and interactive signage, smart phone applications, and internet applications and websites. Each of these technologies has its strengths and weaknesses, which explains why multiple systems are used for redundancy, to make sure all people in a targeted AWN area receive the message. This provides a greater assurance that the intended audience receives the information, in case one of the dissemination paths fails.
Within the message itself, the timing of the expected event (and a timely dispatch of instructions related to it), as well as location, are two of the most critical elements. Within the message content, the inclusion or exclusion of time and location information can greatly enhance or diminish the effectiveness of an AWN. Ideally, once an alert is received by those at risk (in near-real time), attention turns from normal, day-to-day activities, to the unfolding or anticipated incident or situation. Letting people know where the potential threat is, or will be, relative to their current location, provides a more granular level of situational awareness, enabling people to make more informed decisions and for the response to be more effective. In addition to time and location elements, effective alert messages should also include words in order of their relative importance within the message. For example, Source of the incident information should be presented first, followed by Guidance on what to do (e.g., protective action), followed by the Type of Hazard, and the expected Location and Time5. The goal of an effective message is for information to be conveyed with a high level of clarity and delivered across as many dissemination paths as possible to maximize receipt of the message. For example, Wireless Emergency Alert (WEA) messages are limited currently to 90 characters which does not leave a lot of room to include additional information. Radio and television dissemination paths supplement the information provided in WEA messages as well as other short-format messaging technologies.
THE AWN COMMUNITY MODEL
The AWN community consists of key stakeholders that are grouped into roles relevant to sending AWNs. This community includes practitioners of public safety disciplines, such as emergency management, law enforcement, and fire safety; computational modeling and information technology and architecture disciplines; governance and policy experts, who help to ensure accountability during an incident; and additional practitioners in technical and data exchange standards who provide the appropriate structures for sending AWNs. The purpose of the AWN community model (CM) is to provide a high-level view of stakeholder interactions related to the creation and dissemination of AWNS for natural and manmade disasters. The AWN community model enables stakeholders to see how each stakeholder group fits into the overall AWN community and aligns with the GeoCONOPS community model, so that the AWN information environment, actor roles and responsibilities, and high-level processes around information exchange may be identified.
MONITORING AND EARLY WARNING
Within this context, the Monitoring and Early Warning community is comprised of a variety of stakeholders. These stakeholders include: Scientific and engineering organizations such as the U. S. Geological Survey and other support organizations that focus on developing sensors and networks on natural and technological hazards; Public safety organizations that model the potential consequences of a disaster and analyze and define hazard risk areas for inclusion in alerts and targeted dissemination (a.k.a. geotargeting). These various groups employ mechanisms that sense, record, analyze, model, and communicate data to provide the earliest possible insights into a potentially hazardous event and convey that information with speed and confidence to alert originators for dissemination to appropriate audiences.
The Early Warning community includes authoritative organizations such as United States Geological Survey (USGS), Health and Human Services (HHS), and Interagency Modeling and Atmospheric Assessment Center (IMAAC), among others. The community also relies on researchers in academia, government and the private sector for exploring innovative ways to gain distant early warning, improve confidence in results and expand the tools into the study of additional hazard types.
Sensor Networks and Hazard Detection and Modeling are additional methods and tools used collectively to evaluate potential threats. For example, sensors are used to record a type of data such as ground movement, which may be caused by a non-hazard event such as a book dropping to the floor nearby, or an impending earthquake some distance away. This sensor data becomes more meaningful when it is combined with data from a network of sensors to indicate a possible hazard. The sensor data may then be used to model scenarios and predict outcomes, creating information that can be included in the alert message that will be compiled and disseminated to the public.
If and when the collected sensor data and subsequent modeling exercises indicate that a hazard has been detected, and that a threshold has been reached that may cause harm to life and property, the information is then checked against defined standards and guidelines (represented in the community model as the Green Policy and Guidance Filter) for threat data used to issue an AWN. Not all sensor data will meet thresholds of criteria that indicate—through policy and guidance—that the public needs to be made aware. In this case, AWNs may not be issued and the data and information remains at the monitoring and early warning or modeling assessment area. It may be re-evaluated if a more qualified determination is required.
