Early Warning – Some Recent Developments

NEUSSNER, Olafa

a Global Initiative for Disaster Risk Management (GIDRM), Deutsche Gesellschaft fuer Internationale Zusammenarbeit (GIZ), Manila, Philippines, e-mail:olafneus@gmx.net

Abstract — Recent advances in early warning have been observed in science and technology as well as the overall effectiveness of end-to-end early warning systems. The accuracy in predicting earthquakes is slowly increasing, but it is still far from being applicable to routine early warning systems. Tests of landslide warning systems based on motion sensors of slopes are progressing and might be used more widespread in the coming years. Fast communication is not a real issue for today’s technology, but spreading alert information quickly to large communities without access to modern communication equipment remains a problem. However, some experiences of locally anchored early warning systems point to higher effectiveness if they are managed decentralized within communities. Though EWS have to be adjusted to local circumstances the growing number of temporary visitors (e.g. migrants, tourists) in the world, require some type of international standard easily understood by everyone. While the wording may be local plus internationally used languages (e.g. English), signs/icons should follow international conventions.

Keywords — early warning systems, earthquakes, river floods, storm surge, storm, innovations in science and technology

1 Introduction

Most, if not all extreme natural events leading to disasters have precursors which may be used to warn of the approaching danger, take appropriate action and mitigate the effects of such an event. Over the last decades a lot of progress has been made with the establishment of Early Warning Systems (EWS). Many more are in place today and scientific as well as technical developments led to improved quality of hazard detection and communications. However, still many disasters cause avoidable casualties and damages and considerable effort is invested in analysing what went wrong and how systems can be improved.

In this article recent developments in early warning are described using two different ways: assessing progress concerning specific hazards and by analysing the four elements of EWS (risk knowledge, monitoring and warning service, communication, response capacity).

The first of the two options, hazard specific systems, deal predominantly with scientific methods of determining dangerous natural developments. Advances in space technology and more widespread use of ground based sensors led to better understanding of natural phenomena increasing the precision of forecasts.

The second option of describing EWS looks at the complete warning chain starting from risk knowledge (e.g. hazard maps) to the final response capacity (e.g. suitable evacuation centres). The whole EWS is only as strong as the weakest element in the chain and recent calamities like Haiyan in the Philippines demonstrate that weakness in one part can lead to many casualties.

The International Disaster Risk Conference in Davos, Switzerland, in August 2014 contributed to an exchange of professionals involved in early warning with many presentations and lively discussions about new insights. This paper is not intended to give a complete overview of news on early warning efforts. It concentrates on those topics presented in Davos and adds some of the experiences of GIZ.

2 Hazard Spezific Aspects of Early Warning

Observing nature and specifically precursors of potentially dangerous events is the domain of science and technology. The combination of ground based sensors, earth observation from space and computer simulations have expanded the capabilities of institutions tasked with early warning over the last decades. However, there are still many challenges ahead, the biggest probably forecasting earthquakes.

Though science is the main pillar of forecasting in some cases community-based observations are complementing the often expensive and complicated technical systems. Especially in the context of smaller communities in developing countries, inexpensive, simpler solutions with the involvement of volunteer observers have proven to be adequate approaches for some hazards.

2.1 Earthquakes

For all globally important natural hazards some methods of practical applicable early warning are available (storm, floods, volcano, tsunami, drought), but strong earthquakes still elude forecasts to a large extent. Researchers investigate many different phenomena associated with precursors of earthquakes, but though some of them have a significant correlation with earthquakes, the spatial and temporal precision is far from applicable for concrete preparedness measures. In recent years it has been tried to combine many such indicators leading towards a marked increase in accuracy.

Three researchers (Ouzounov, 2014: 168; Perminov, 2014: 170; Pulinets, 2014: 175) took a large array of indicators, most of them space based, and applied the “Lithosphere-Atmosphere-Ionosphere Coupling (LAIC) model” (Figure 1) with these data. They tested the model in recent years with earthquakes of magnitude 5.5+.

