Development of Real-Time Distribution System of Asian Region on Disaster Reduction Information

Shigeya Yoshikawa*, Kiichi Hirono*, Susumu Yokomaku*
Hirotaka Suzuki**, Yujiro Ogawa**, Bambang Rudyanto**

*Asian Air Survey Co., Ltd. System Engineering Division, Aoki Daisan Building 2F, 7-3 Tamura-cho, Atsugi-shi, Kanagawa 243-0016, Japan
**Urban Disaster Reduction Research Center Asian Disaster Reduction Center, IHD Building 3F, 1-5-1 Wakihama Kaigan Dori, Chuo-ku, Kobe-shi, Hyogo 651-0073, Japan



The Asian Region is hit by extensive disasters every year, which leave massive social and economic damages. The disaster management administrative officers and researchers of each country in Asia are thus faced with the task of establishing mechanisms which allow disaster information for a broad area extending over many countries to be obtained for swift and easy tracking and analysis of the scale and size of the damage incurred.
We therefore developed a system that can distribute disaster reduction information to Asian regions, using Internet technology for effective distribution and GIS technology that expresses the geographical spread of disaster for easier analysis of the state of disaster.
In order to minimize damage when disasters occur, there is a need to collect and distribute disaster information promptly. We therefore sorted out pre-existing and state-of-the-art technology for collecting disaster information promptly and verified the accuracy of this information. We then conducted experiments to measure the time taken for the period from actual acquisition of the data to distribution. Based on these results, we studied effective data sources and data collection technology for distributing disaster information promptly to Asian countries.
This paper reports the results of the study on the real-time collection and distribution of disaster information as well as the features of the system developed.

Keywords: Real-Time distribution system; Disaster reduction information

1. Introduction

The Asian region is often hit by climatic disasters, volcanic disasters, earthquake disasters, etc., and thus suffers considerable social and economical damage. Since damage sometimes extend from the country hit to other countries, the exchange of disaster information through multilateral cooperation is essential for the collection, processing, and transmission of the information and for implementing various measures. The establishment of mechanisms that allows such exchanges is thus a challenge. Though methods for developing individual systems by each country are conceivable, to realize cooperation in implementing relief for damages and rescue activities the multinational sharing of disaster information is imperative, as are the standardization of system functions and data formats. Geographical information on topography, roads, and railways required for tracking the geographical position of information also needs to be consolidated and stored for the system of each country. This means that massive time and costs would be required for establishing the required systems and environment.
In order to minimize the spread of damages when disasters occur, disaster management administrators and researchers of the respective countries in Asia must be able to acquire disaster information on the scope of damage and disaster immediately. However, there now exists no method for obtaining such information broadly across the Asian region.
In order to resolve these problems, we have developed a real-time distribution system of disaster reduction information for the Asian region using a server-client type of Internet GIS technology (VENTEN (Vehicle through Electronic Network of disasTer gEographical information). The VENTEN system manages disaster information at only one location (server) to eliminate the need for standardizing system functions and data formats. In addition, because the geographical information of each country is managed at the server, there is no need to consolidate or store the information by country.
By using this system, disaster management administrators and researchers of each country need only establish the environment for a Web browser to run. The use of the Internet for distributing data enables these administrators and researchers to promptly obtain information on damage that has occurred in the Asian region.
To collect and distribute disaster information in real-time using this system, raw data "extracted" from the disaster area is immediately required after the event. However, the sources of such data of disaster information that are suitable for the Asian region are not clear.
In addition, to realize the prompt distribution of disaster information, it is also necessary to determine the time taken for extracting disaster information from data sources. For this reason, we investigated data sources that enable prompt collection of information on the state of the earth's surface over a broad area, and the level of disaster information that can be extracted from these data sources. Furthermore, we conducted experiments using actual data on this system to check the time taken from obtaining data to extracting disaster information and distributing it.
Real-time distribution here aims at distributing information within several days from the time of a disaster.


2. Research on Real-Time Collection and Distribution of Disaster Information

2.1.Use of Disaster Information Collected and Distributed at Real-Time on the System

Figure 1 shows an example of the use of disaster information collected and distributed in real-time on this system. This system provides disaster information generated in the Asian region by the following process.

1)Images, aerial photographs, ground vehicles and people, disaster information on the scope of floods and high tides, destruction of homes, and the state of road blocks is extracted.
2)The extracted disaster information, as well as the satellite images, aerial photographs and geographical images used for extraction, is registered in the database server.
3)Disaster reduction management officers and researchers (the clients) access this system through the Internet, search for the disaster information, and using this information and geographical information, calculate the number of victims and shortest route for evacuation and rescue.


