Decision Support Tool for Disaster Management in the Case of Strong Earthquakes

Chen Yong*, Chen Q.F.*, Frolova N.**, Larionov V.***,
Nikolaev A.**, Pejcoch J.****, Suchshev S.*****, A.N. Ugarov*****

* China Seismological Burea, 100036, Beijing, China, e-mail:
** Seismological Center of IGE, Russian Academy of Sciences, Nikoloyamskaya str. 51, Moscow 109004, Russia, e-mail:
*** All Russian Scientific Research Institute on Civil Defense and Emergency Problems, Davydkovskaya str. 7, Moscow 121352, Russia, e-mail:
**** T-SOFT s.r.o., Novodvorska 1010/14, 142 01 PRAHA, Czech Republic, e-mail:
***** Extreme Situations Research Center, Yunykh Lenintsev str, Moscow, Russia, e-mail:



The paper provides a description of the Web Based Decision Support Tool "WaveLet" for expected damage and loss assessment, as well as for the identification of effective response measures to strong earthquakes. This System takes advantages of GIS "Extremum", "GloSeisRisk" and "EIG/GEM Infobook". The "WaveLet" System development is supported by the World Agency on Planetary Monitoring and Earthquake Risk Reduction.

Keywords: Web based decision support tool; possible damage extent; seismic and complex risk; expected social and economic losses due to strong earthquakes; effective response measures; geographical information system

1. Introduction

Strong earthquakes and secondary processes result in catastrophic human suffering, loss of property and other negative consequences. Over 1.6 million people have died in earthquakes during the 20th century. Analysis of natural emergencies including earthquakes occurring all over the world during the last 5 years show that these types of emergencies are growing at an average rate of about 5% per year.
The recent disasters in El Salvador and India have reaffirmed the urgency and importance of expected loss assessment due to strong earthquakes, and have shown once again that a decision support tool is necessary, which allows the possible consequences to be estimated before and just after the event. The goal may be achieved by two ways. The first one is to provide all decision makers with their own tool for consequences simulation and risk assessment. The second variant is to develop a web based tool, which may be accessible to all interested decision makers on the regional or global level.
The "Extremum" System is an example of the second type of tool. Its operation over 10 years as been as a local system (1990 - 1997) and as a web-based system (1998 -2000). It was developed by joint efforts of Extreme Situations Research Center Ltd. and organizations belonging to the Russian Ministry of Emergency Situations, Russian Academy of Sciences, which are combined on an informational level into the Russian Agency of Monitoring and Forecast of Emergencies. In 1999 - 2000, the System was advanced within the framework of EUR-OPA EDRIM ("Electronic Discussions in Risk Managements") Program. The System is based on the application of new information technologies.
The System may be utilized for loss and risk assessment by registered end-users. The end-user may identify the scenario event that meet specified conditions. The dialogue between the end-user and the System is realized by means of the Internet.
The System is now used in order to provide operative information on damage and casualties assessment of strong earthquakes all over the world:

In order to increase the accuracy and reliability of loss and risk assessment with the GIS "Extremum" application, it is proposed that additional expert block be developed. Within the block the estimations of possible consequences obtained at least by two independent approaches are analyzed and an expert decision is taken. There are plans to make use of distributed systems' advantages in order to access reliable information about seismic and collateral hazards, population distribution, human-made environment and other information. The development of the advanced System was started under the umbrella of the World Agency of Planetary Monitoring and Earthquake Risk Reduction.


2. Web Based Decision Support Tool "Extremum"

According to the request of EUR-OPA, one of the end-users, the System is now used in automatic mode. It allows:

The computation time performance is 10 min. after receiving alert data about a strong event. The response scenario may be prepared within 1.5-2 hours after a strong earthquake in any country; the average error in social losses assessment is 60 %. The example of the System application for losses assessment due to the August 17, 1999 in Turkey showed high efficiency (Fig. 1).