Figure 1: Lithosphere-Atmosphere-Ionosphere Coupling (LAIC) model (Pulinets, 2009)

The results indicate that large earthquakes were mostly predicted with a temporal precision of some weeks and a spatial resolution of a circle of 400km diameter. This is a marked improvement compared to previous applications of the LAIC model and other methods which often did not even show a significant correlation between an indicator and actual earthquakes. For practical applicability especially the date of an expected strong earthquake is important and if the researchers manage to reach this precision an early warning system may be established.

2.2 Seasonal Rain Forecasts and Floods

Disaster institutions and agriculture are interested in seasonal weather forecasts. They need to allocate resources and take precautions for extreme conditions and events. Seasonal forecasts are provided in many countries and representatives from the Nigerian Meteorological Agency (NiMet) reported about their method of seasonal rain prediction (SRP) in the International Disaster Risk Conference in Davos in August 2014 (Alozie, 2014: 44). NiMet produces SRPs with about one month lead time. Disaster risk managers like the National Emergency Management Agency, (NEMA) are known to use seasonal climate forecasts as a starting point for preparedness and resource planning, and to inform agricultural management decisions such as crop and variety choice.

Seasonal forecasts as practised in Nigeria are a good basis for preparedness for floods and other rain-related extreme events, but the backbone for early warning are short-term weather predictions used for flood alerts.

Flood Early Warning Systems (FEWS) for inland/river floods do not pose a real scientific challenge any more. All major rivers with significant numbers of residents in flood prone areas are equipped with EWS and technically they can easily achieve a perfect record (no missed floods, warning for all actual floods). However, certain constellations are still not fully covered by FEWS, such as flash floods, ponding, and floods in smaller rivers.

Toronto experienced torrential rains in July 2013 when two thunderstorms collided over the city. This resulted in serious transportation interruptions and it was the most costly calamity in the history of the Canadian city. According to the researchers (Asgary, 2014: 109) the population was not warned of the heavy rains and accordingly did not take any advance action. The researchers found that a better early warning system would have reduced the losses caused by the floods.

Many smaller rivers all over the world experience regular floods, but they do not have FEWS and residents may only receive a general warning of inclement weather conditions approaching. As the number of such small rivers is huge a full coverage of all these rivers with the spectrum of available technologies would be economically not viable as the establishment and maintenance is too costly. Furthermore, complex systems require expertise (e.g. hydrologists) for operation and the permanent assignment of these specialists is also not sustainable. Low cost and simple EWS have been tried and tested in several places and the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) promoted them together with the Philippine weather agency PAGASA (Philippine Atmospheric, Geophysical and Astronomical Service Administration)(Hernando, 2013; GIZ, 2012). A total of 16 such local FEWS were established from 2007 – 2014 in the country. With guidance and advice from PAGASA and GIZ municipalities, cities and provinces set up local FEWS and are operating them now with their own staff and resources. The model also involves volunteers to detect water level and rain upstream (Figure 2). Some of them serve as a backup if the automated devices fail. By and large the FEWS have proven to be reliable and they issued many flood warnings giving the population additional hours to prepare their households and evacuate (Neussner, 2012, 153). In a GIZ commissioned survey in 2011 two thirds of the respondents of flood affected areas said that they feel their lives are better protected than before.

Figure 2: Leyte, Philippines. Left: Community volunteer with the display of a tipping bucket rain gauge on the roof of the house; right: a community member paints a staff gauge at the pillar of a bridge.

Currently GIZ is in the process of replicating the local FEWS model in other Asian countries (Myanmar, Thailand, Vietnam)1.

2.3 Storm and Storm Surge

Tropical cyclones are tracked and their paths predicted with days advance notice to those expected to be affected by the storm. Very often wind speeds, location and time are forecasted quite well, but this does not necessarily apply to the accompanying rain or the occurrence and height as well as extend of storm surges. The most deadly disaster of 2013, the typhoon Haiyan (local Philippine name Yolanda) crossed the Philippines on 8 November 2013 and illustrated these problems as described by researchers (Tohoku University, 2014; Neussner, 2014).

