key: cord-021081-yqu1ykc9
authors: nan
title: Early Warning Systems A State of the Art Analysis and Future Directions
date: 2012-11-02
journal: nan
DOI: 10.1016/j.envdev.2012.09.004
sha: 
doc_id: 21081
cord_uid: yqu1ykc9

nan

United Nations' International Strategy for Disaster Reduction (ISDR), it integrates (United Nations, 2006) :

1. Risk knowledge: Risk assessment provides essential information to set priorities for mitigation and prevention strategies and designing early warning systems. 2. Monitoring and predicting: Systems with monitoring and predicting capabilities provide timely estimates of the potential risk faced by communities, economies and the environment. 3. Disseminating information: Communication systems are needed for delivering warning messages to the potentially affected locations to alert local and regional governmental agencies. The messages need to be reliable, synthetic and simple to be understood by authorities and the public. 4. Response: Coordination, good governance and appropriate action plans are key points in effective early warning. Likewise, public awareness and education are critical aspects of disaster mitigation.

Failure of any part of the system will imply failure of the whole system. For example, accurate warnings will have no impact if the population is not prepared or if the alerts are received but not disseminated by the agencies receiving the messages.

The basic idea behind early warning is that the earlier and more accurately we are able to predict short-and long term potential risks associated with natural and human induced hazards, the more likely we will be able to manage and mitigate a disaster's impact on society, economies, and environment.

Environmental hazards can be associated with: ongoing and rapid/sudden-onset threats and slow-onset (or ''creeping'') threats:

1. Ongoing and Rapid/sudden-onset: These include such hazards as follows: accidental oil spills, nuclear plant failures, and chemical plant accidents -such as inadvertent chemical releases to the air or into rivers and water bodies -geological hazards and hydro-meteorological hazards (except droughts). 2. Slow-onset (or ''creeping''): Incremental but long-term and cumulative environmental changes that usually receive little attention in their early phases but which, over time, may cause serious crises. These include such issues as deteriorating air and water quality, soil pollution, acid rain, climate change, desertification processes (including soil erosion and land degradation), drought, ecosystems change, deforestation and forest fragmentation, loss of biodiversity and habitats, nitrogen overloading, radioactive waste, coastal erosion, pressures on living marine resources, rapid and unplanned urban growth, environment and health issues (emerging and re-emerging infectious diseases and links to environmental change), land cover/land changes, and environmental impacts of conflict, among others. Such creeping changes are often left unaddressed as policy-makers choose or need to cope with immediate crises. Eventually, neglected creeping changes may become urgent crises that are more costly to deal with. Slow-onset threats can be classified into location-specific environmental threats, new emerging science and contemporary environmental threats (Table 1) .

Rapid/sudden-onset and slow-onset events will provide different amounts of available warning time. Fig. 1 shows warning times for climatic hazards. Early Warning systems may provide seconds of available warning time for earthquakes to months of warning for droughts, which are the quickest and slowest onset hazards, respectively. Specifically, early warning systems provide tens of seconds of warning for earthquakes, days to hours for volcanic eruptions, and hours for tsunamis. Tornado warnings provide minutes of lead-time for response. Hurricane warning time varies from weeks to hours. The warning time provided by warning systems, increases to years or even decades of leadtime available for slow-onset threats (such as El Niñ o, global warming etc., as shown in Fig. 1 ). Drought warning time is in the range of months to weeks. Slow-onset (or creeping) changes may cause serious problems to environment and society, if preventive measures are not taken when needed. Such creeping environmental changes require effective early warning technologies due to the high potential impact of incremental cumulative changes on society and the environment. (Golnaraghi, 2005) . The graph shows the timeliness of early warning systems for hydrometeorological hazards and the area of impact (by specifying the diameter of the spherical area) for climatic hazards.

Early warning systems help to reduce economic losses and mitigate the number of injuries or deaths from a disaster, by providing information that allows individuals and communities to protect their lives and property. Early warning information empowers people to take action prior to a disaster. If well integrated with risk assessment studies and communication and action plans, early warning systems can lead to substantive benefits.

Effective early warning systems embrace the following aspects: risk analysis; monitoring and predicting location and intensity of the disaster; communicating alerts to authorities and to those potentially affected; and responding to the disaster. The early warning system has to address all aspects. Monitoring and predicting is only one part of the early warning process. This step provides the input information for the early warning process that needs to be disseminated to those whose responsibility is to respond (Fig. 2) . Early warnings may be disseminated to targeted users (local early warning applications) or broadly to communities, regions or to media (regional or global early warning applications). This information gives the possibility of taking action to initiate mitigation or security measures before a catastrophic event occurs. When monitoring and predicting systems are associated with communication systems and response plans, they are considered early warning systems (Glantz, 2003) . Commonly, however, early warning systems lack one or more elements. In fact, a review of existing early warning systems shows that in most cases communication systems and adequate response plans are missing.

To be effective, warnings also must be timely so as to provide enough lead-time for responding; reliable, so that those responsible for responding to the warning will feel confident in taking action; and simple, so as to be understood. Timeliness often conflicts with the desire to have reliable predictions, which become more accurate as more observations are collected from the monitoring system . Thus, there is an inevitable trade-off between the amount of warning time available and the reliability of the predictions provided by the EWS. An initial alert signal may be sent to give the maximum amount of warning time when a minimum level of prediction accuracy has been reached. However, the prediction accuracy for the location and size of the event will continue to improve as more data are collected by the monitoring system part of the EWS network. It must be understood that every prediction, by its very nature, is associated with uncertainty. Because of the uncertainties associated with the predicted parameters that characterize the incoming disaster, it is possible that a wrong decision may be made. Two kinds of wrong decisions may occur : Missed Alarm (or False Negative), when the mitigation action is not taken when it should have been or False Alarm (or False Positive), when the mitigation action is taken when it should not have been.

Finally, the message should communicate the level of uncertainty and expected cost of taking action but also be stated in simple language so as to be understood by those who receive it. Most often, there is a communication gap between EW specialists who use technical and engineering language and the EWS users, who are generally outside of the scientific community. To avoid this, these early warnings need to be reported concisely, in layman's terms and without scientific jargon.

An effective early warning system needs an effective communication system. Early warning communication systems have two main components (Second International Conference on Early Warning (EWCII), 2003):

Communication infrastructure hardware that must be reliable and robust, especially during the disaster; and appropriate and effective interactions among the main actors of the early warning process, such as the scientific community, stakeholders, decision-makers, the public, and the media.

Redundancy of communication systems is essential for disaster management, while emergency power supplies and back-up systems are critical in order to avoid the collapse of communication systems after disasters occur. In addition, to ensure the communication systems operate reliably and effectively during and after a disaster occurs, and to avoid network congestion, frequencies and channels must be reserved and dedicated to disaster relief operations.

Many communication tools are currently available for warning dissemination, such as Short Message Service (SMS) (cellular phone text messaging), e-mail, radio, TV and web service. Information and communication technology (ICT) is a key element in early warning, which plays an important role in disaster communication and disseminating information to organizations in charge of responding to warnings and to the public during and after a disaster (Tubtiang, 2005) .

Today, the decentralization of information and data through the World Wide Web makes it possible for millions of people worldwide to have easy, instantaneous access to a vast amount of diverse online information. This powerful communication medium has spread rapidly to interconnect our world, enabling near-real-time communication and data exchanges worldwide. According to the Internet World Stats database, as of December 2011, global documented Internet usage was 2.3 billion people. Thus, the Internet has become an important medium to access and deliver information worldwide in a very timely fashion.

In addition, remote sensing satellites now provide a continuous stream of data. They are capable of rapidly and effectively detecting hazards, such as transboundary air pollutants, wildfires, deforestation, changes in water levels, and natural hazards. With rapid advances in data collection, analysis, visualization and dissemination, including technologies such as remote sensing, Geographical Information Systems (GIS), web mapping, sensor webs, telecommunications and ever-growing Internet connectivity, it is now feasible to deliver relevant information on a regular basis to a worldwide audience relatively inexpensively. In recent years, commercial companies such as Google, Yahoo, and Microsoft have started incorporating maps and satellite imagery into their products and services, delivering compelling visual images and providing easy tools that everyone can use to add to their geographic knowledge.

EWS: decision making procedure based on cost-benefit analysis.

To improve the performance of EWS, a performance based decision making procedure needs to be based on the expected consequences of taking action, in terms of the probability of a false and missed alarm. An innovative approach sets the threshold based on the acceptable probability of false (missed) alarms, from a cost-benefit analysis .

Consider the case of a EWS decision making strategy based on raising the alarm if a critical severity level, a, is predicted to be exceeded at a site. The decision of whether to activate the alarm or not is based on the predicted severity of the event.

A decision model that takes into account the uncertainty of the prediction and the consequences of taking action will be capable of controlling and reducing the incidence of false and missed alerts. The proposed decision making procedure intends to fill this gap. The EWS will provide the user with a real-time prediction of the severity of the event,ŜðtÞ, and its error, e tot ðtÞ.

During the course of the event, the increase in available data will improve prediction accuracy. The prediction and its uncertainty are updated as more data come in. The actual severity of the event, e tot , is unknown and may be defined by adding the prediction error to the predicted value,Ŝ.

The potential probability of false (missed) alarm is given by the probability of being less (greater) than the critical threshold; it becomes an actual probability of false (missed) alarm if the alarm is (not) raised:

Referring to the principle of maximum entropy (Jaynes, 2003) , the prediction error is modelled by Gaussian distribution, representing the most uninformative distribution possible due to lack of information. Hence, at time t, the actual severity of the event, S, may be modelled with a Gaussian distribution, having mean equal to the predictionŜðtÞ and uncertainty equal to s tot ðtÞ, that is the standard deviation of the prediction error e tot ðtÞ. Eqs. (1) and (2) may be written as : 

where F represents the Gaussian cumulative distribution function. The tolerable level at which mitigation action should be taken can be determined from a cost-benefit analysis by minimizing the cost of taking action:

where C save are the savings due to mitigation actions and C fa is the cost of false alert. Note that the tolerable levels a and b sum up to one which directly exhibits the trade-off between the threshold probabilities that are tolerable for false and missed alarms. The methodology offers an effective approach for decision making under uncertainty focusing on user requirements in terms of reliability and cost of action.

Information is now available in a near-real-time mode from a variety of sources at global and local levels. In the coming years, the multi-scaled global information network will greatly improve thanks to new technological advances that facilitate the global distribution of data and information at all levels. Globalization and rapid communication provides an unprecedented opportunity to catalyse effective action at every level by rapidly providing authorities and the general public with high-quality and scientifically credible information in a timely fashion.

The dissemination of warnings often follows a cascade process, which starts at the international or national level and then moves outwards or downwards in scale to regional and community levels (Twigg, 2003) . Early warnings may activate other early warnings at different authoritative levels, flowing down in responsibility roles, although all are equally necessary for effective early warning.

Standard protocols play a fundamental role in addressing the challenge of effective coordination and data exchange among the actors in the early warning process and it aids in the process for warning communication and dissemination. The Common Alerting Protocol (CAP), Really Simple Syndication (RSS) and Extensible Markup Language (XML) are examples of standard data interchange formats for structured information that can be applied to warning messages for a broad range of information management and warning dissemination systems.