Table 1. Possible losses assessment due to the August 17, 1999 earthquake in Turkey

No Name Computed values
NEIC, Obninsk, EMSC
Real data Error, in percent
1 Total losses 36,630 -48,860 persons  61,080 persons* 40 - 30
6,690 - 17,340 persons 90 - 72
57,880 - 71,190 persons 5.3 - 16
2 Fatalities  5,410 - 22,340 persons 17,127 persons* 68 - 30
510 -3,930 persons 97 - 77
9,760 - 34,910 persons 43 - 103
3 Injured 26,520 - 31,220 persons 43,953 persons* 39 - 29
6,180 - 13,410 persons 86 - 70
36,280 - 48,120 persons 17 - 9
4 Homeless 138,190 -155,110 persons 150,000 persons 4 - 8
5 Total number of damaged buildings 55 %  244,500 buildings* -
6 Number of buildings with d = 5 15 % 77,300 buildings* -
7 The ratio of buildings with d = 5 to the total number of damaged buildings (in %) 27 % 32 % 5

*)Source: Relief Web OCHA - Turkey: Earthquake Situation Report No. 24

Table 1 shows that a more reliable estimation of possible consequences was made with NEIC alert data usage. The maximum errors in fatalities computations may exceed 100%. The total losses and number of injured, as well as the number of buildings which suffered different damage, are the most reliably forecast.
At least one Survey gave the input data, which allows to estimate possible losses with error in range of 20-100%. The task of the expert is to choose the most reliable survey for the area under consideration.
The accuracy of possible damage and losses assessment depends on the reliability of information about existing building stock in the stricken area. In order to identify the data on existing buildings stock, the high-resolution space images application is proposed. The possibility of loading the high resolution space images of the striсken areas into the system for operative (real time) losses assessment was analysed. Development proceeded for a procedure of space image decoding, which allows in a short-time interval to update the information about existing building stock, as well as distribution of buildings with different number of stories. At present the procedure allows to estimate the vulnerability of different classes of buildings with high probability. If a building's size is 12 meters (6 pixels), its vulnerability class may be determined with probability = 0.8 (Fig. 2).
The System's conceptual framework, its databases on cartographic information, population distribution, data on existing buildings' stock inventory, on natural and technological hazards, as well as mathematical models are described in (Chen Yong et al., 2001; Larionov et al., 1992; Larionov et al., 1993; Shojgu et al., 1992; Larionov, 1999; Larionov et al., 2000; Shakhramanian et al., 2000a; Shakhramanian et al., 2000b).
Its database includes the information massifs (IM), which are combined into four groups. The first IM group allows one to describe in details the space under study. This group contains the digital topographic data. As the accuracy, completeness and reliability of this data is concerned, they correspond to the standards for the scales of maps. M. 1 : 5 000 000; 1 : 1 000 000; 1 : 100 000; 1 : 10 000, 1: 2 000. Small-scaled maps give general information about the regional topography. Large-scaled maps allow the structure of cities and towns to be described.
The second IM group is assigned to describe seismic hazards. It contains catalogs and the information from seismic zonation maps of different scale (review, detailed and micro-zonation). The data forms the set of thematic maps, tables, networks and lists.
The third IM group provides the description of the different elements at risk: population, buildings and structures, lifeline systems, hazardous facilities, et al. The information about buildings may be detailed (type of structure, materials, date of construction, height and so on) and generalized--for instance, the distribution of buildings characterized by different vulnerability classes within the city districts. The information about the population distribution in the buildings and city districts within 24 hours is also included.
The fourth IM group combines the parameters of mathematical models for population distribution, building damage distribution, casualties and fatalities, for rescue team operations and others.
All four IM groups of information massifs are interrelated by single coordinate space (coordinate system B, L, H) and by a unified code system. The Internet is also used as a technical tool for data collecting and the presentation of the obtained results.
The mathematical models allow one to:

On the basis of the results of computations for scenario and real events, the decision is taken about immediate response and/or preventive measures. Figure 3 shows expected building damage distribution due to the scenario earthquake for Petropavlovsk-Kamchatsky City. Fig. 4 shows possible damage to roads due to scenario event from zone PET. Fig. 5. shows individual complex risk zonation (earthquakes, fires hazardous and chemical hazardous facilities).
 Contribution of hazardous facilities to complex risk is about 3% of seismic one. The results were used by local authorities for verification of preventive measures planned in the case of a strong event.
The output results of computations are presented as thematic maps, tables, graphs, as well as pages by means of the Internet.
 The present Tool is running in a Windows environment. It is made up of the unified complex with commercial databases (dBase, Access, Oracle, Informix) and management systems. The exchange of data with desktop GIS, ARCINFO, MAPINFO is provided. The equipment, which is used for GIS realization, corresponds to the modern standard IBM PC for graphics stations.