The weather bureau Philippine Atmospheric, Geophysical and Astronomical Service Administration (PAGASA) issued severe wind warnings three days before the storm made landfall. On 7 November a storm surge warning was added, but warning of heavy rains and landslides were already issued earlier. All this would have given enough time to take appropriate action and leave the danger zone, but 30% (Tohoku University, 2014) or in some places up to 50% (Neussner, 2014) did not evacuate.

Interviews with residents revealed a multitude of reasons why people did not leave their homes near the coast. Many apparently underestimated the seriousness of the upcoming storm and thought that they are safe in their houses. As approximately 95% of the deceased were killed by the storm surge in Leyte, the combination of underestimating the inundation area, unclear messages (and some other reasons) highlighted the need for improvements needed especially with regards to the inundation danger zone. The survey results and computer modelling can improve the maps considerably. There are already efforts on the way to change the wording of warning messages to simpler terms in order to avoid communication problems in the future.

2.4 Landslides

Though landslides are very limited in their extent they are often devastating in a small area. They may be triggered by earthquakes, excessive rain or even human activities. In most cases they happen with precursor signs which may be used for early warning purposes. Typically a slope moves slowly a little before a big failure happens. The small, creeping movements are often indicated by tilted trees or fences, crevices in the ground or changes in the behaviour of small streams (increased turbidity). The detection of dangerous small movements with different types of motion sensors is currently being tested in some parts of the world. The installed systems are in the development stage and it will take many years before they can be used on a routine basis for reliable predictions and warnings.

The Deutsche Gesellschaft fuer Internationale Zusammenarbeit (GIZ) promotes the development of landslide early warning systems in the Philippines and has installed five sets of motion sensors in Leyte Island2(examples in Figure 3). The selected sites show signs of previous motions, normally cracks in the top soil.

Figure 3: Landslide motion sensors in Southern Leyte, Philippines. Left: Installation and functionality test. Right: Location of sensor in tube near crevices (red arrows) indicating recent soil movements.
Figure 4: Flume tests of landslide motion sensors in the facilities of the Department of Public Works and Highways in Manila, Philippines.

The sensors detect tilting, acceleration and changes in atmospheric pressure and the latest generation of sensors can also monitor soil humidity. Data are transmitted via radio to operation centers nearby. The sensors are functioning but no actual suspicious movement happened in the test sites up to now. They were also tested under laboratory conditions (Figure 4).

Further testing and development of the systems is needed. Initial thresholds concerning suspicious movements have been set, but they have not been verified yet. This will be important as the ratio between false alarms and actual landslides is determined by these thresholds. Only if a reasonable reliability is achieved landslide warning systems could be established in bigger numbers for routine operations.

2.5 Atmospheric threats

A topic usually not receiving much attention in the disaster risk community was presented by Fernandez de Arroyabe (Fernandez de Arroyabe, 2014: 247) in Davos. There are a number of specific health risks linked to atmospheric conditions. Variations in oxygen content, air humidity, atmospheric pressure, sun radiation, ozone levels, electromagnetic fields and others may pose a threat to some persons and Fernandez described a correlation between an influenza outbreak in Iberian Peninsula in 2004/5 and meteorological contrast index parameters. As weather is predicted by meteorologists the data may be utilized for biometeorological forecasting, but how individuals are reacting to changes in atmospheric conditions depends largely on their specific condition. Therefore Fernandez suggests establishing individually customized early warning systems. This could be facilitated with Apps using individual bio-data and processing them with parameters from weather forecasts resulting in individualized biometeorological alerts.

3 Effectiveness of early warning systems

Though the methods of detecting approaching dangerous natural events are very hazard specific, the other elements of a functioning Early Warning System (EWS) are common for all types of hazards. UNISDR facilitated the publication of a checklist for EWS (UNISDR, 2006) enumerating four elements constituting an operational EWS:

  • Risk Awareness
  • Monitoring And Warning Service
  • Dissemination And Communication
  • Response Capability

Furthermore, governance and institutional arrangements were identified as a cross cutting issue. EWS are only working properly if all for elements are in place, operational and handled by competent staff. With this the EWS might disseminate timely warnings, but the most important factor for successful preparations for a potentially disastorous event are the people in the affected area. The term ‘People Centered Early Warning System’ has been coined to highlight that EWS have to be established with and for the communities they are expected to serve. Failure to do so might produce technically well working systems issuing warnings residents do not understand or ignore because of mistrust.