The advantage of standard format alerts is that they are compatible with all information systems, warning systems, media, and most importantly, with new technologies such as web services. CAP, for example, defines a single standard message format for all hazards, which can activate multiple warning systems at the same time and with a single input. This guarantees consistency of warning messages and would easily replace specific application-oriented messages with a single multihazard message format. CAP is compatible with all types of information systems and public alerting systems (including broadcast radio and television), public and private data networks, multi-lingual warning systems and emerging technologies such as Internet Web services and existing systems such as the US National Emergency Alert System and the National Oceanic and Atmospheric Organization (NOAA) Weather Radio. CAP uses Extensible Markup Language (XML), which contains information about the alert message, the specific hazard event, and appropriate responses, including the urgency of action to be taken, severity of the event, and certainty of the information.

For early warning systems to be effective, it is essential that they be integrated into policies for disaster mitigation. Good governance priorities include protecting the public from disasters through the implementation of disaster risk reduction policies. It is clear that natural phenomena cannot be prevented, but their human, socio-economic and environmental impacts can and should be minimized through appropriate measures, including risk and vulnerability reduction strategies, early warning, and appropriate action plans. Most often, these problems are given attention during or immediately after a disaster. Disaster risk reduction measures require long term plans and early warning should be seen as a strategy to effectively reduce the growing vulnerability of communities and assets.

The information provided by early warning systems enables authorities and institutions at various levels to immediately and effectively respond to a disaster. It is crucial that local government, local institutions, and communities be involved in the entire policy-making process, so they are fully aware and prepared to respond with short and long-term action plans.

The early warning process, as previously described, is composed of 4 main stages: risk assessment, monitoring and predicting, disseminating and communicating warnings, and response. Within this framework, the first phase, when short-and long-term actions plans are laid out based on risk assessment analysis, is the realm of institutional and political actors. Then EW acquires a technical dimension in the monitoring and predicting phase, while in the communication phase, EW involves both technical and institutional responsibility. The response phase then involves many more sectors, such as national and local institutions, non-governmental organizations, communities, and individuals.

Below is a summary of recommendations for effective decision making within the early warning process (Sarevitz et al., 2000) .

Prediction efforts by the scientific community alone are insufficient for decision making. The scientific community and policy-makers should outline the strategy for effective and timely decision making by indicating what information is needed by decision-makers, how predictions will be used, how reliable the prediction must be to produce an effective response, and how to communicate this information and the tolerable prediction uncertainty so that the information can be received and understood by authorities and public. A miscommunicated or misused prediction can result in costs to society. Prediction, communication, and use of the information are necessary factors in effective decision making within the early warning process.

Wishing not to appear ''alarmist'' or to avoid criticism, local and national governments have sometimes kept the public in the dark when receiving technical information regarding imminent threats. The lack of clear and easy-to-use information can sometimes confuse people and undermine their confidence in public officials. Conversely, there are quite a few cases where the public may have refused to respond to early warnings from authorities, and have therefore exposed themselves to danger or forced governments to impose removal measures. In any case, clear and balanced information is critical, even when some level of uncertainty remains. For this reason, the information's uncertainty level must be communicated to users together with the early warning .

Resources must be allocated wisely and priorities should be set, based on risk assessment, for long-and short-term decision making, such as investing in local early warning systems, education, or enhanced monitoring and observational systems. In addition, decision-makers need to be able to set priorities for timely and effective response to a disaster when it occurs based on the information received from the early warning system. Decision-makers should receive necessary training on how to use the information received when an alert is issued and what that information means.

Institutional networks should be developed with clear responsibilities. Complex problems such as disaster mitigation and response require multidisciplinary research, multi-sector policy and planning, multi-stakeholder participation, and networking involving all the participants of the process, such as the scientific research community (including social sciences aspects), land use planning, environment, finance, development, education, health, energy, communications, transportation, labour, and social security and national defence. Decentralization in the decision making process could lead to optimal solutions by clarifying local government and community responsibilities. Collaboration will improve efficiency, credibility, accountability, trust, and costeffectiveness. This collaboration consists of joint research projects, sharing information and participatory strategic planning and programming.

Because there are numerous actors involved in early warning response plans (such as governing authorities, municipalities, townships, and local communities), the decision making and legal framework of responsibilities should be set up in advance in order to be prepared when a disaster occurs. Hurricane Katrina in 2005 showed gaps in the legal frameworks and definition of responsibilities that exacerbated the disaster. Such ineffective decision making must be dealt with to avoid future disasters such as the one in New Orleans.

Earth observation (EO), through measuring and monitoring, provides an insight and understanding into Earth's complex processes and changes. EO includes measurements that can be made directly or by sensors in-situ or remotely (i.e. satellite remote sensing, aerial surveys, land or oceanbased monitoring systems, Fig. 3 ), to provide key information to models or other tools to support decision making processes. EO assists governments and civil society to identify and shape corrective and new measures to achieve sustainable development through original, scientifically valid assessments and early warning information on the recent and potential long-term consequences of human activities on the biosphere. At a time when the world community is striving to identify the impacts of human actions on the planet's life support system, time sequenced satellite images help to determine these impacts and provide unique, visible and scientifically convincing evidence that human actions are causing substantial changes to the Earth's environment and natural resource base (i.e. ecosystems changes, urban growth, transboundary pollutants, loss of wetlands, etc.). By enhancing the visualization of scientific information on environmental change, satellite imagery will enhance environmental management and raise the awareness of emerging environmental threats. EO provides the opportunity to explore, to discover, and to understand the world in which we live from the unique vantage point of space.

The following section discusses the potential role of EO for each type of environmental threat.

2.1. Ongoing and rapid/sudden-onset environmental threats 2.1.1. Oil spills Earth observation is increasingly used to detect illegal marine discharges and oil spills. Infra-red (IR) video and photography from airborne platforms, thermal infrared imaging, airborne laser fluorosensors, airborne and satellite optical sensors, as well as airborne and satellite Synthetic Aperture Radar (SAR) are used for this purpose. SAR has the advantage of also providing data during cloud cover conditions and darkness, unlike optical sensors. In addition, optical-sensor techniques applied to oil spills detection are associated to a high number of false alarms, more often cloud shadows, sun glint, and other conditions such as precipitation, fog, and the amounts of daylight present also may be erroneously associated with oil spills. For this reason, SAR is preferred over optical sensors, especially when spills cover vast areas of the marine environment, and when the oil cannot be seen or discriminated against the background. SAR detects changes in sea-surface roughness patterns modified by oil spills. The largest shortcoming of oil spills detection using SAR images is accurate discrimination between oil spills and natural films (Brekke and Soldberg, 2005) . To date, operational application of satellite imagery for oil spill detection still remains a challenge due to limited spatial and temporal resolution. In addition, processing times are often too long for operational purposes, and it is still not possible to measure the thickness of the oil spill (Mansor et al., 2007; US Fig. 3 . Illustration of multiple observing systems in use on the ground, at sea, in the atmosphere and from space for monitoring and researching the climate system (WMO, 2011). Interior, Minerals Management Service, 2007) . Existing applications are presented in the Section 3.

Chemical and nuclear accidents may have disastrous consequences, such as the 1984 accident in Bhopal, India, which killed more than 2000 and injured about 150 000, and the 1986 explosion of the reactors of the nuclear power plant in Chernobyl, Ukraine, which was the worst such accident to date, affecting part of the Soviet Union, eastern Europe, Scandinavia, and later, western Europe. Meteorological factors such as wind speed and direction, turbulence, stability layers, humidity, cloudiness, precipitation and topographical features, influence the impact of chemical and nuclear accidents and have to be taken into account in decision models. In some cases, emergencies are localized while in others, transport processes are most important. EO provides key data for monitoring and forecasting the dispersion and spread of the substance.

Geohazards associated with geological processes such as earthquakes, landslides, and volcanic eruptions are mainly controlled by ground deformation. EO data allows monitoring of key physical parameters associated with geohazards, such as deformation, plate movements, seismic monitoring, baseline topographic, and geoscience mapping. EO products are useful for detection and mitigation before the event, and for damage assessment during the aftermath. For geohazards, stereo optical and radar interferometry associated with ground-based Global Positioning System (GPS) and seismic networks are used. For volcanic eruptions additional parameters are observed such as temperature and gas emissions. Ground based measurements have the advantage of being continuous in time but have limited spatial extent, while satellite observations cover wide areas but are not continuous in time. These data need to be integrated for an improved and more comprehensive approach (Committee on Earth Observation Satellites (CEOS), 2002; Integrated Global Observing Strategy (IGOS-P), 2003).

Earthquakes are due to a sudden release of stresses accumulated around the faults in the Earth's crust. This energy is released through seismic waves that travel from the origin zone, which cause the ground to shake. Severe earthquakes can affect buildings and populations. The level of damage depends on many factors, such as the intensity and depth of the earthquake, and the vulnerability of structures and their distance from the earthquake's origin. For earthquakes, information on the location and magnitude of the event first needs to be conveyed to responsible authorities. This information is used by seismic early warning systems to activate security measures within seconds after the earthquake's origin and before strong shaking occurs at the site. Shakemaps generated within five minutes provide essential information to assess the intensity of ground shaking and the damaged areas. The combination of data from seismic networks and GPS may help to increase reliability and timeliness of this information. Earthquake frequency and probability shakemaps based on historical seismicity and base maps (geological, soil type, active faults, hydrological, and DEMs), assist in the earthquake mitigation phase and need to be included in the building code design process for improved land use and building practices. For responses, additional data are needed, such as seismicity, intensity, strain, DEMs, soil type, moisture conditions, infrastructure and population, to produce post-event damage maps. Thermal information needs to continuously be monitored. This is obtained from low/medium resolution IR imagery from polar and geostationary satellites for thermal background characterization (Advanced Very High Resolution Radiometer (AVHRR), ATSR, MODIS and GOES) together with deformation from EDM and/or GPS network; borehole strainmetres; and SAR interferometry.

Landslides are displacements of earth, rock, and debris caused by heavy rains, floods, earthquakes, volcanoes, and wildfires. Useful information for landslides and ground instability include the following: hazard zonation maps (landslides, debris flows, rockfalls, subsidence, and ground instability scenarios) during the mitigation phase, associated with landlside inventories, DEM, deformation (GPS network; SAR interferometry; other surveys such as leveling, laser scanning, aerial, etc.), hydrology, geology, soil, geophysical, geotechnical, climatic, seismic zonation maps, land cover, land use, and historical archives. Forecasting the location and extent of ground instability or landslides is quite challenging. Landslides can be preceded by cracks, accelerating movement, and rock fall activity. Real-time monitoring of key parameters thus becomes essential. The observed acceleration, deformation or displacement, exceeding a theoretical pre-fixed threshold is the trigger for issuing an alert signal. An alternative approach is based on hydrologic forecasting. It should be said that for large areas site-specific monitoring is not feasible. In this case, hazard mapping associated with monitoring of high risk zones remains the best option for warning. Local rapid mapping of affected areas, updated scenarios and real-time monitoring (deformation, seismic data, and weather forecasts) assist during the response phase.