2.2. Results of Study on Real-Time Collection and Distribution

(1) Pre-Existing Technologies for Collecting Data Sources and Extractable Disaster Information

We investigated the platform for collecting data sources that can be used for extracting disaster information applicable to the Asian region and the range of observation from existing literature. The results are shown in Table 1. It shows that the platform suitable for the Asian region that can also collect data over a broad area is the satellite.

As shown in Table 2, the resolution is low when using satellite images collected by this platform for extracting detailed disaster information. For this reason, satellite images are often used for collecting disaster information on extensive damages such as river floods, lava flow, pyroclastic flow, high tides, tsunami disasters, etc.

Table 3 shows the pre-existing technology for collecting data sources with higher resolution than satellite images vs. the latest technology. Tables 1 and 3 indicate that aerial photography technologies and on-vehicle video images (mobile mapping) technologies can extract detailed disaster information in real time, but with the observation area being narrower than that of satellite images. Disaster information that can be extracted from these collection technologies includes buildings damaged
or tilted by earthquake disasters, rubble, the state of fires, and more.

(2) Time taken for collecting, extracting, and distributing information

@ Satellite image information
Based on data on the changes in the number of natural disasters by area and percentage of disasters worldwide from 1963 to 1992, as disclosed in "Disasters Around the World-Global and Regional View," a survey was conducted on the Asian region. The results indicated that river floods are the largest natural disaster incurring economical and individual damage to the Asian region.
Through experiments, we also calculated the time for obtaining satellite images of actual river flooding, extracting the data, and registering the information in this system. The following describes details of the experiment performed using the satellite image and the results.
- Data used
Landsat satellite images (TM sensor)
- Studied disaster
Flood in Wuhan, China (Exposed August 14, 1998)
- Extraction method
Extracted flood data by obtaining changes in territorial area images before and after damage by land-cover categorization.
- Results of experiment
*Time taken to obtain images: About 1 month from time of disaster
*Pre-processing and extraction: About 1 day
*Data registration for distribution: 0.5 to 1 hour

APre-existing and recent technologies for collecting detailed information
Using both pre-existing and recent technologies for collecting detailed information, we calculated the time taken to obtain the actual images exposed, extract and distribute the disaster information through experiments. The following describes the experiments performed using various collection techniques.
#Aerial photograph information
- Studied disaster
Approx. 1 km2 area at the west side of JR Nishinomiya Station during the Great Hanshin-Awaji Earthquake (exposed on January 18, 1995) (see Figure 2).
- Extraction method
Determined the damage, tilting of homes, rubble, and fires from the color and shape of the surrounding areas.
- Results of experiment
Time taken to obtain aerial photographs:
- Several hours to about 1 day after exposure
- Pre-processing and extraction (targeted areas this time): About 10 or more hours
- Data registration for distribution: 0.5 to 1 hour
- On-vehicle video images (mobile mapping)
According to the lack of data sources on the state of disasters after the technology was put to practical use, we selected a model area, and took images with the on-vehicle video. From this data source, we extracted buildings that satisfy certain simulated conditions as disaster information, and obtained the time taken for distribution.
# Studied disaster
Results of exposing 10 km area of Maebashi City streets
# Extraction method

From the exposed video images, we generated images for every linear-scanned area by automatic mosaic (see Figure 3), and read from these images the positions of parking areas facing roads and the presence of parked cars as simulated information on the damaged state of buildings.
# Results of Experiment
- Time taken to obtain on-vehicle video image information: Several hours to about one day from exposure
- Pre-processing and extraction (targeted areas): Several hours
- Data registration for distribution: 0.5 to 1 hour


2.3 Results of System Development

(1) Information provided by system
The information provided by this system is as follows.

@Disaster information
Data sources are collected; disaster information is extracted and distributed after inputting it into the database.
# Specific disaster information provided:
For disasters whose scope is a closed area, such as areas submerged by floods and high tides, the affected range is provided as polygon information.
Point damage such as the tilting of buildings and damage due to earthquake, rubble, and fires are provided as point information.

AGeographical information
For tracking the position of the disaster data and the surrounding conditions, information such as national borders of the disaster-stricken area, water systems, roads, railways, transportation facilities, public facilities (hospitals, schools, etc.), names of cities, population dot graph/line map/polygon graph (see Table 4) are provide

BAnalysis results
The system user utilizes the buffering, route search, and counting functions to obtain the following analysis results from disaster information and geographical information (see Figure 4).
# Search for route for evacuation and rescue
Shortest route between two points specified on a road on the map or railway.
# Count of number of victims and damaged facilities included in the scope of disaster or scope of disaster effects.
Results of totaling geograp information included in the scope of disaster information, such as submerged area and random area.

(2) Features of System
The following describes the features of this system.

This system covers the geographical information of a broad area of 22 countries in Asia, and is able to provide information on disasters occurring in this area.