This chapter looks at the elements of EWS and some recent developments, among them those presented in the IDRC in Davos in August 2014.

3.1 Risk awareness

A good EWS is based on sound risk knowledge with designated institutions responsible for collecting and providing such information to the government, the public (and the international community). It covers hazard characteristics (intensity, frequency, probability and geographic extent), vulnerability analysis of communities who might be affected, and also a risk assessment derived from the interaction of vulnerability and hazard. An example of incomplete hazard information leading to many casualties is the tropical cyclone Haiyan in Leyte, Philippines, on 8 November 2013. The high death toll has more than one reason, but a lack of risk awareness certainly contributed substantially to the magnitude of the catastrophe. It has been estimated that 30% (Tohoku University, 2014) or in some places up to 50% (Neussner, 2014) of the residents of coastal areas inundated by the storm surge did not evacuate. The storm surge proofed to be the main cause of death. Approximately 95% (= 5000 persons) of the deceased in Leyte died from the surge and 5% from the high wind speed of the storm (collapsing buildings, flying debris, falling trees)(Neussner, 2014).

One reason for the non-evacuation might be the official storm surge hazard map. An example is displayed in Figure 5 together with a map of the actual inundation area. Obviously it was believed that many people lived in safe areas although they did not.

Figure 5: Tacloban, Leyte, Philippines. Left: Storm surge hazard map by PAGASA, potential inundation area in purple (1m > 4m); right: actual inundation on 8 Nov. 2013 from physical signs and interviews (Tohoku, 2014)

The maps show clear differences between the official hazard map and the actual extent of the storm surge. Usually officials tasked with the information of communities in need of evacuation would look at the governmental hazard map and inform the respective neighborhoods accordingly. Based on interviews conducted by GIZ most staff responsible for disaster risk management in local government units along the eastern coast of Leyte said that they consulted the map before the storm hit the island.

Officials identified a number of evacuation centers outside the area marked as storm surge prone, but the water of the surge flooded some of those centers and many died there.

The underestimation of the storm surge inundation area on the official hazard map highlights the need to design EWS in a comprehensive, end-to-end manner, with due attention to risk assessments.

3.2 Monitoring and Warning Service

The detection of an upcoming threat is hazard specific and has been discussed in the Chapter 2, Hazard Specific Aspects of Early Warning. For many hazards the sources of information and data gathering involve a multitude of sensors and sources with the recent addition of crowd sourced information. Many of those are not yet accessible to staff tasked with warning decisions. Usländer suggested in his presentation in Davos (IDRC, 2014: 729) that the seamless interconnection of devices to the Internet, being sensors of all types ranging from in-situ measurement devices, sensors on smart phones up to hyper-spectral cameras mounted on satellites, offers an enormous potential for the improvement in recognizing and assessing risks, for the targeted launch of preventive measures e.g. improved quality, preciseness and personalization of early warnings. The same is true for the decision support in disaster management, especially concerning early warning. However, to exploit this potential, there is an urgent need to improve interoperability. Therefore, Usländer argues for an open approach of sensor-based global information management based upon international standards.

In some cases decision makers do not have too little but potentially too much information and the current diverse structure of data in the internet does not provide them with the appropriate tools. The measures proposed by Usländer would solve this problem at least partly.

Technology and the capability of handling it is not as widespread as it should be and development assistance of highly industrialized countries (e.g. Britain, Germany, Japan, US) to developing countries is essential for the advancement of early warning systems. An example of cooperation to this effect was provided in Davos by the British Met Office. It supports many countries in their technical needs for improved weather services, including those relevant for early warning (Donovan, 2014: 218).