A tsunami is a series of ocean waves generated by sudden displacements in the sea floor, landslides, or volcanic activity. Although a tsunami cannot be prevented, the impact of a tsunami can be mitigated through community preparedness, timely warnings, and effective response. Observations of seismic activity, sea floor bathymetry, topography, sea level data (Tide Gauge observations of sea height; Real-time Tsunami Warning Buoy Data; Deep Ocean Assessment and Reporting of Tsunamis (DART) buoys and sea-level variations from the TOPEX/Poseidon and Jason, the European Space Agency's Envisat, and the US Navy's Geosat Follow-On), are used in combination with tsunami models to create inundation and evacuation maps and to issue tsunami watches and warnings.

Volcanic eruptions may be mild, releasing steam and gases or lava flows, or they can be violent explosions that release ashes and gases into the atmosphere. Volcanic eruptions can destroy land and communities living in their path, affect air quality, and even influence the Earth's climate. Volcanic ash can impact aviation and communications. Data needs for volcanic eruptions include hazard zonation maps, real-time seismic, deformation (Electronic Distance Measurement (EDM) and/or GPS network; leveling and tilt networks; borehole strainmeters; gravity surveys; SAR interferometry), thermal (Landsat, ASTER, Geostationary operational environmental satellites (GOES), MODIS); air borne IR cameras; medium-high resolution heat flux imagery and gas emissions (COSPEC, LICOR surveys); Satellite imagery (i.e., ASTER) and digital elevation maps (DEM). As soon as the volcanic unrest initiates, information needs to be timely and relatively high-resolution. Once the eruption starts, the flow of information has to speed up. Seismic behaviour and deformation patterns need to be observed throughout the eruption especially to detect a change of eruption site (3-6 seismometers ideally with 3-directional sensors; a regional network).

Hydro-meteorological hazards include the wide variety of meteorological, hydrological and climate phenomena that can pose a threat to life, property and the environment. These types of hazards are monitored using the meteorological, or weather, satellite programs, beginning in the early 1960s. In the United States, NASA, NOAA, and the Department of Defense (DoD) have all been involved with developing and operating weather satellites. In Europe, ESA and EUMETSAT (European Organisation for the Exploitation of Meteorological Satellites) operate the meteorological satellite system (US Centennial of Flight Commission). Data from geostationary satellite and polar microwave derived products (GOES) and polar orbiters (microwave data from the Defense Meteorological Satellite Program (DMSP), Special Sensor Microwave/Imager (SSM/I), NOAA/Advanced Microwave Sounding Unit (AMSU), and Tropical Rainfall Measuring Mission (TRMM)) are key in weather analysis and forecasting. GOES has the capability of observing the atmosphere and its cloud cover from the global scale down to the storm scale, frequently and at high resolution. Microwave data are available on only an intermittent basis, but are strongly related to cloud and atmospheric properties.

is key for monitoring meteorological processes from the global scale to the synoptic scale to the mesoscale and finally to the storm scale. (Scofield et al., 2002) . GOES and POES weather satellites provide useful information on precipitation, moisture, temperature, winds and soil wetness, which is combined with ground observation.

Floods are often triggered by severe storms, tropical cyclones, and tornadoes. The number of floods has continued to rise steadily; together with droughts, they have become the most deadly disasters over the past decades. The increase in losses from floods is also due to climate variability, which has caused increased precipitation in parts of the Northern Hemisphere (Natural Hazards Working Group, 2005) . Floods can be deadly, particularly when they arrive without warning. In particular, polar orbital and geostationary satellite data are used for flood observation. Polar orbital satellites include optical low (AVHRR), medium (Landsat, SPOT, IRS) and high resolution (IKONOS) and microwave sensors (high (SAR-RADARSAT, JERS and ERS) and low resolution passive sensors (SSMI). Meteorological satellites include GOES 8 and 10, METEOSAT, GMS, the Indian INSAT and the Russian GOMS; and polar orbitals such as NOAA (NOAA 15) and SSMI. For storms, additional parameters are monitored, such as sea surface temperature, air humidity, surface wind speed, rain estimates (from DMSP/SSMI, TRMM, ERS, QuikScat, AVHRR, RADARSAT). TRMM offers unique opportunities to examine tropical cyclones. With TRMM, scientists are able to make extremely precise radar measurements of tropical storms over the oceans and identify their intensity variations, providing invaluable insights into the dynamics of tropical storms and rainfall.

Epidemics such as malaria and meningitis are linked to environmental factors. Satellite data can provide essential information on these factors and help to better understand diseases. As an example, the ESA Epidemio project, launched in 2004, utilizes data from ESA's Envisat or the French Space Agency's Spot, and field data to gather information on the spread of epidemics, helping to better prepare for epidemic outbreaks. GEO, with WHO and other partners, are working together on the Meningitis Environmental Risk Information Technologies (MERIT) project to better understand the relationship between meningitis and environmental factors using remote sensing.

Wildfires pose a threat to lives and properties and are often connected to secondary effects such as landslides, erosion, and changes in water quality. Wildfires may be natural processes, human induced for agriculture purposes, or just the result of human negligence. Wildfire detection using satellite technologies is possible thanks to significant temperature difference between the Earth's surface (usually not exceeding 10-25 1C) and the heat of fire (300-900 1C), which results in a thousand times difference in heat radiation generated by these objects. NOAA (AVHRR radiometer with 1100 m spatial resolution and 3000 km swath width) and Earth Observing Satellites (EOS) (Terra and Aqua satellites with MODIS radiometer installed on them with 250, 500 and 1000 m spatial resolution and 2330 km swath width) are the most widely used modern satellites for operative fire monitoring (Klaver et al., 1998) . High-resolution sensors, such as the Landsat Thematic Mapper, SPOT multispectral scanner, or National Oceanic and Atmospheric Administration's AVHRR or MODIS, are used for fire potential definition. Sensors used for fire detection and monitoring include AVHRR, which has a thermal sensor and daily overflights, the Defense Meteorological Satellite Program's Optical Linescan System (OLS) sensor, which has daily overflights and operationally collects visible images during its nighttime pass, and the MODIS Land Rapid Response system. AVHRR and higher resolution images (SPOT, Landsat, and radar) can be used to assess the extent and impact of the fire.

Smog is the product of human and natural activities, such as industry, transportation, wildfires, volcanic eruptions, etc. and can have serious effects on human health and the environment.

A variety of EO tools are available to monitor air quality. The National Aeronautics and Space Administration (NASA) and the European Space Agency (ESA) both have instruments to monitor air quality. The Canadian MOPITT (Measurements of Pollution in the Troposphere) aboard the Terra satellite monitors the lower atmosphere to observe how it interacts with the land and ocean biospheres, distribution, transport, sources, and sinks of carbon monoxide and methane in the troposphere. The Total Ozone Mapping Spectrometer (TOMS) instrument measures the total amount of ozone in a column of atmosphere as well as cloud cover over the entire globe. Additionally, TOMS measures the amount of solar radiation escaping from the top of the atmosphere to accurately estimate the amount of ultraviolet radiation that reaches the Earth's surface. The Ozone Monitoring Instrument (OMI) on Aura will continue the TOMS record for total ozone and other atmospheric parameters related to ozone chemistry and climate. The OMI instrument distinguishes between aerosol types, such as smoke, dust, and sulphates, and can measure cloud pressure and coverage. ESA's SCHIAMACHY (Scanning Imaging Absorption Spectro-Meter for Atmospheric ChartographY) maps atmosphere over a very wide wavelength range (240-2380 nm), which allows detection of trace gases, ozone and related gases, clouds and dust particles throughout the atmosphere (Athena Global, 2005) . The Moderate Resolution Imaging Spectroradiometer (MODIS) sensor measures the relative amount of aerosols and the relative size of aerosol particles-solid or liquid particles suspended in the atmosphere. Examples of such aerosols include dust, sea salts, volcanic ash, and smoke. The MODIS aerosol optical depth product is a measure of how much light airborne particles prevent from passing through a column of atmosphere. New technologies are also being explored for monitoring air quality, such as mobile phones equipped with simple sensors to empower citizens to collect and share real-time air quality measurements. This technology is being developed by a consortium called Urban Atmospheres.

The traditional methods of monitoring coastal water quality require scientists to use boats to gather water samples, typically on a monthly basis because of the high costs of these surveys. This method captures episodic events affecting water quality, such as the seasonal freshwater runoff, but is not able to monitor and detect fast changes. Satellite data provide measures of key indicators of water qualityturbidity and water clarity -to help monitor fast changes in factors that affect water quality, such as winds, tides and human influences including pollution and runoff. GeoEYE's Sea-viewing Wide Field-ofview Sensor (SeaWiFS) instrument, launched aboard the OrbView-2 satellite in 1997, collects ocean colour data used to determine factors affecting global change, particularly ocean ecology and chemistry. MODIS sensor, launched aboard the Aqua satellite in 2002, together with its counterpart instrument aboard the Terra satellite, collects measurements from the entire Earth surface every 1-2 days and can also provide measurements of turbidity (Bjorn-Hansen, 2007) . Overall, air and water quality monitoring coverage still appears to be irregular and adequate and available in real-time only for some contaminants (Global Earth Observation System of Systems, 2005).

2.2.3.1. Droughts. NOAA's National Weather Service (NWS) defines a drought as ''a period of abnormally dry weather sufficiently prolonged for the lack of water to cause serious hydrologic imbalance in the affected area.'' Drought can be classified by using 4 different definitions: meteorological (deviation from normal precipitation); agricultural (abnormal soil moisture conditions); hydrological (related to abnormal water resources); and socio-economic (when water shortage impacts people's lives and economies).

A comprehensive and integrated approach is required to monitor droughts, due to the complex nature of the problem. Although all types of droughts originate from a precipitation deficiency, it is insufficient to monitor solely this parameter to assess severity and resultant impacts (World Meteorological Organization, 2006) . Effective drought early warning systems must integrate precipitation and other climatic parameters with water information such as streamflow, snow pack, groundwater levels, reservoir and lake levels, and soil moisture, into a comprehensive assessment of current and future drought and water supply conditions (Svoboda et al., 2002) . In particular, there are 6 key parameters that are used in a composite product developed from a rich information stream, including climate indices, numerical models, and the input of regional and local experts. These are as follows:

1) Palmer Drought Severity Index (based on precipitation data, temperature data, division constants (water capacity of the soil, etc.) and previous history of the indices). 2) Soil Moisture Model Percentile (calculated through a hydrological model that takes observed precipitation and temperature and calculates soil moisture, evaporation and runoff. The potential evaporation is estimated from observed temperature). 3) Daily stream flow percentiles. 4) Percent of normal precipitation. 5) Standardized precipitation index, and 6) Remotely sensed vegetation health index.

Additional indicators may include the Palmer Crop Moisture Index, Keetch-Byram Drought Index, Fire Danger Index, evaporation-related observations such as relative humidity and temperature departure from normal, reservoir and lake levels, groundwater levels, field observations of surface soil moisture, and snowpack observations. Some of these indices and indicators are computed for point locations, and others are computed for climate divisions, drainage (hydrological) basins, or other geographical regions (Svoboda et al., 2002) . A complete list of drought products can be found on NOAA's National Environmental Satellite, Data, & Information Service (NOAANESDIS) web page.