3. Web Based Decision Support Tool "WaveLet"

At resent the System is under reconstruction in order to become the Distributed Decision Support System (DDSS). The advanced system is called "WaveLet". In order to increase the reliability of possible consequence estimation due to strong earthquakes, at least two approaches are planned to be used to simulate possible damage and losses. It is proposed to use the detailed simulation approach foreseen in the System "Extremum" (idea authored by Dr. V. Larionov) and approach based on macroeconomic indicators (idea authored by Prof. Chen Yong). The application of the approaches will allow experts to avoid the terrible mistakes in loss estimations. Figure 6 shows the general framework of "WaveLet" System. The new blocks are shown by blue color. The block no. 2 was added by the approach based on macroeconomic indicators, which does not require a detailed inventory database of the structures and facilities in the region. It uses an alternative means of estimating earthquake losses based on several macroeconomic indices such as the Gross Domestic Product (GDP) and population (Chen et al, 1998).
Based on the published earthquake loss data during 1980-1998, the relations between GDP and earthquake loss have been formulated empirically for several intensity ranges. The world's land surface was divided into unit cells with 0.5° ´ 0.5° in size, the GDP of each cell was apportioned based on its population and the GDP and population of the region to which it belongs. The predicted seismic loss of the cell was then estimated from the seismic hazard probability function, its GDP, and the empirical relation between GDP and seismic loss. Accordingly a global seismic loss map is compiled for intensity VI and above. Employing readily available social economic data as the basis for the vulnerability analysis, the method generates seismic loss estimates for regions without a detailed inventory of exposed structures or the required collateral geological information. Seismic loss estimates can also be upgraded easily with social economic data collection for fast-developing areas of the world.
The approach uses the "GloSeisRisk" software package for Global Seismic Hazard and Risk Analysis. The "GloSeisRisk" is designed to analysis and demonstration of seismic hazard and risk. The "GloSeisRisk" is a program running in a Windows environment, therefore it is easy to combine with the System "Extremum". "GloSeisRisk" includes the following world-wide databases:

・ Historic earthquake catalog (USGS, since 1800).
・ Instrument earthquake catalog (Mmin=4.5, 1964-1998).
・ Population data (resolution 0.5° ´ 0.5°, 1998).
・ Average attenuation of intensity or Peak Ground Acceleration (PGA).
・ National Gross Domestic Product (GDP, 1998).
・ World wide earthquake loss data (since 1990).

The advantage of "GloSeisRisk" is that it can provide a reasonable assessment of global as well as regional seismic hazard and risk by using limited data and a simplified method. A key contribution of "GloSeisRisk" is the series of global or regional seismic hazard and risk (loss estimate) maps that are produced and updated periodically with new and refined information. "GloSeisRisk" now provides users with the quantitative products and information of seismic hazard and risk, including:

・ The probability that a certain value of a macroseismic intensity or of a ground motion parameter (i.e. particle acceleration, velocity or displacement) will not be exceeded at any site (oceanic or continental) in the world in various periods of time.
・ The expected loss caused by future earthquakes at any site in the world in various periods of time.
・ Assess the seismic hazard and loss impact from earthquake scenario in the world (or in a specific region).
・ Incorporate the results with GIS for specific map generation.

The table 2 demonstrates the reliability of results obtained with the "GloSeisRisk" application.
The "GloSeisRisk" offers the maximum reliable estimation of possible social and economic losses, which may be used as upper limit (threshold) in comparison with similar estimations provided by the "Extremum" System.