The Deutsche Gesellschaft fuer Internationale Zusammenarbeit (GIZ) addressed another problem of warning decisions. In case operation centers are far away from the site the staff may not be very familiar with local circumstances, the question of who in what area should be alerted first or whom to contact if a primary contact fails or communication to the far-away countryside is interrupted, is solved if the operation center is located near the jeopardized site and locals are working in the center. GIZ promoted Local Flood Early Warning Systems (LFEWS) in 16 river basins in the Philippines with locals in charge of the operations. Commitment to the affected communities and familiarity with the area and the rivers proofed to be an advantage for the smooth operation of the systems.

Decisions on early warning are normally provided in alert stages. Starting from a low level, usually asking the public to be on alert and wait for further instructions, up to the highest level, commonly ordering evacuation and to take shelter. In many countries different hazards have different warning stages (e.g. 3 for floods, 4 for storm and 5 for volcano). This might be clear and logic for some experts but most people may be confused by different alert stages. In interviews after the storm Haiyan in the Philippines GIZ found that many people did not know how many alert levels storm has or they did not know what warning stage was raised before the typhoon hit land (Neussner, 2014).

The author proposed a simplification and using a universal four stage system (Neussner, 2014). This would also consider the fact that in today’s world many people are migrants or travelling (e.g tourists) and are therefore not familiar with local peculiarities. A worldwide standard would reduce the problem of diverse schemes leading to confusion.

Figure 6: Proposed four stage alert scheme. Symbols for illustration purposes only.

The scheme displayed in Figure 6 shows four hazards with sample symbols, but it could be easily expanded to other hazards. The colors for stages 2-4 appear to have some universal acceptance while level 1 is blue in some warning schemes, but it seems that the color green is more widely used for a stage that should signal that things are fine (but attention is needed).

3.3 Communication

Once a warning decision has been taken by a designated office and a specific alert level is in force for a certain area, this has to be communicated to disaster professionals and the general public without delay. This communication faces two challenges. It is a technical challenge as the news has to be spread fast to ALL concerned and it has to be understood by everyone.

Modern communication technology is fast, but some hazards like local tsunamis, landslides or flash floods require widespread alerts within minutes after a warning decision is taken. One aspect of this issue was addressed by Asgary and co-worker with reference to a flash flood in Toronto in June 2013. Asgary used AnyLogic simulation software and showed that a better coordination between the emergency managers and media of the emergency alerts and warnings, and general public attention to the emergency warning messages and subsequent actions could have significantly minimized the overall impacts of the flash flood.

Communication of warning often is not always a matter of minutes. For many hazards hours or days are acceptable warning times (e.g. distant tsunamis, tropical cyclones, floods in bigger rivers) and technology is not the limiting factor. This means simple and cheap communication chains are achieving good results. An Example is displayed in Figure 7.

Figure 7: Communication chain from the Operation Center to households in Local Flood Early Warning Systems (LFEWS) promoted by GIZ in the Philippines.

The alert levels are communicated in a largely two way system from the Operation Centre to municipalities, then villages and finally households. In addition there is public radio and TV, but there is no provision for feedback with these channels and therefore the backbone of the communication is the upper chain in Figure 7. Communication is mostly via handheld radios, mobile and landline phones, while on community level, megaphones, bells or messengers going from house to house are used. It is noteworthy that these systems do not include internet-based devices yet (smart phones, tablets, computers) as they are not commonly used yet in rural areas of the Philippines. These simple communication chains were successful in transmitting warnings to all those who needed to be alerted.

The second challenge of communicating warning messages to the public is the issue of understanding these messages. Different types of hazards and different alert levels have to be clearly distinguished, but lengthy explanations are too time consuming and might not be clear enough. A compromise between being too simple and too complicated needs to be found and the terms used must be unambiguous.