2.2.3.2. Desertification. Desertification refers to the degradation of land in arid, semi-arid, and dry sub-humid areas due to climatic variations or human activity. Desertification can occur due to inappropriate land use, overgrazing, deforestation, and over-exploitation. Land degradation affects many countries worldwide and has its greatest impact in Africa. In spite of the potential benefits of EO information, the lack of awareness of the value and availability of information, inadequate institutional resources and financial problems are the most frequent challenges to overcome in detecting desertification (Sarmap, 2003) . In 2004, through a project called DesertWatch, ESA has developed a set of indicators based principally on land surface parameters retrieved from satellite observations for monitoring land degradation and desertification. DesertWatch is being tested and applied in Mozambique, Portugal, and Brazil. The UN Food and Agriculture Organization's Global Information and Early Warning System (GIEWS). These provide information on food availability, market prices and livelihoods.

The observations of climate-related variables on a global scale have made it possible to document and analyse the behaviour of Earth's climate, made available through programs as follows: the IOC-WMO-UNEP-ICSU Global Ocean Observing System (GOOS); the FAO-WMO-UNESCO-UNEPICSU Global Terrestrial Observing System (GTOS); the WMO Global Observing System (GOS) and Global Atmosphere Watch (GAW); the research observing systems and observing systems research of the WMO-IOC-ICSU World Climate Research Programme (WCRP) and other climate relevant international programs; and WMO-UNESCO-ICSUIOC-UNEP Global Climate Observing System (GCOS). The Intergovernmental Panel on Climate Change (IPCC) periodically reviews and assesses the most recent scientific, technical and socio-economic information produced worldwide relevant to the understanding of climate change. Hundreds of scientists worldwide contribute to the preparation and review of these reports.

According to the recent IPCC report, the atmospheric buildup of greenhouse gases is already shaping the earth's climate and ecosystems from the poles to the tropics, which face inevitable, possibly profound, alteration. The IPCC has predicted widening droughts in southern Europe and the Middle East, sub-Saharan Africa, the American Southwest and Mexico, and flooding that could imperil low-lying islands and the crowded river deltas of southern Asia. It stressed that many of the regions facing the greatest risks are among the world's poorest. Information about the impacts of climate variability impact information is needed by communities and resource managers to adapt and prepare for larger fluctuations as global climate change becomes more evident. This information includes evidence of changes occurring due to climate variability, such as loss of ecosystems, ice melting, coastal degradation, and severe droughts. Such information will provide policy-makers scientifically valid assessment and early warning information on the current and potential long-term consequences of human activities on the environment.

(i.e., ecosystem changes, loss of biodiversity and habitats, land cover/land changes, coastal erosion, urban growth, etc.). Landsat satellites (series 1-7) are extensively used to monitor location-specific environmental changes. They have the great advantage of providing repetitive, synoptic, global coverage of high-resolution multi-spectral imagery (Fadhil, 2006) . Landsat can be used for change detection applications to identify differences in the state of an object or phenomenon by comparing the satellite imagery at different times. Change detection is key in natural resources management (Singh, 1989) . Central to this theme is the characterization, monitoring and understanding of land cover and land use change, since they have a major impact on sustainable land use, biodiversity, conservation, biogeochemical cycles, as well as land-atmosphere interactions affecting climate and they are indicators of climate change, especially at a regional level (IGOS-P, 2004) .

The United Nations Environment Programme's (UNEP) bestselling publication One Planet, Many People: Atlas of Our Changing Environment, which shows before and after satellite photos to document changes to the Earth's surface over the past 30 years, proves the importance and impact of visual evidence of environmental change in hotspots. The Atlas contains some remarkable Landsat satellite imagery and illustrates the alarming rate of environmental destruction. Through the innovative use of some 271 satellite images, 215 ground photos and 66 maps, the Atlas provides visual proof of global environmental changes -both positive and negative -resulting from natural processes and human activities. Case studies include themes such as atmosphere, coastal areas, waters, forests, croplands, grasslands, urban areas, and tundra and Polar Regions. The Atlas demonstrates how our growing numbers and our consumption patterns are shrinking our natural resource base.

The aim of this report is to identify current gaps and future needs of early warning systems through the analysis of the state of the art of existing early warning and monitoring systems for environmental hazards. Among existing early warning/monitoring systems, only systems that provide publicly accessible information and products have been included in the analysis. For the present study, several sources have been used, such as the Global Survey of Early Warning Systems (United Nations, 2006) together with the online inventory of early warning systems on ISDR's Platform for the Promotion of Early Warning (PPEW) website, and several additional online sources, technical reports and scientific articles listed in the references. For each hazard type, a gap analysis has been carried out to identify critical aspects and future needs of EWS, considering aspects such as geographical coverage, and essential EWS elements such as monitoring and prediction capability, communication systems and application of early warning information in responses. Below is the outcome of the review of existing early warning/monitoring systems for each hazard type. Details of all systems, organized in tables by hazard type, are listed in the Appendix. The current gaps identified for each hazard type could be related to technological, organizational, communication or geographical coverage aspects. To assess the geographical coverage of existing systems for each hazard type, the existing systems have been imposed on the hazard's risk map. For this analysis, the maps of risks of mortality and economic loss were taken from Natural Disaster Hotspots: A Global Risk Analysis, a report from the World Bank (Dilley et al., 2005) .

To detect operational oil spills, satellite overpasses and aerial surveillance flights need to be used in an integrated manner. In many countries in Northern Europe, the KSAT manual approach is currently used to identify oil spills from the satellite images. KSAT has provided this operational service since 1996, and in Europe, use of satellites for oil spill detection is well established and well integrated within the national and regional oil pollution surveillance and response chains. Operational algorithms utilizing satellite-borne C-band SAR instruments (Radarsat-1, Envisat, Radarsat-2) are also being developed for oil-spill detection in the Baltic Sea area.

Releases of a hazardous substance from industrial accidents can have immediate adverse effects on human and animal life or the environment. WMO together with IAEA provides specialized meteorological support to environmental emergency response related to nuclear accidents and radiological emergencies. The WMO network of eight specialized numerical modelling centres called Regional Specialized Meteorological Centres (RSMCs) provides predictions of the movement of contaminants in the atmosphere. The Inter-Agency Committee on the Response to Nuclear Accidents (IACRNA) of the IAEA, coordinates the international intergovernmental organizations responding to nuclear and radiological emergencies. IACRNA members are: the European Commission (EC), the European Police Office (EUROPOL), the Food and Agriculture Organization of the United Nations (FAO), IAEA, the International Civil Aviation Organization (ICAO), the International Criminal Police Organization (INTERPOL), the Nuclear Energy Agency of the Organization for Economic Co-operation and Development (OECD/NEA), the Pan American Health Organization (PAHO), UNEP, the United Nations Office for the Co-ordination of Humanitarian Affairs (UN-OCHA), the United Nations Office for Outer Space Affairs (UNOOSA), the World Health Organization (WHO), and WMO. The Agency's goal is to provide support during incidents or emergencies by providing near real-time reporting of information through the following: the Incident and Emergency Centre (IEC), which maintains a 24 h oncall system for rapid initial assessment, and if needed, triggers response operations; the Emergency Notification and Assistance Convention Website (ENAC), which allows the exchange of information on nuclear accidents or radiological emergencies; and the Nuclear Event Webbased System (NEWS), which provides information on all significant events in nuclear power plants, research reactors, nuclear fuel cycle facilities and occurrences involving radiation sources or the transport of radioactive material. The Global Chemical Incident Alert and Response System of the International Programme on Chemical Safety, which is part of WHO, focuses on disease outbreaks from chemical releases and also provides technical assistance to Member States for response to chemical incidents and emergencies. Formal and informal sources are used to collect information and if necessary, additional information and verification is sought through official channels: national authorities, WHO offices, WHO Collaborating Centres, other United Nations agencies, and members of the communicable disease Global Outbreak Alert and Response Network (GOARN), Internet-based resources, particularly the Global Public Health Intelligence Network (GPHIN) and ProMED-Mail. Based on this information, a risk assessment is carried out to determine the potential impact and if assistance needs to be offered to Member States.

3.1.3. Geological hazards 3.1.3.1. Earthquakes. Earthquake early warning systems are a relatively new approach to seismic risk reduction. They provide a rapid estimate of seismic parameters such as magnitude and location associated with a seismic event based on the first seconds of seismic data registered at the epicentre. This information can then be used to predict ground motion parameters of engineering interest including peak ground acceleration and spectral acceleration. Earthquake warning systems are currently operational in Mexico, Japan, Romania, Taiwan and Turkey (Espinosa Aranda et al., 1995; Wu et al., 1998; Wu and Teng, 2002; Odaka et al., 2003; Kamigaichi, 2004; Nakamura, 2004; Horiuchi et al., 2005) . Systems are under development for seismic risk mitigation in California and Italy. Local and national scale seismic early warning systems, which provide seismic information between a few seconds and tens of seconds before shaking occurs at the target site, are used for a variety of applications such as shutting down power plants, stopping trains, evacuating buildings, closing gas valves, and alerting wide segments of the population through the TV, among others.

On the global scale, multi-national initiatives, such as the US Geological Survey (USGS) and GEO-FOrschungs Netz (GEOFON), operate global seismic networks for seismic monitoring but do not provide seismic early warning information. Today, the USGS in cooperation with Incorporated Research Institutions for Seismology (IRIS) operates the Global Seismic Networks (GSN), which comprises more than 100 stations providing free, real-time, open access data. GEOFON collects information from several networks and makes this information available to the public online. USGS Earthquake Notification Service (ENS) provides publicly available e-mail notification for earthquakes worldwide within 5minutes for earthquakes in US and within 30 min for events worldwide. USGS also provides near-real-time maps of ground motion and shaking intensity following significant earthquakes. This product, called ShakeMap, is being used for post-earthquake response and recovery, public and scientific information, as well as for preparedness exercises and disaster planning.

Effective early warning technologies for earthquakes are much more challenging to develop than for other natural hazards because warning times range from only a few seconds in the area close to a rupturing fault to a minute or so (Heaton, 1985; Allen and Kanamori, 2003; Kanamori, 2005) . Several local and regional applications exist worldwide but no global system exists or could possibly exist for seismic early warning at global scale, due to timing constraints. Earthquake early warning systems applications must be designed at the local or regional level. Although various early warning systems exist worldwide at the local or regional scale, there are still high seismic risk areas that lack early warning applications, such as Peru, Chile, Iran, Pakistan, and India. The International Consortium on Landslides (ICL), created at the Kyoto Symposium in January 2002, is an international non-governmental and non-profit scientific organization, which is supported by the United Nations Educational, Scientific and Cultural Organization (UNESCO), the WMO, FAO, and the United Nations International Strategy for Disaster Reduction (UN/ISDR). ICL's mission is to promote landslide research for the benefit of society and the environment and promote a global, multidisciplinary programme regarding landslides. ICL provides information about current landslides on its website, streaming this information from various sources such as the Geological Survey of Canada. This information does not provide any early warning since it is based on news reports after the events have occurred. Enhancing ICL's existing organizational infrastructure by improving landslide prediction capability would allow ICL to provide early warning to authorities and populations. Technologies for slopes monitoring has greatly improved, but currently only few slopes are being monitored at a global scale. The use of these technologies would be greatly beneficial for mitigating losses from landslides worldwide.