Governance within the AWN community implies that a layer of standards and control (policy) are applied to the output from the Early Warning community to ensure that information received from the Early Warning community (and its technologies) flows into the dissemination infrastructure. For data to be used effectively and consistently, a governance structure must be in place to ensure quality and availability of the data so that precious time is not wasted interpreting data during an emergency. Governance stakeholders in the AWN community include Federal entities (e.g. National Oceanic and Atmospheric Administration (NOAA), United States Geological Survey (USGS), Federal Emergency Management Agency (FEMA), Department of Interior (DOI)); Non-governmental organizations (NGO) such as NEMA, NAPSG, and NSGIC, and State level, Local, Tribal and Territorial governments and organizations (SLTT); and private sector entities. Governance also includes State and enterprise level Emergency Operations Centers (EOC) and Executive Leadership. The private sector provides the infrastructure (networks, transmitters, stations and devices) that are utilized to deliver AWNs to the public; private sector capabilities, limitations and use policy also impacts governance.
All of these AWN community stakeholders are accountable for ensuring that current policy is implemented effectively and that new policy reflects the increasing capabilities and integration of AWN community stakeholders. At its core, the AWN Community relies on policy and mandates to define expectations, requirements, and workflows between stakeholder organizations that work in concert to deliver this essential information. To be a reliable partner in an alerting structure, an organization must have the proper authorities and legal foundation to ensure it is funded, empowered, and held accountable for its contributions. At the Federal level, policy driving the AWN community is defined across government-funded programs in the Department of Homeland Security (e.g., FEMA IPAWS, DHS NTAS, DHS S&T), Department of Commerce (NOAA, NIST), Department of Justice (FBI, NCMEC AMBER), Public Laws and Regulations such as the WARN Act, Stafford Act, Federal Code of Regulations (47 Part 10 and Part 11), Executive Orders (i.e. EO 13407) and Presidential Policy Directives such as PPD-8. Each state, tribe, territory and community will have its own legal framework which must be considered when collaborating to improve AWN dissemination.
Within the Messaging portion of the AWN community model, the raw scientific information previously collected by Early Warning stakeholders is joined with other contextual information and geotargeted to a specific population at risk. Messaging technology includes a segment of software known as Alert Origination Service Providers (AOSP), which are systems and software that alerting authorities use to generate compliant messages and interface to systems that disseminate AWNs. AOSPs deliver messages to either a single dissemination system or to an aggregator for dissemination to a range of dissemination systems. For example, FEMA’s Integrated Public Alert and Warning System (IPAWS) is an alert aggregator capability that Federal, State, Local, tribal and territorial authorities can use to disseminate critical public alerts and warnings to cellular phones as Wireless Emergency Alerts (WEA), via radio and television as Emergency Alert System (EAS) broadcasts, to NOAA All-Hazards weather radios, and to an All Hazards Alert and Information Feed for internet applications, services, and websites in specific jurisdictions. Alerting Authorities use IPAWS through alert origination software or systems that meets IPAWS system requirements. There is no cost to send messages through IPAWS, although there may be costs associated with acquiring compatible alert origination software6.
Alert dissemination technology includes all networks and devices used to purvey messages composed by alerting agencies to the general public to provide public warning and information to enhance personal safety and property protection. Examples of dissemination technology include cellular networks that deliver Wireless Emergency Alert (WEA) to mobile phones, radio, television, and cable stations that compose the Emergency Alert System, telephone networks, internet applications and websites, audible siren systems, NOAA’s All-Hazards Weather Radios, and other communications technologies.
LEVERAGING GEOSPATIAL INFORMATION
AWNs must be able to promptly and reliably convey the current or likely future location of a hazard in order to allow adequate and effective protection of life and property. This is why defining the hazardous location area with increased accuracy is important—for people inside the at-risk area and outside the at-risk area—so they may avoid moving into the hazardous area unknowingly. First responder organizations could also use this information to pre-position response capabilities to allow immediate access to the impacted area or to plan assessments immediately following the event.
AWNs must leverage geospatial information at least twice in provisioning messages. First, the potential hazard location must be clearly described within the message delivered to the public according to one of the methods outlined below. Second, the people within the at-risk area must be “geo-targeted” using a method compatible with the messaging technology. For example, this can include using wireless network technology to selectively send AWNs only from cell towers in a given area, using location-aware devices that only alert if their location is inside the geo-targeted alert area, or only triggering sirens, radios, TVs, etc. in a specific geographic area. This appendix focuses primarily on the first category. For a full discussion on the second category, please refer to Geotargeted Alerts and Warnings from the National Research Council (NRC 2013). In order to communicate Locations at Risk within a message, organizations may use one or more of the following methods:
Common Place Names – Town county or community names that are familiar to the people in the at-risk area
Encoded Geographies – Standardized geographic locations that have corresponding unique identifying codes to ensure they remain unambiguous and are interpretable by computer systems, and can be efficiently communicated via low bandwidth connections. This includes 5-digit zip and Federal Information Processing Standard (FIPS) codes.