An example of a problem of understanding the warning of a governmental institution is the English term “storm surge” used by Philippine authorities before the tropical cyclone Haiyan made landfall on 8 November 2013 in the country. Many coastal residents complained later that they did not understand this term (Tohoku, 2014; Neussner, 2014) and that it describes a wave with the characteristics of a tsunami. Interestingly no appropriate translation of the English term “storm surge” exists in local languages (Waray or Cebuano) and it was suggested that the authorities should have warned of a tsunami. The proposition met resistance from scientists who do not want the two phenomena to be confused. The discussion is still going on and efforts are made to find terms to explain natural phenomena in simple words and also clear symbols without losing important information. The author believes that especially the symbols should be following worldwide standards in order to accommodate the growing number of temporary visitors (e.g. migrants, tourists), while the wording may be local plus internationally used languages (e.g. English). This topic might be suitable for deliberations in the 3rd United Nations World Conference on DRR scheduled to take place in March 2015 in Sendai, Japan.

3.4 Response Capacity

Early warning only makes sense if there are means of reacting properly to a warning. This concerns mainly the provision of evacuation centres and respective evacuation routes, but also search and rescue services and stocks of relief goods for emergencies.

There are few buildings constructed with the sole purpose of serving as evacuation centres. Evacuations are rare events and it would be a waste to have buildings unused most of the time. Therefore many public buildings are temporarily used to host evacuees. Schools, gymnasiums, community centers or in some cases religious buildings (e.g. churches) are utilized.

Having appropriate places for shelter is not really a challenge for technology, but more one of proper organization and coordination. Of course, an evacuation center, should provide safety and needs to be outside of the hazard area or sturdy enough to withstand the forces of nature. Furthermore, buildings designated as evacuation centers should provide some minimum facilities (toilets, water, space for cooking)(The Sphere Project, 2004).

3.5 Governance and Institutional Arrangements

Occasionally Early Warning Systems (EWS) are primarily seen as technical arrangements of detecting a threat and using a communication chain for the dissemination of a warning. However, many cases of non-ideal performance of EWS are not technical issues but aspects involving the institutional arrangements of EWS. Very often many different offices and institutions are involved in EWS. The threat detection is usually a scientific task, but decision making may be done by a politician, while the dissemination of the alert may be carried out by another office (e.g. Ministry of the Interior or a designated disaster office), and public media (private or governmental). Clear definition of roles is essential and may facilitate the fast and comprehensive information of all concerned.

Some contributions during the International Disaster Risk Conference in Davos in August 2014 assessed previous disasters and noted that institutional issues reduced the effectiveness of the EWS (Asgary, 2014; Neussner, 2014). In the experience of Deutsche Gesellschaft fuer Internationale Zusammenarbeit (GIZ) a locally run Flood Early Warning System (FEWS) reduces the number of players and increases the effectiveness of the FEWS (GIZ, 2012). In this type of setup the detection, decision making, and initial warning are in one hand no disputes about responsibilities arise.

4 Conclusion

Early Warning Systems (EWS) have not reached their full potential yet. There are still some scientific and technical issues to be solved for rapid onset extreme natural events (earthquakes, local tsunamis, flash floods, landslides), and mature existing technology for storms, river floods, volcanoes, distant tsunamis needs to be applied more widespread, including remote communities and developing countries. Furthermore, the diversity of early warning alert stages, warning messages and signs/icons, is confusing and some standardization would increase the effectiveness of EWS.

For earthquakes the main challenge remains to be the short-term prediction of an event. The spatial precision of the Lithosphere-Atmosphere-Ionosphere Coupling (LAIC) model may be already good enough, but the temporal accuracy is not yet in the range of practical applications. Landslide motion sensors are tested in many places, but their reliability and accuracy has also not reached the level for mass application yet.

All rapid onset events need fast and widespread communication structures. Though it is possible to inform some recipients on short notice, the problem on how to secure that everyone in a danger zone gets the message, is still a challenge.

Many areas in the world require EWS but none is operational there yet. Political decision makers should consider establishing smaller, local EWS in order to fill remain gaps.

Though EWS have to be adjusted to local circumstances the growing number of temporary visitors (e.g. migrants, tourists) in the world, require some type of international standard easily understood by everyone. While the wording may be local plus internationally used languages (e.g. English), signs/icons should follow international conventions. This topic might be suitable for deliberations in the 3rd United Nations World Conference on DRR scheduled to take place in March 2015 in Sendai, Japan.