3.1.3.3. Tsunamis. The Indian Ocean tsunami of December 2004 killed 220 000 people and left 1.5 million homeless. It highlighted gaps and deficiencies in existing tsunami warning systems. In response to this disaster, in June 2005 the Intergovernmental Oceanographic Commission (IOC) secretariat was mandated by its member states to coordinate the implementation of a tsunami warning system for the Indian Ocean, the northeast Atlantic and Mediterranean, and the Caribbean. Efforts to develop these systems are ongoing. Since March 2011, the Indonesian meteorological, climatological and geophysical agency has been operating the German-Indonesian Tsunami Early Warning System for the Indian Ocean. Milestones, such as the development of the automatic data processing software and underwater communication for the transmission of pressure data from the ocean floor to a warning centre, have been reached. These systems will be part of the Global Ocean Observing System (GOOS), which will be part of GEOSS.

The Pacific basin is monitored by the Pacific Tsunami Warning System (PTWS), which was established by 26 Member States and is operated by the Pacific Tsunami Warning Center (PTWC), located near Honolulu, Hawaii. PTWC monitors stations throughout the Pacific basin to issue tsunami warnings to Member States, serving as the regional centre for Hawaii and as a national and international tsunami information centre. It is part of the PTWS effort. NOAA National Weather Service operates PTWC and the Alaska Tsunami Warning Center (ATWC) in Palmer, Alaska, which serves as the regional Tsunami Warning Center for Alaska, British Columbia, Washington, Oregon, and California. PTWS monitors seismic stations operated by PTWC, USGS and ATWC to detect potentially tsunamigenic earthquakes. Such earthquakes meet specific criteria for generation of a tsunami in terms of location, depth, and magnitude. PTWS issues tsunami warnings to potentially affected areas, by providing estimates of tsunami arrival times and areas potentially most affected. If a significant tsunami is detected, the tsunami warning is extended to the Pacific basin. The International Tsunami Information Center (ITIC), under the auspices of IOC, aims to mitigate tsunami risk by providing guidance and assistance to improve education and preparedness. ITIC also provides a complete list of tsunami events worldwide. Official tsunami bulletins are released by PTWC, ATWC, and the Japan Meteorological Agency (JMA). Regional and national tsunami information centres exist worldwide; the complete list is available from IOC.

Currently, no global tsunami warning system is in place. In addition, fully operational tsunami early warning systems are needed for the Indian Ocean and the Caribbean. Initial steps have been taken in this direction. In 2010, NOAA established the Caribbean Tsunami Warning Program as the first step towards the development of a Caribbean Tsunami Warning Center. Since 2005, steps have been taken to develop an Indian Ocean tsunami system, such as establishing 26 tsunami information centres and deploying 23 real-time sea level stations and 3 deep ocean buoys in countries bordering Indian Ocean. In 2005, the United States Agency for International Development (USAID) launched the US Indian Ocean Tsunami Warning Systems Program as the US Government's direct contribution to the international effort led by the IOC. Since then, there are ongoing activities, such as Germany's 5-year German-Indonesia Tsunami Early Warning System program with Indonesia, the Tsunami Regional Trust Fund established in 2005, and the United Kingdom's tsunami funds reserved for early warning capacity building.

Nevertheless, on 17 July 2006, only one month after the announcement that the Indian Ocean's tsunami warning system was operational, a tsunami in Java, Indonesia, killed hundreds of people. On that day, tsunami warnings were issued to alert Jakarta but there was not enough time to alert the coastal areas. The July 2006 tsunami disaster illustrates that there are still operational gaps to be solved in the Indian Ocean tsunami early warning system, notably in warning coastal communities on time.

3.1.3.4. Volcanic eruptions. Volcanic eruptions are always anticipated by precursor activities. In fact, seismic monitoring, ground deformation monitoring, gas monitoring, visual observations, and surveying are used to monitor volcanic activity. Volcano observatories are distributed worldwide. A complete list of volcano observatories is available at the World Organization of Volcanic Observatories (WOVO) website. However, there is still a divide between developed and developing countries. In particular, a large number of observatories and research centres monitor volcanoes in Japan and the United States very well. Most Central and South American countries (Mexico, Guatemala, El Salvador, Nicaragua, Costa Rica, Colombia, Ecuador, Peru, Chile, Trinidad, and the Antilles) have volcano observatories that provide public access to volcanic activity information. In Africa, only two countries (Congo and Cameroon) have volcano monitoring observatories and they do not provide public access to information. Only a small number, probably fewer than 50, of the world's volcanoes are well monitored, mostly due to inadequate resources in poor countries (National Hazards Working Group, 2005) . There is a need to fill this gap by increasing the coverage of volcanic observatories.

Currently, there is no global early warning system for volcanic eruptions except for aviation safety. Global volcanic activity information is provided by the Smithsonian institution, which partners with the USGS under the Global Volcanism Program to provide online access to volcanic activity information, collected from volcano observatories worldwide. Reports and warnings are available on a daily basis. Weekly and monthly summary reports are also available, but these only report changes in volcanic activity level, ash advisories, and news reports. The information is also available through Google Earth. This information is essential for the aviation sector, which must be alerted to ash-producing eruptions. There are several ash advisory centres distributed worldwide, in London, Toulouse, Anchorage, Washington, Montreal, Darwin, Wellington, Tokyo, and Buenos Aires. However, there is a need to coordinate interaction and data sharing among the approximately 80 volcano observatories that make up WOVO. ESA is developing GlobVolcano, an Information System to provide earth observations for volcanic risk monitoring.

3.1.3.5. Wildfires. Early warning methodologies for wildfires are based on the prediction of precursors, such as fuel loads and lightning danger. These parameters are relevant for triggering prediction, but once the fire has begun, fire behaviour and pattern modelling are fundamental for estimating fire propagation patterns. Most industrial countries have EW capabilities in place, while most developing countries have neither fire early warning nor monitoring systems in place (Goldammer et al., 2003) . Local and regional scale fire monitoring systems are available for Canada, South America, Mexico and South Africa. An interactive mapping service based on Google maps and EO imagery from INPE, the Brazilian Space Research Institute, has been available since September 2008. Individuals can contribute with information from the ground; in only 3 months the service has received 41 million reports on forest fires and illegal logging, making it one of the most successful web sites in Brazil; it has had real impact through follow up legal initiatives and Parliamentary enquiries.

Wildfire information is available worldwide through the Global Fire Monitoring Center (GFMC), a global portal for fire data products, information, and monitoring. This information is accessible to the public through the GFMC website but is not actively disseminated. The GFMC provides global fire products through a worldwide network of cooperating institutions. GFMC fire products include the following: fire danger maps and forecasts, which provide assessment of fire onset risk; near realtime fire events information; an archive of global fire information; and assistance and support in the case of a fire emergency. Global fire weather forecasts are provided by the Experimental Climate Prediction Center (ECPC), which also provides national and regional scale forecasts. NOAA provides experimental, potential fire products based on estimated intensity and duration of vegetation stress, which can be used as a proxy for assessment of potential fire danger. The Webfire Mapper, part of FAO's Global Fire Information Management System (GFIMS), initially developed by the University of Maryland and financially supported by NASA, provides near real-time information on active fires worldwide, detected by the MODIS rapid response system. The Webfire Mapper integrates satellite data with GIS technologies for active fire information. This information is available to the public through the website and e-mail alerts.

The European forest fire information system also provides information on current fire situations and forecasts for Europe and the Mediterranean area. Although global scale fire monitoring systems exist, an internationally standardized approach is required to create a globally comprehensive early fire warning system. Integration of existing fire monitoring systems could significantly improve fire monitoring and early warning capabilities. An information network must be developed to disseminate early warnings about wild-fire danger at both the global and local levels, to quickly detect and report fires, and to enhance rapid fire detection and classification capabilities at national and regional levels. The Global Early Warning System for Wild-Fires, which is under development as part of the Global Earth Observation System of Systems (GEOSS) effort, will address these issues.

3.1.4.1. Floods. Among natural hazards that are currently increasing in frequency, floods are the deadliest. This study shows there is inadequate coverage of flood warning and monitoring systems, especially in developing or least developed countries such as China, India, Bangladesh, Nepal, West Africa, and Brazil. At the local scale, there are several stand-alone warning systems, for example, in Guatemala, Honduras, El Salvador, Nicaragua, Zimbabwe, South Africa, Belize, Czech Republic, and Germany. However, they do not provide public access to information. The European Flood Alert System (EFAS), which is an initiative by EC-JRC, provides information on the possibility of river flooding occurring within the next three days. EFAS also provides an overview of current floods based on information received from the National Hydrological Services and the Global Runoff Data Center in Germany.

Floods are monitored worldwide from the Dartmouth Flood Observatory, which provides public access to major flood information, satellite images and estimated discharge. Orbital remote sensing (Advanced Scanning Microradiometer (AMSR-E and QuickScat)) is used to detect and map major floods worldwide. Satellite microwave sensors can monitor, at a global scale and on a daily basis, increases of floodplain water surface without cloud interference. The Dartmouth Flood Observatory provides estimated discharge and satellite images of major floods worldwide but does not provide forecasts of flood conditions or precipitation amounts that could allow flood warnings to be issued days in advance of events. NOAA provides observed hydrologic conditions of major US river basins and predicted values of precipitation for rivers in the United States. NOAA also provides information on excessive rainfall that could lead to flash-flooding and if necessary warnings are issued within 6 h in advance. IFnet Global Flood Alert System (GFAS) uses global satellite precipitation estimates for flood forecasting and warning. The GFAS website publishes useful public information for flood forecasting and warning, such as precipitation probability estimates, but the system is currently running on a trial basis. At a global scale, flood monitoring systems are more developed than flood early warning systems. For this reason, existing technologies for flood monitoring must be improved to increase prediction capabilities and flood warning lead times.

3.1.4.2. Severe weather, storms and tropical cyclones. At the global level, the World Weather Watch (WWW) and Hydrology and Water Resources Programmes coordinated by WMO provide global collection, analysis and distribution of weather observations, forecasts and warnings. The WWW is composed of the Global Observing System (GOS), which provides the observed meteorological data; the Global Telecommunications System (GTS), which reports observations, forecasts and other products and the Global Data Processing System (GDPS), which provides weather analyses, forecasts and other products. The WWW is an operational framework of coordinated national systems, operated by national governments.

The Tropical Cyclone Programme (TCP) is also part of the WWW. TCP is in charge of issuing tropical cyclones and hurricanes forecasts, warnings and advisories, and seeks to promote and coordinate efforts to mitigate risks associated with tropical cyclones. TCP has established tropical cyclone committees that extend across regional bodies (Regional Specialized Meteorological Centres (RSMC)), which, together with National Meteorological and Hydrological Services (NMHSs), monitor tropical cyclones globally and issue official warnings to the Regional Meteorological Services of countries at risk. Regional bodies worldwide have adopted standardized WMO-TCP operational plans and manuals, which promote internationally accepted procedures in terms of units, terminology, data and information exchange, operational procedures, and telecommunication of cyclone information. Each member of a regional body is normally responsible for its land and coastal waters warnings. A complete list of WMO members and RSMCs is available on the WMO-TCP website. WMO then collects cyclone information and visualizes it on world maps. The University of Hawaii collects information from WMO and provides online information on cyclone categories, wind speed, and current and predicted courses.