AReal-time information
The registered disaster information can be distributed immediately via the Internet.

The user can check the disaster information immediately if in an environment that can access the Internet.

COperational ability
By specifying the name of the desired country or the desired country from the index map, an enlarged map centering around that country will be displayed (see Figure 5). This means that the user need not magnify or move items on the screen to display the information of a certain country.
Disaster information can also be searched for by country name, type of disaster, and date the disaster occurred, thus speeding a check of registered disaster information.

DAnalysis function
This system not only displays disaster information and geographical information, but it also has a function which uses this information to interactively perform advanced analysis on the Internet, such as buffering and route searches. The following shows the analysis functions of this system.
# Function for creating the scope of disaster effects by buffering
Buffers (of the desired size) that consist of mainly the desired drawing of the disaster information or geographical information are displayed on the map. Using this function to overlap the scope of disaster effects compiled with geographical information, the overall geographical information for that area can be made to appear graphically.
# Function for overlapping disaster information with the desired geographical information
This function displays disaster information and the desired geographical information overlapped. It allows tracking of the geographical state of the surrounding area--which is important for disaster management.
# Damage totaling function
This function totals geographical information (for example the number of victims and number of damaged facilities) included in the scope of disaster information, like flooded areas, for the desired area. This function can be used to calculate damage.
# Evacuation and rescue route search function
By specifying the start and end points of roads or railroads on the map, the shortest route and distance can be calculated. It is also possible to search routes outside buffer areas compiled according to the area of the disaster information or at the desired size. This function enables evacuation and rescue routes to be reviewed.


3. Discussion on Real-Time Collection and Distribution of Disaster Information

As a result of the study on the real-time collection and distribution of disaster information (see Table 2), it was found that satellite images are not necessarily able to provide information on conditions immediately after a disaster occurs because the satellite only has the chance to shoot the same area once every several days to once every more than ten days. It is however a comparatively easy-to-use data source in Asian countries.
It was also clarified that the collection and distribution of detailed technical information by pre-existing and recent technologies has high real-time performance and the potential to serve as an effective means of collecting information immediately after a disaster has occurred.
These results indicate that it is advantageous to use satellite images for accumulating disaster records, tracking past disasters and capturing the surrounding conditions, because of its availability in the Asian region. However, aerial photographs can be applied when collecting data sources in real time.
In the future, the latest technologies such as on-vehicle video technologies, which have higher real-time performance than satellite images and are easier to operate than from airplanes, should become available.


4. Discussion on Real-time Distribution System of Asian Disaster Reduction Information

By constructing this system, we have established a mechanism for providing useful information to disaster management administrators and researchers in Asia for tracking and analyzing disaster information. The geographical information currently available is in the small scale of 1/1 million, and can be applied to large-scale disasters over several 10 km2. For disaster countermeasures requiring analysis from an even smaller scale, there is a need to compile large-scale map data in the range of 1/several thousand to 1/several tens of thousands, with the cooperation of countries in Asia.


5. Conclusion

To enable multinational sharing of disaster information, and collect and distribute the disaster information at real-time, we developed a server/client-type real-time Asian region disaster reduction information distribution system. The system centrally manages other systems, disaster information and geographical information at the server and distributes it promptly. This has, as a result, established the mechanism for disaster management administrators and researchers in Asia to obtain disaster information in real-time.
To collect and distribute disaster information in real time using this system, it was necessary to clarify the technologies for collecting data sources that can extract information from disasters immediately after occurring, as well as clarify the disaster information that can be extracted and the time taken from collection to extraction.
We thus investigated the data sources that can be used and disaster information that can be extracted, and calculated the time requiring for obtaining disaster information through experiments using actual data. As a result, we were able to clarify the disaster information that can be obtained by disaster management administrators and researchers in Asia using this system, depending on the type of disaster, as well as the time taken to obtain the information.
This system has been running at the ADRC homepage since April 2000. It is available at the following URL.

Figure 1 Using Disaster Information


Figure 2 Great Hanshin-Awaji Earthquake--Overlapping of Collapsed buildings and aerial photography (January 17, 1995)


Figure 3 Image Exposed by On-Vehicle Video Camera (mosaic-processed)


Figure 4 Search for Evacuation and Rescue Routes


Figure 5 Selection of Country from Index Map


Table 1 Data Source Collection Platform


Table 2 Satellites that can be used for extracting disaster information


Table 3 Technologies that can reveal the state of disaster


Table 4 Geographical Information




We would like to express our appreciation to Professor Takeda, Environmental Planning Technology Department, Faculty of Science and Technology, Tohoku Bunka Gakuen University, for his support in the experiments using Landsat satellite images in our study on real-time collection and distribution of disaster information.