References

Alozie, J. et al. (2014): The Seasonal Rainfall Prediction: Early Warning Tool for Disaster Risk Reduction in Nigeria, IDRC, DAVOS 2014, Extended Abstracts, 44

Asgary, A. et al. (2014): Simulating Disaster Warning and Alerting Effectiveness, an Agent Based Approach, IDRC, DAVOS 2014, Programme & Short Abstracts, 109

Donovan, T. (2014): Working together to deliver sustainable development through weather and climate services, IDRC, DAVOS 2014, Extended Abstracts, 218

Fernandez de Arroyabe, P. (2014): Atmospheric determinants of health risks and customized early warning systems, IDRC, DAVOS 2014, Extended Abstracts, 247

Hernando, H. et al. (2013): Community involvement and local flood early warning with low-tech approaches for small rivers in the Philippines, Presentation at the World Bank workshop Flood Risk Management and Urban Resilience in Seoul, Korea (http://www.gfdrr.org) (15 Oct 2014)

GIZ (2012): Local Flood Early Warning Systems, Manila, 54 pages (http://www.preventionweb.net/files/29230_29230gizphilippineslocalfloodearlyw.pdf) (16 Oct 2014)

Neussner, O. (2012): Community participation in local flood early warning systems based on low-technology approaches in the Philippines, in: Government of Germany: Experiences in Disaster Risk Management within the German Development Cooperation (Special Joint G20 Publication for DRM), 153-155 (http://www.gfdrr.org/gfdrr/sites/gfdrr.org/files/Chapter_9-Germany-Experiences_in_Disaster_Risk_Management_within_the_German_Development_Cooperation_0.pdf) (17 Oct 2014)

Neussner, O. (2014): Assessment of Early Warning Efforts in Leyte for Typhoon Haiyan/Yolanda, March 2014, 64 pages (http://www.preventionweb.net/files/36860_36860gizassessmentofearlywarningyol.pdf) (18 Oct 2014)

Ouzounov, D. et al. (2014): Testing new methodologies for alerting large earthquakes: An integrated geo space approach and ground observations, IDRC, DAVOS 2014, Programme & Short Abstracts, 168

Perminov, A.N. et al. (2014): Methodology of Short-term Powerful Earthquakes Forecasting and Warnings Through the Using of Space-ground Monitoring Data: Russian Approach, IDRC, DAVOS 2014, Programme & Short Abstracts, 170

Pulinets, S. et al.: Novel conception of short-term earthquake forecast based on Lithosphere-Atmosphere-Ionosphere Coupling (LIAC) model, IDRC, Chengdu, http://idrc.info/fileadmin/user_upload/idrc/former_conferences/idrc2009/presentations/Pulinets_Sergey_Ouzounov_Dimitar_IDRC_Chengdu_2009.pdf, 3 (17 Oct 2014)

Pulinets, S., et al. (2014): Physical background for multiparameter approach in earthquake precursors monitoring, IDRC, DAVOS 2014, Programme & Short Abstracts, 175

The Sphere Project (2004): Humanitarian Charter and Minimum Standards in Disaster Response, Geneva, 344 pages

Tohoku University, International Research Institute of Disaster Science (2014): Initial Report of the IRIDeS Fact-finding missions to Philippines, (http://irides.tohoku.ac.jp/media/files/IRIDeS_Report_Haiyan_20140527.pdf) (16 Oct 2014)

UNISDR (2006): Developing Early Warning Systems: A Checklist, Third International Conference on Early Warning, March 2006, 13 pages

Usländer, T. (2014): The trend towards the Internet of Things: what does it help in Disaster and Risk Management?, IDRC, DAVOS 2014, Extended Abstracts, 729

Citation

Neussner, O. (2015): Early Warning – Some Recent Developments. In: Planet@Risk, 3(1): 24-32, Davos: Global Risk Forum GRF Davos.


1
Funded by the Federal Ministry for Economic Development (BMZ) in the framework of the project Global Initiative for Disaster Risk Management
2
Currently with the project Global Initiative for Disaster Risk Management