Although comprehensive coverage of early warning systems for storms and tropical cyclones is available, recent disasters such as Hurricane Katrina of 2005 have highlighted inadequacies in early warning system technologies for enabling effective and timely emergency response. There is a pressing need to improve communication between the sectors involved by strengthening the links between scientific research, organizations responsible for issuing warnings, and authorities in charge of responding to these warnings. While the WWW is an efficient framework of existing RSMC, NMHSs and networks, national capacities in most developing countries need improvements in order to effectively issue and manage early warnings. Action plans must also be improved.

Epidemics pose a significant threat worldwide, particularly in those areas that are already affected by other serious hazards, poverty, or under-development. Epidemics spread easily across country borders. Globalization increases the potential of a catastrophic disease outbreak: there is the risk that millions of people worldwide could potentially be affected.

A global disease outbreak early warning system is urgently needed. WHO is already working in this field through the Epidemic and Pandemic Alert and Response, which provides real-time information on disease outbreaks, and GOARN. The 192 WHO member countries, disease experts, institutions, agencies, and laboratories, part of an Outbreak Verification List, are constantly informed of rumoured and confirmed outbreaks. The WHO constantly monitors anthrax, avian influenza, Crimean-Congo hemorrhagic fever (CCHF), dengue haemorrhagic fever, Ebola haemorrhagic fever, Hepatitis, Influenza, Lassa fever, Marburg Haemorrhagic Fever, Meningococcal disease, plague, Rift Valley fever, Severe Acute Respiratory Syndrome (SARS), Tularaemia, and Yellow fever.

A global early warning system for animal diseases transmissible to humans was formally launched in July 2006 by the FAO, the World Organization for Animal Health (OIE), and WHO. The Global Early Warning and Response System for Major Animal Diseases, including Zoonoses (GLEWS) monitors outbreaks of major animal diseases worldwide. A malaria early warning system is not yet available and the need for system development is pressing, especially in Sub-Saharan Africa where malaria causes more than one million deaths every year. The IRI institute at Columbia University provides malaria risk maps based on rainfall anomaly, which is one of the factors influencing malaria outbreak and distribution, but no warning is disseminated to the potentially affected population. In addition, the Malaria Atlas Project (MAP) supported by the Wellcome Trust, the Global Fund to Fight AIDS, Tuberculosis and Malaria, the University of Oxford-Li Ka Shing Foundation Global Health Programme and others, aims to disseminate free, accurate and up-to-date information on malaria. The MAP is a joint effort of researchers from around the globe working in different fields (from public health to mathematics, geography and epidemiology). MAP produces and makes available a range of maps and estimates to support effective planning of malaria control at national and international scales.

Air pollution affects developing and developed countries without exception. For this reason, air quality monitoring and early warning systems are in place in most countries worldwide. Nevertheless, there is still a technological divide between developed and developing countries; in fact, these systems are most developed in the United States, Canada, and Europe. There are several successful cases to mention in Asia (Taiwan, China, Hong Kong, Korea, Japan, and Thailand), a few in Latin America (Argentina, Brazil, and Mexico City) and only one in Africa (Cape Town, South Africa).

Most of the existing systems focus on real-time air quality monitoring by collecting and analysing pollutant concentration measurements from ground stations. Satellite observation is extremely useful for aviation and tropospheric ozone monitoring, which is done by NASA and ESA. Air quality information is communicated mainly through web services. The US Environmental Protection Agency (EPA) provides an e-mail alert service (EPA AIRNow) only available in the US and the Ministry of Environment of Ontario, Canada, also provides e-mail alerts. The EPA AIRNow notification service provides air quality information in real-time to subscribers via e-mail, cell phone or pager, allowing them to take steps to protect their health in critical situations.

While current air quality information is provided by each of the air quality monitoring systems listed in the Appendix, few sources provide forecasts. The following agencies provide forecasts, which are fundamental for early warning: US EPA, ESA, Prev'Air, and the Environmental Agencies of Belgium, Germany, and Canada (see the Appendix). Prediction capability is an essential component of the early warning process. Existing air quality monitoring systems need to be improved in order to provide predictions to users days in advance so they can act when unhealthy air quality conditions occur.

3.2.2.1. Drought. Drought early warning systems are the least developed systems due its complex processes and environmental and social impacts. The study of existing drought early warning systems shows that only a few such systems exist worldwide. On a regional scale, the FEWS Net for Eastern Africa, Afghanistan, and Central America reports on current famine conditions, including droughts, by providing monthly bulletins that are accessible on the FEWS Net web page. For the United States, the US Drought Monitor (Svoboda et al., 2002) provides current drought conditions at the national and state level through an interactive map available on the website accompanied by a narrative on current drought impacts and a brief description of forecasts for the following week. The US Drought Monitor, a joint effort between the US Department of Agriculture (USDA), NOAA, the Climate Prediction Center, the University of Nebraska Lincoln and others, has become the best available product for droughts (Svoboda et al., 2002) . It has a unique approach that integrates multiple drought indicators with field information and expert input, and provides information through a single easy-to-read map of current drought conditions and short notes on drought forecast conditions. For China, the Beijing Climate Center (BCC) of the China Meteorological Administration (CMA) monitors drought development. Based on precipitation and soil moisture monitoring from an agricultural meteorological station network and remote-sensing-based monitoring from CMA's National Satellite Meteorological Center, a drought report and a map on current drought conditions are produced daily and made available on their website. The European Commission Joint Research Center (EC-JRC) provides publicly available drought-relevant information through the following real-time online maps: daily soil moisture maps of Europe; daily soil moisture anomaly maps of Europe; and daily maps of the forecasted top soil moisture development in Europe (7-day trend).

At a global scale, two institutions (FAO's Global Information and Early Warning System on Food and Agriculture (GIEWS) and Benfield Hazard Research Center of the University College London) provide some information on major droughts occurring worldwide. The FAO-GIEWS provides information on countries facing food insecurity through monthly briefing reports on crop prospects and food situations, including drought information, together with an interactive map of countries in crisis, available through the FAO website. Benfield Hazard Research Center uses various data to produce a monthly map of current drought conditions accompanied by a short description for each country. In addition, the WMO provides useful global meteorological information, such as precipitation levels, cloudiness, and weather forecasts, which are visualized on a clickable map on the WMO website.

Existing approaches for drought early warning must be improved. Due to the complex nature of droughts, a comprehensive and integrated approach (such as the one adopted by the US Drought Monitor) that would consider numerous drought indicators is required for drought monitoring and early warning. In addition, for large parts of the world suffering from severe droughts, early warning systems are not yet in place, such as in western and southern Africa, and in eastern Africa where FEWS Net is available but no drought forecast is provided. Parts of Europe (Spain, parts of France, southern Sweden, and northern Poland) are characterized by high drought risk but have no system in place. India, parts of Thailand, Turkey, Iran, Iraq, eastern China, areas of Ecuador, Colombia, and the south-eastern and western parts of Australia also require a drought warning system.

3.2.2.2. Desertification. The United Nations Convention to Combat Desertification (UNCCD), signed by 110 governments in 1994, aims to promote local action programs and international activities. National Action Programmes at the regional or sub-regional levels are key instruments for implementing the convention and are often supported by action programmes at sub-regional and regional levels. These programs lay out regional and local action plans and strategies to combat desertification. The UNCCD website provides a desertification map together with documentation, reports, and briefing notes on the implementation of action programs for each country worldwide. Currently no desertification early warning system is fully implemented, despite their potential to mitigate desertification.

3.2.2.3. Food security. The FAO's GIEWS supports policy-makers by delivering periodic reports through the GIEWS web page and an e-mail service. GIEWS also promotes collaboration and data exchange with other organizations and governments. The frequency of briefs and reports -which are released monthly or bimonthly -may not be adequate for early warning purposes. The WFP is also involved in disseminating reports and news on famine crises through its web service. No active dissemination is provided by WFP. Another service is FEWS net, a collaborative effort of the USGS, United States Agency for International Development (USAID), NASA, and NOAA, which reports on food insecurity conditions and issues watches and warnings to decision-makers. These bulletins are also available on their website. Food security prediction estimates and maps would be extremely useful for emergency response, resources allocation, and early warning. The Food Security and Nutrition Working Group (FSNWG) serves as a platform to promote the disaster risk reduction agenda in the region and provides monthly updates on food security through maps and reports available on its website. FEWSnet and FSNWG were instrumental in predicting the food crisis in 2010-2011 in the East African Region in a timely manner. Nevertheless these early warnings did not lead to early action to address the food crisis. If they had been used adequately, the impacts of the serious humanitarian crisis in the Horn of Africa could have been partially mitigated (Ververs, 2011) .

Nearly all efforts to cope with climate change or variability focus on either mitigation to reduce emissions or on adaptation to adjust to changes in climate. Although it is imperative to continue with these efforts, the on-going pace of climate change and the slow international response suggests that a third option is becoming increasingly important: to protect the population against the immediate threat and consequences of climate-related extreme events, including heat waves, forest fires, floods and droughts, by providing it with timely, reliable and actionable warnings. Although great strides have been made in developing climate-related warning systems over the past few years, current systems only deal with some aspects of climate related risks or hazards, and have large gaps in geographic coverage. Information exists for melting glaciers, lake water level, sea height and sea surface temperature anomalies, El Nino and La Nina. The National Snow and Ice Data Center (NSIDC)-Ice Concentration and Snow Extent provides near real-time data on daily global ice concentration and snow coverage. The USDA, in cooperation with the NASA and the University of Maryland, routinely monitors lake and reservoir height variations for approximately 100 lakes worldwide and provides online public access to lake water level data. Information on Sea Height Anomaly (SHA) and Significant Wave Height data are available from altimeter JASON-1, TOPEX, ERS-2, ENVISAT and GFO on a near-real time basis with an average 2-day delay. This information is provided by NOAA. Additionally, near real-time Sea Surface Temperature (SST) products are available from NOAA's GOES and POES, as well as NASA's EOS, Aqua and Terra. The International Research Institute (IRI) for Climate and Society provides a monthly summary of the El Nino and La Nina Southern Oscillation, providing forecast summary, probabilistic forecasts, and a sea surface temperature index. However, these systems are still far from providing the coverage and scope that is needed and technically feasible. Large parts of the world's most vulnerable regions are still not covered by a comprehensive early warning system. Most systems only deal with one aspect of climate-related risks or hazards, e.g., heat waves or drought. Finally, most systems do not cover the entire early warning landscape from collection of meteorological data to delivery and response of users. Recently, the World Meteorological Organization proposed a Global Framework for Climate Services, which aims to strengthen the global cooperative system for collecting, processing and exchanging observations and for using climate-related information (WMO, 2011).

Early warning technologies appear to be mature in certain fields but not yet in others. Considerable progress has been made, thanks to advances in scientific research and in communication and information technologies. Nevertheless, a significant amount of work remains to fill existing technological, communication, and geographical coverage gaps. Early warning technologies are now available for almost all types of hazards, although for some hazards (such as droughts and landslides) these technologies are still less developed. Most countries appear to have early warning systems for disaster risk reduction. However, there is still a technological and national capacity divide between developed and developing countries. From an operational point of view, some elements of the early warning process are not yet mature. In particular, it is essential to strengthen the links between sectors involved in early warning, including organizations responsible for issuing warnings and the authorities in charge of responding to these warnings, as well as promoting good governance and appropriate action plans.