Table 2 Comparison between actual loss and expected loss of recent

In order to evaluate the best choice for emergency response and preventive measures, GIS "WaveLet" will take advantages of "EIS/GEM Infobook" (idea authored by Dr. Jaroslav Pejcoch). "EIS/GEM Infobook" is used for integrated decision support in the crisis situations caused by earthquakes with taking into account secondary hazards. The procedure is realized in Block no. 7 (Fig. 7). When the system is applied to scenario identification, the following issues are covered (Pejcoch, 2000):

・ Logging and messages
・ Mapping
・ Checklists
・ Earthquake damage areas
・ Secondary Incident Response Flowchart for Incident Commander
・ Hazmat information - protective clothing, first aid, fire fighting etc
・ Building and structure damage
・ Locations of schools and hospitals
・ Briefing and deployment of resources & personnel to stricken area

The information gathered within the process of emergency response is analyzed and stored in the System database. The efficiency of all actions is estimated. As a result, the examples of positive and negative response measures are available. It is expected that these estimates can be taken advantage of during operative response to future events.
The "EIS/GEM Infobook" is a set of programs running in Windows environment, therefore it is also easy to combine with the System "Extremum". The minimal hardware configuration needed is as follows: Windows/9x or NT; Processor 200 MHz; 32MB Memory; 1 GB Hard Disk; LAN;WAN. The "EIS/GEM Infobook" contains the following programs: "Human Resources"; "Financial"; "Logistics"; "Humanitarian Aid"; "Field Command"; "Command and Control". All the programs are inter-operated through the "Message Broker" program. Fig. 6 shows the flowchart with steps for the disaster/incident commander in the case of a chemical facility incident due to strong earthquake.
At present the "EIS/GEM Infobook" is used for crisis management by many state and local governments, federal and military authorities and more than 500 companies.
It is planned that the advanced system "WaveLet" will provide the results of seismic and complex risk assessment, as well as results of computations for scenario and real events according to the request of the registered end-users in the form of tables and maps.


4. Conclusions

As a result of joint efforts within the framework of the World Agency of Planetary Monitoring and Earthquake Risk Reduction, development has proceeded with the advanced Web Based Decision Support Tool "WaveLet" for risk and consequences assessment due to strong earthquakes, as well as for identifying effective response and preventive measures. The three systems "Extremum", "GloSeisRisk" and "EIS/GEM Infobook" are for use at the decision level. In the future, the combination of these systems' databases and related models is planned.


Fig 1. Results of possible losses assessment due to 17 August, 1999 earthquake in Turkey with "Extremum" tool application


Fig. 2. Dependence of the probability F(x) of building vulnerability class identification on buildings' size


Fig. 3. Damage due to scenario event from zone AVS for Petropavlovsk-Kamchatsky City (black: total collapse; brown: partial collapse, red: heavy damage; yellow: moderate damage, green: slight damage; blue: no damage).


Analysis of possible consequences due to scenario events from AVS and PET zones led to these conclusions:oThe total length of damaged roads may reach 40 km in the case of PET event; it is three times more than in the case of an event in an AVS zone.o The recommendations were given that in the case of an event in a PET zone, the traffic will be totally destroyed and the special road repairing teams should be involved in urgent rescue operations. oThe most vulnerable road were identified near Kultuchoe Lake.oThe recommendations were given that sea way along the Avacha Bay should be used in the case of an emergency.

Fig. 4. Possible damage to roads due to scenario event from zone


Fig. 5. Individual complex risk zonation (earthquakes, fires hazardous and chemical hazardous facilities).


Fig. 6. General GIS framework


Fig. 7. Example of information flowchart in the case of chemical facility incident due to strong earthquake:
1-- Enter the incident in the Ops Log;
2-- Specialized checklists;
3-- Alert Hazmat Team personnel;
4-- Get chemical response info for fire hazards, health hazards, Personal Protective equipment, first aid measures, and much more;
5-- Run plume for chemical (NOAA's Aloha is included with EIS/GEM);
6-- Schools to evacuate;
7-- Hospitals.