Latitude/Longitude Coordinates – Common coordinates used in mapping and display systems. Can be individual points or combined to define areas.
Prescribed Hazard Areas – Areas of known hazard exposure that can be delineated on maps, pre-defined within the community through signs and made part of educational programs.
Hazard Templates – Tools that reflect the predictable behavior of a hazard and can be employed to define the hazard area based on specific situational parameters such as magnitude and weather conditions.
Regardless of the method employed, geographic areas at immediate risk of hazard need to be defined quickly, confidently, and communicated clearly to help ensure public safety. Imperfection in precision geotargeting is frequently traded for levels of “over-alerting” to gain speed of message delivery and/or higher assurance that intended area receives the message, which is why geographic alerting areas tend to be larger and less precise than typical model outputs. Geospatial tools also contribute to the AWN community in a number of other significant ways that are still evolving in their application, but hold significant promise when fully developed and integrated. Some of these include geofencing, probabilistic and deterministic modeling, remote sensing, and field data collection.
Commonly, both people and hazards are in motion, often unpredictably, resulting in a complex scenario, especially for alerting people in motion who may travel into and out of a hazard area. A geofence is a virtual boundary encompassing a geographic area within a geographic information system (GIS). A geofence around the potentially-impacted hazard region means that anyone with location-aware devices can either display or ignore the AWN message as he/she moves inside or outside of the geofence. Geofences may cover a wider area to alert the person before they encounter the hazard directly. Geofencing is also a key tool in the growing industry of location-based services, and is often used in marketing directly to those in proximity of a vendor. This tie to commercial growth suggests it will advance quickly and be increasingly available for emergency management applications as well.
AWNs may also leverage probabilistic and deterministic modeling methods in order to better prepare for future disaster events (e.g. pre-designation of evacuation routes and Tsunami safe areas). Probabilistic modeling employs statistics gathered in past events to predict the likelihood of future events across the area studied. This can be used to identify areas for more detailed study, siting of alerting infrastructure (i.e. sirens), and for educational outreach planning. An example of this would be the assessment of the Gulf Coast to identify the likelihood of a category 3 or above hurricane striking in any given year. Determinative models explore the potential impact of a specific event or hypothetical scenario on a location based upon observed, anticipated, historical or invented parameters. An example of a determinative model would be assessing the impact that a historic storm (like Hurricane Katrina) strikes Mobile Bay, and planning for worst case scenarios, including the identification of evacuation zones. Deterministic models are frequently used to define hazard areas associated with floods and weather events, but they have considerable untapped potential for use in other hazards, such as wildland fire and hazardous material releases.
The actual geographic location, intensity, magnitude, and impacts of a hazard event can be defined by field observations, in-situ sensors (i.e. weather stations), and overhead (satelliteaerialUAV) imagery. Following an event, the impact area can be determined by direct observation of the extent of damages by field teams such as mapping high water marks from a flood with hand-held GPS units. Remotely sensed imagery from satellites and aircraft can capture and delineate high water and other hazards. In addition, imagery is increasingly used to collect an overhead perspective of impacts that can provide a record of the event for purposes such as recovery planning, expediting the insurance payment and claims process, for recovering loans, and for fraud monitoring.
These types of studies are very precisely delineated and accurate, but they can only be completed after the event has run its course, and they must be done quickly, since the recovery efforts of the victims and nature itself can remove the evidence. For these reasons, while these studies are not directly useful in the early warning of an observed event, they are helpful in after-action studies and model validation efforts than can improve future alerting techniques.
When people understand risks and proper courses of action, they are better able to execute disaster plans, make more informed, rational decisions and take action to protect themselves and their property. Geospatial information provides the critical element of “where” and identification and understanding of “what matters” to all aspects of a disaster and, therefore, is an essential element of disaster preparedness and protection of lives and property. The quality of the geographic information conveyed in AWNs is important.