It is generally recognized that it is fundamental to establish effective early warning systems to better identify the risk and occurrence of hazards and to better monitor the population's level of vulnerability. Although several early warning systems are in place at the global scale in most countries for most hazard types, there is the need ''To work expeditiously towards the establishment of a worldwide early warning system for all natural hazards with regional nodes, building on existing national and regional capacity such as the newly established Indian Ocean Tsunami Warning and Mitigation System'' (2005 UN World Summit Outcome). By building upon ongoing efforts to promote early warning, a multi-hazard early warning system will have a critical role in preventing hazardous events from turning into disasters. A globally comprehensive early warning system can be built, based on the many existing systems and capacities. This will not be a single, centrally planned and commanded system, but a networked and coordinated assemblage of nationally owned and operated systems. It will make use of existing observation networks, warning centres, modelling and forecasting capacities, telecommunication networks, and preparedness and response capacities (United Nations, 2006) . A global approach to early warning will also guarantee consistency of warning messages and mitigation approaches globally thus improving coordination at a multi-level and multi-sector scale among the different national actors such as the technical agencies, the academic community, disaster managers, civil society, and the international community. The next section provides an analysis of existing global early warning/monitoring systems that aggregate multi-hazard information.

This section presents the results of a comparative analysis of multi-hazard global monitoring/ early warning systems. The aim of this analysis is to assess the effectiveness of existing global scale multi-hazard systems and define the set of needs suggested by comparing existing services. It assesses existing monitoring/early warning systems (Singh and Grasso, 2007) , chosen to be multi-hazard with global coverage, such as WFP's (the UN food aid agency's) HEWS; AlertNet, the humanitarian information alert service by Reuters; ReliefWeb, the humanitarian information alert service by UN-OCHA; GDACS (Global Disaster Alert and Coordination System), a joint initiative of the United Nations Office for the Coordination of Humanitarian Affairs (UN-OCHA) and the EC-JRC; and USGS ENS. These systems have been analysed for the type of events covered, variety of output/ communication forms and range of users served.

These systems cover a range of hazards and communicate results using a variety of methods. USGS ENS only provides information on earthquakes and volcanic eruptions through the Volcano Hazards Program (VHP) in collaboration with Smithsonian Institution. ReliefWeb focuses on natural hazards (earthquakes, tsunamis, severe weather, volcanic eruptions, storms, floods, droughts, cyclones, insect infestation, fires, and technological hazards) and health; AlertNet additionally provides information on food insecurity and conflicts. GDACS provides timely information on natural hazards (earthquakes, tsunamis, volcanic eruptions, floods, and cyclones). HEWS informs users on earthquakes, severe weather, volcanic eruptions, floods, and locusts.

Existing systems, such as HEWS, post the information on a website, provide mobile phone and RSS services, and GDACS sends emails, SMS and faxes to users. The GDACS notification service mostly addresses humanitarian organizations, rescue teams or aid agencies. AlertNet provides information to users through a web service, e-mail, SMS and reports. ReliefWeb uses web services, e-mail and reports to disseminate information to users. ReliefWeb and AlertNet also use new communications tools (Facebook and Twitter). OCHA's Virtual On-site Operations Coordination Centre (Virtual-OSOCC) enables real-time information exchange by all actors of the international disaster response community during the early phases following disasters. This service has been integrated within GDACS but is restricted to disaster managers.

The only natural event notification provided by USGS e-mail service is earthquakes. HEWS offers no e-mail notifications for natural events. AlertNet and ReliefWeb inform users on natural hazards and on health, food and security issues. USGS, ReliefWeb, AlertNet and GDACS serve a wide a variety of users such as international organizations, humanitarian aid, policy/decision-makers, and civil society (Fig. 4) .

The optimal global coverage multi-hazard system has to be as comprehensive as possible in terms of content, output and range of users. It will enhance existing systems by streaming data and information from existing sources and it will deliver this information in a variety of user-friendly formats to reach the widest range of users. By building on existing systems the multi-hazard system will inherit both the technological and geographical coverage gaps and limitations of existing early warning systems. The review analysis performed in the section ''Inventory of early warning systems'' has shown that for some hazards (such as droughts and landslides) these technologies are still less developed and for tsunamis these systems are still under development for areas at risk. The analysis has shown that there is still a technological and national capacity divide between developed and developing countries. From an operational point of view, the links and communication networks between all sectors involved (organizations responsible for issuing warnings and the authorities in charge of responding to these warnings) need improvement. Similarly, good governance and appropriate action plans need to be promoted.

Overcoming these gaps and enhancing, integrating, and coordinating existing systems is the first priority for the development of a global scale multi-hazard early warning system .

Early warning technologies have greatly benefited from recent advances in communication and information technologies and an improved knowledge on natural hazards and the underlying science. Nevertheless many gaps still exist in early warning technologies and capacities, especially in the developing world, and a lot more has to be done to develop a global scale multi-hazard system. Operational gaps need to be filled for slow-onset hazards both in monitoring, communication and response phases. Effective and timely decision making is needed for slow onset hazards. Below are some recommendations:

(1) Fill existing gaps: The Section 3 identified the weaknesses and gaps in existing early warning systems. Technological, geographical coverage, and capacity gaps exist, in addition to operational gaps for slow-onset hazards. In particular, actions need to be taken to improve prediction capabilities for landslides hazard aimed at developing a landslides early warning system. Likewise, there is a pressing need to improve existing prediction capabilities for droughts. A global early fire warning system is not yet in place, the tsunami early warning systems for the Indian Ocean and the Caribbean are not yet fully operational and a desertification early warning system has not been developed yet. There are ongoing efforts to develop these systems, such as the GFMC effort for the global fire EWS, the Indian Ocean Tsunami Warning System operated by Indonesia and a NOAA led effort in the Caribbean. A malaria early warning system is mandatory for Africa, where one million deaths occur every year due to malaria. Climate variability impacts need to be monitored within a global and coordinated effort, and the Global Framework for Climate Services needs to be further elaborated and operationalized. Local earthquake early warning systems applications are needed in high seismic risk areas, where early warning systems are not yet in place. Air quality and flood systems require improvements in prediction capabilities. Dust storms and transboundary early warning systems do not yet exist. A coordinated volcanic early warning system that would integrate existing resources is also needed as well as an increase in coverage of volcanic observatories. Particular attention should be paid to fill gaps in decision making processes for slow-onset hazards. Their extent and impact are challenging to quantify. For this reason, actions and response are far more difficult tasks for slow-onset hazards than they are for other natural hazards. An institutional mechanism to regularly monitor and communicate slow-onset changes is needed to keep changes under review and to enable rational and timely decisions to be taken based on improved information. (2) Build capacity: The evaluation study of existing early warning systems (the Section 3) highlighted that a technological divide between developed and developing countries still exists. It is critical to develop basic early warning infrastructures and capacities in parts of the developing world most affected by disasters; it is also important to promote education programs on disaster mitigation and preparedness and integrate disaster mitigation plans into the broader development context. Poor countries suffer greater economic losses from disasters than rich countries. Development plays a key role and has a significant impact on disaster risk. Almost 85 percent of the people exposed to the deadliest hazards, earthquakes, floods, cyclones and droughts live in the developing world. The impact of disasters is mostly influenced by previous development choices. By integrating disaster mitigation strategies into planning and policies, the effects of disasters can be sensibly reduced and managed. ''Disaster risk is not inevitable, but on the contrary can be managed and reduced through appropriate development actions'' (United Nations Development Programme-UNDP, 2004) . It is through ''risk-sensitive development planning that countries can help reduce disaster risks''. Key targets for capacity building include the following:

1. Developing national research, monitoring and assessment capacity, including training in assessment and early warning. 2. Supporting national and regional institutions in data collection, analysis and monitoring of natural and man-made hazards. 3. Providing access to scientific and technological information, including information on stateof-the-art technologies. 4. Education and awareness-raising, including networking among universities with programmes of excellence in the field of the emergency management. 5. Organizing of training courses for local decision-makers and communities. 6. Bridging the gap between emergency relief and long-term development.

(3) Bridge the gaps between science and decision making, and strengthen coordination and communication links: Scientific and technological advances in modelling, monitoring and predicting capabilities could bring immense benefits to early warning if science were effectively translated into disaster management actions. Bridging the gap between scientific research and decision making will make it possible to fully exploit capacities of early warning technologies for societal benefit. The major challenge is to ensure that early warnings result in prompt responses by governments and potentially the international community. This requires that information be effectively disseminated in an accessible form down to the end user. This is achievable by adopting standard formats and easy-to-use tools for information dissemination, such as interactive maps, emails, SMS, etc. The adoption of standard formats (such as the Common Alerting Protocol CAP) for disseminating and exchanging information has to be promoted. The advantage of standard format alerts is their compatibility with all information systems, warning systems, media, and most importantly, with new technologies such as web services. The adoption of standard formats guarantees consistency of warning messages and is compatible with all types of information systems and public alerting systems, including broadcast radio and television as well as public and private data networks, with multi-lingual warning systems and emerging technologies. This would easily replace specific application oriented messages and will allow the merging of warning messages from several early warning systems into a single multihazard message format. Finally, it is critical to strengthen coordination and communication links by defining responsibility mechanisms and appropriate action plans. More often, time-sequenced warning messages are released in early warning processes, implying a decrease in warning times available for action and in reliability of the information. This trade-off needs to be addressed.

The content and views expressed in this publication are those of the authors and do not necessarily reflect the views or policies, or carry the endorsement of the contributory organizations or the United Nations Environment Programme (UNEP). The designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of UNEP concerning the legal status of any country, territory or city or its authorities, or concerning the delimitation of its frontiers and boundaries. Reference to a commercial company or product in this publication does not imply the endorsement of UNEP.

& Maps, photos, and illustrations as specified.

This publication may be reproduced in whole or in part and in any form for educational or non-profit purposes without special permission from the copyright holder, provided acknowledgement of the source is made. UNEP would appreciate receiving a copy of any publication that uses this publication as a source. No use of this publication may be made for resale or any other commercial purpose whatsoever without prior permission in writing from UNEP. Applications for such permission, with a statement of purpose and intent of the reproduction, should be addressed to the Director, Division of Communications and Public Information (DCPI), UNEP, P.O. Box 30552, Nairobi 00100, Kenya. The use of information from this publication concerning proprietary products for publicity or advertising is not permitted.

Citation: UNEP (2012 Adaptation: (1) Actions taken to help communities and ecosystems cope with changing climate conditions (UNFCCC).