EXAMPLES OF AWNS
Though the alerting discipline is still developing, a number of well-known sources of AWNs currently exist and are likely to be recognized by the general public. Weather alerts are common, as NOAA’s National Weather Service (NWS) has capabilities that span hazard monitoring, hazard modeling, and the capability (and mandate) to issue weather related alert and warnings. NOAA has been using IPAWS to send tornado or flash flood warnings to mobile devices since June 2012. WEAs appear as 90 character messages on the home screen of capable cellular phones. NOAA has sent over 12,000 warnings via IPAWS as WEA messages, Flash Flood Warnings being the most frequent. An example of a NOAA Flash Flood Warning received as a WEA is depicted below:
Many people have seen an AMBER Alert, or received notification of a child abduction in a region. AMBER Alerts are sent by State AMBER Coordinators, typically in the state police agency, in collaboration with the National Center for Missing and Exploited Children (NCMEC). AMBER Alerts are typically only sent as WEA when authorities have specific identifying information about the victim or the potential abductor such as a car description and license plate. The NCMEC has attributed the recovery of 19 children directly to WEA messages received by citizens.
Earthquake Early Warning (EEW) systems will use a network of monitoring systems and the science of earthquakes to alert people via their devices when shaking waves are detected. Depending on a person’s location relative to the epicenter of an earthquake, the EEW may provide anywhere from seconds to minutes of lead time to enable protective actions.
The Canada-U.S. Enhanced Resiliency Experiment (CAUSE) is a unique example in the context of this Appendix because it is not an alerting system or AWN by definition, but rather a partnership between countries to perform a series of experiments that evaluate technologies and processes enabling cross-border information sharing. The IPAWS is used to share CAP messages between the US and Canadian authorities and systems. The CAUSE exercises have demonstrated cross border alert and incident information sharing and interoperability between IPAWS and a similar Canadian system known as the Multi Agency Situational Awareness System (MASAS). For instance, in the scenario where a strong storm is predicted to make landfall in the Northeastern U.S. and Nova Scotia, Alerts and Warnings information generated by U.S.-based emergency management agencies would be shared with Canadian emergency management agencies to further inform the AWNs issued to the Canadian public. Cross-border information-sharing benefits all involved by eliminating redundant data collection, identifying patterns occurring in the hazard event, and having the ability to disseminate warnings with even more advance notice.
THE FUTURE OF AWN
The goal for this work is to encourage conversations and enhance cooperation among stakeholders. This is important, so that more accurate information about a variety of hazards may be established and contained within AWNs, beyond the well-established weather related events, such as hazardous material incidents, landslides and wildfire. It is anticipated that the demand for more accurate information will result in closer, successful cooperation among hazard modelers, the emergency management community and the geospatial community, and that in the not too distant future, the number of hazards that are modeled and mapped for alerting purposes is expanded to significantly enhance public safety. The ultimate goal is to create a more cohesive and integrated discipline among people who are responsible for monitoring and responding to emergencies, and communicating with the public. As the science of creating alerts areas continues to expand, so will the content, examples, and recommendations described in future versions of the AWN section within the GeoCONOPS.
Geotargeted Alerts and Warnings, National Research Council, 2013: http://www.nap.edu/openbook.php?record_id=18414&page=R1
1 All AWNs communicated via the FEMA Integrated Public Alert and Warning System (IPAWS) have geospatial information in accordance with the IPAWS specification in the Common Alerting Protocol (CAP) standard. However, not all of the messaging systems that communicate the AWN message content to the public communicate the geospatial information. This means that while WEA messages contain geospatial information (as per the CAP), it is not always forwarded to individual cell phones.
2 The National Center for Missing and Exploited Children (NCMEC) issues child abduction emergency alerts, called AMBER Alerts.
3 University of Maryland/START Center of Excellence, Comprehensive Testing of Imminent Threat Public Messages: Updated Findings August 2015. The report also makes clear that at the beneficial effects of maps for longer WEA messages may disappear, and that maps should not be added to WEA messages without further research.
4 Source for definition of Alert and Warning: National Research Council (NRC).
5 Ibid, University of Maryland/START Center of Excellence report, August 2015.
6 For FEMA IPAWS information, see http://www.fema.gov/integrated-public-alert-warning-system. For Alert Authority information, see http://www.fema.gov/integrated-public-alert-warning-system-authorities.