(2) Genetically determined characteristic that enhances the ability of an organism to cope with its environment (CBD). Aerosols: A collection of airborne solid or liquid particles, with a typical size between 0.01 and 10 mm, that resides in the atmosphere for at least several hours. Aerosols may be of either natural or anthropogenic origin. Air quality: Smog is the product of human and natural activities such as industry, transportation, wild-fires, volcanic eruptions, etc. and can have serious effects on human health and the environment. US EPA uses and Air Quality Index (AQI) which is calculated on five major air pollutants regulated by the Clean Air Act: ground-level ozone, particle pollution (also known as particulate matter), carbon monoxide, sulfur dioxide, and nitrogen dioxide. For each of these pollutants, EPA has established national air quality standards to protect public health. Groundlevel ozone and airborne particles are the two pollutants that pose the greatest threat to human health in this country. Biodiversity: The variety of life on Earth, including diversity at the genetic level, among species and among ecosystems and habitats. It includes diversity in abundance, distribution and in behaviour. Biodiversity also incorporates human cultural diversity, which can both be affected by the same drivers as biodiversity, and itself has impacts on the diversity of genes, other species and ecosystems. Biofuels: Fuel produced from dry organic matter or combustible oils from plants, such as alcohol from fermented sugar, black liquor from the paper manufacturing process, wood and soybean oil. Biomass: Organic material, both above ground and below ground, and both living and dead, such as trees, crops, grasses, tree litter and roots. Capacity: The combination of all the strengths, attributes and resources available within a community, society or organization that can be used to achieve agreed goals. Comment: Capacity may include infrastructure and physical means, institutions, societal coping abilities, as well as human knowledge, skills and collective attributes such as social relationships, leadership and management. Capacity also may be described as capability. Capacity assessment is a term for the process by which the capacity of a group is reviewed against desired goals, and the capacity gaps are identified for further action. Capacity building: Process of developing the technical skills, institutional capability, and personnel. Climate change: Change of climate, which is attributed directly or indirectly to human activity that alters the composition of the global atmosphere and which is in addition to natural climate variability observed over comparable time periods. Climate Variability: Variations in the mean state and other statistics (such as standard deviations and the occurrence of extremes) of the climate on all temporal and spatial scales beyond that of individual weather events. Variability may be due to natural internal processes in the climate system (internal variability), or to variations in natural or anthropogenic external forcing (external variability). Common Alerting Protocol: The Common Alerting Protocol (CAP) provides an open, nonproprietary digital message format for all types of alerts and notifications. It does not address any particular application or telecommunications method. The CAP format is compatible with emerging techniques, such as Web services, as well as existing formats including the Specific Area Message Encoding (SAME) used for the United States' National Oceanic and Atmospheric Administration (NOAA) Weather Radio and the Emergency Alert System (EAS). Cost-Benefit Analysis: A technique designed to determine the feasibility of a project or plan by quantifying its costs and benefits.

Cyclone: An atmospheric closed circulation rotating counter clockwise in the Northern Hemisphere and clockwise in the Southern Hemisphere. Deforestation: The direct human-induced conversion of forested land to non-forested land. Desertification: Degradation of land in arid, semi-arid and dry sub-humid areas, resulting from various factors, including climatic variations and human activities. Disaster: A serious disruption of the functioning of a community or a society involving widespread human, material, economic or environmental losses and impacts, which exceeds the ability of the affected community or society to cope using its own resources. Comment: Disasters are often described as a result of the combination of: the exposure to a hazard; the conditions of vulnerability that are present; and insufficient capacity or measures to reduce or cope with the potential negative consequences. Disaster impacts may include loss of life, injury, disease and other negative effects on human physical, mental and social well-being, together with damage to property, destruction of assets, loss of services, social and economic disruption and environmental degradation. Disaster risk: The potential disaster losses, in lives, health status, livelihoods, assets and services, which could occur to a particular community or a society over some specified future time period. Comment: The definition of disaster risk reflects the concept of disasters as the outcome of continuously present conditions of risk. Disaster risk comprises different types of potential losses which are often difficult to quantify. Nevertheless, with knowledge of the prevailing hazards and the patterns of population and socio-economic development, disaster risks can be assessed and mapped, in broad terms at least. Disaster risk reduction: The concept and practice of reducing disaster risks through systematic efforts to analyse and manage the causal factors of disasters, including through reduced exposure to hazards, lessened vulnerability of people and property, wise management of land and the environment, and improved preparedness for adverse events. Comment: A comprehensive approach to reduce disaster risks is set out in the United Nations-endorsed Hyogo Framework for Action, adopted in 2005, whose expected outcome is ''The substantial reduction of disaster losses, in lives and the social, economic and environmental assets of communities and countries.'' The International Strategy for Disaster Reduction (ISDR) system provides a vehicle for cooperation among Governments, organisations and civil society actors to assist in the implementation of the Framework. Note that while the term ''disaster reduction'' is sometimes used, the term ''disaster risk reduction'' provides a better recognition of the ongoing nature of disaster risks and the ongoing potential to reduce these risks. Droughts: A period of abnormally dry weather sufficiently prolonged for the lack of water to cause serious hydrologic imbalance in the affected area. Early warning system: The set of capacities needed to generate and disseminate timely and meaningful warning information to enable individuals, communities and organizations threatened by a hazard to prepare and to act appropriately and in sufficient time to reduce the possibility of harm or loss. Comment: This definition encompasses the range of factors necessary to achieve effective responses to warnings. A people-centred early warning system necessarily comprises four key elements: knowledge of the risks; monitoring, analysis and forecasting of the hazards; communication or dissemination of alerts and warnings; and local capabilities to respond to the warnings received. The expression ''end-to-end warning system'' is also used to emphasize that warning systems need to span all steps from hazard detection through to community response. Earth observation: Earth Observation, through measuring and monitoring, provides an insight and understanding into Earth's complex processes and changes. EO include measurements that can be made directly or by sensors in-situ or remotely (i.e. satellite remote sensing, aerial surveys, land or ocean-based monitoring systems, Fig. 3 ), to provide key information to models or other tools to support decision making processes. Earthquakes: Earthquakes are due to a sudden release of stresses accumulated around the faults in the Earth's crust. This energy is released through seismic waves that travel from the origin zone, which cause the ground to shake. Severe earthquakes can affect buildings and populations. The level of damage depends on many factors such as intensity of the earthquake, depth, vulnerability of the structures, and distance from the earthquake origin. Ecosystem: Dynamic complex of plant, animal, microorganism communities and their non-living environment, interacting as a functional unit. Ecosystems are irrespective of political boundaries.

El Niñ o-southern oscillation: A complex interaction of the tropical Pacific Ocean and the global atmosphere that results in irregularly occurring episodes of changed ocean and weather patterns in many parts of the world, often with significant impacts over many months, such as altered marine habitats, rainfall changes, floods, droughts, and changes in storm patterns. Comment: The El Nino part of the El Nino-Southern Oscillation (ENSO) phenomenon refers to the well-above average ocean temperatures that occur along the coasts of Ecuador, Peru and northern Chile and across the eastern equatorial Pacific Ocean, while La Nina part refers to the opposite circumstances when well-below-average ocean temperatures occur. The Southern Oscillation refers to the accompanying changes in the global air pressure patterns that are associated with the changed weather patterns experienced in different parts of the world. Emergency management: The organization and management of resources and responsibilities for addressing all aspects of emergencies, in particular preparedness, response and initial recovery steps. Comment: A crisis or emergency is a threatening condition that requires urgent action. Effective emergency action can avoid the escalation of an event into a disaster. Emergency management involves plans and institutional arrangements to engage and guide the efforts of government, non-government, voluntary and private agencies in comprehensive and coordinated ways to respond to the entire spectrum of emergency needs. The expression ''disaster management'' is sometimes used instead of emergency management. Extensible Markup Language: A markup language that defines a set of rules for encoding documents in a format that is both human-readable and machine-readable. E-waste: A generic term encompassing various forms of electrical and electronic equipment that has ceased to be of value and is disposed of. A practical definition of e-waste is ''any electrically powered appliance that fails to satisfy the current owner for its originally intended purpose.'' False Alarm: In the context of Early Warning Systems, a false alarm is defined as the situation in which an alarm is activated when it should not have been.

Fine Particle: Particulate matter suspended in the atmosphere less than 2.5 mm in size (PM2.5).

Floods: An overflow of water onto normally dry land. Theinundation of a normally dry area caused by rising water in an existing waterway, such as a river, stream, or drainage ditch. Ponding of water at or near the point where the rain fell. Flooding is a longer term event than flash flooding: it may last days or weeks. Floods are often triggered by severe storms, tropical cyclones, and tornadoes. Food security: When all people at all times have access to sufficient, safe, nutritious food to maintain a healthy and active life. Forecast: Definite statement or statistical estimate of the likely occurrence of a future event or conditions for a specific area. Comment: In meteorology a forecast refers to a future condition, whereas a warning refers to a potentially dangerous future condition. Forest: Land spanning more than 0.5 ha with trees higher than 5 m and a canopy cover of more than 10 percent, or trees able to reach these thresholds in situ. It does not include land that is predominantly under agricultural or urban land use. Gaussian distribution: The Gaussian (normal) distribution was historically called the law of errors.

It was used by Gauss to model errors in astronomical observations, which is why it is usually referred to as the Gaussian distribution. Geological hazard: Geological process or phenomenon that may cause loss of life, injury or other health impacts, property damage, loss of livelihoods and services, social and economic disruption, or environmental damage. Comment: Geological hazards include internal earth processes, such as earthquakes, volcanic activity and emissions, and related geophysical processes such as mass movements, landslides, rockslides, surface collapses, and debris or mud flows. Hydrometeorological factors are important contributors to some of these processes. Tsunamis are difficult to categorize; although they are triggered by undersea earthquakes and other geological events, they are essentially an oceanic process that is manifested as a coastal water-related hazard. Within this report, tsunamis are included in the geological hazards group. Geographic Information System: A computerized system organizing data sets through a geographical referencing of all data included in its collections.

The potential for earthquake early warning in Southern California

Oregon Department of Environmental Quality. Water Quality Credit Trading in Oregon: A Case Study Report

Disaster Management Support Group (DMSG) CEOS Disaster Management Support Group Report. The Use of Earth Observing Satellites for Hazard Support: Assessments & Scenarios. Final Report of the CEOS Disaster Management Support Group

Natural Disaster Hotspots: A Global Risk Analysis

Mexico City Seismic Alert System

Environmental Change Monitoring by Geoinformation Technology for Baghdad and its Neighboring Areas

GEOSS 10-year Implementation Plan Reference Document

Early warning systems

Early Warning of Wildland Fires

Usable Science: Early warning systems: Do's and Don'ts

Automated decision procedure for earthquake early warning

Automated decision procedure for earthquake early warning

A model for a seismic computerized alert network

An automatic processing system for broadcasting earthquake alarms

Global Observations of the Land, A proposed theme to the IGOS-Partnership-Version 2

Probability Theory: The Logic of Science

JMA Earthquake Early Warning

Real-time seismology and earthquake damage mitigation

Global forest fire watch: wildfire potential, detection, monitoring and assessment

Oil Spill Detection and Monitoring from Satellite Image

On a rational strong motion index compared with other various indices

The Role Of

A new method of quickly estimating epicentral distance and magnitude from a single seismic record

Prediction, Science, Decision Making and the Future of Nature

University of Valencia,EOS.D2C. Chinese Academy of Forestry

Effective Early Warning -use of hazard maps as a tool for effective risk communication among policy makers and communities -Session organized by the Government of Japan

Global Environmental Alert Service

Digital change detection techniques using remotely sensed data

The drought monitor

ICT in pre-disaster awareness and preparedness. Presented at the APT-ITU Joint Meeting for ICT Role in Disaster Management

Early Warning Systems for Natural Disasters Reduction

UNDP-United Nations Development Programme Bureau for Crisis Prevention and Recovery

United Nations Convention to Combat Desertification

The East African food crisis: did regional early warning systems function

Climate Knowledge for Action: A Global Framework for Climate Services-Empowering the most vulnerable

Quick and reliable determination of magnitude for seismic early warning

A virtual sub-network approach to earthquake early warning