WEB-BASED TRAINING SYSTEM FOR
FOREST FIRE OFFICE STAFF
Agustin Yagüe, Pedro P. Alarcón, Juan Garbajosa,
Universidad Politécnica de Madrid, OEI. E.U. Informatica. Carr. de Valencia, Km 7, E-28031 Madrid, Spain
Fraunofer-Institute for Production Systems and
Design Technology (IPK) Pascalstrasse 8-9 D - 10587 Berlin GERMANY
Keywords: System Modelling, Validation, Model-based, m
odel-driven, GIS, Simulation, System Integration, XML,
Technical Networks, XNETMOD
Abstract: The objective of this paper is to present an approach for a web-based training system for Forest fire offices.
The d
evelopment of a modelling and simulation technology for systems with a network-like architecture is
growing day by day. Forest fire offices represent an appropriate application such approach. It is based on an
XML languages family defined in a research project and is applied to a number of systems that have been
modelled and simulated. This paper introduces two different related issues: the system architecture; and the
XML-based language and its use for simulation.
1 INTRODUCTION
Every summer forest fires ends up to being the
centre of attention in press and media. In the period
1997-2001 more than 565.000 (Guardia Civil de
España, 2003), (Ministerio de Medio Ambiente de
España, 2003) (Instituto Nacional de Estadística de
España, 2003) have taken place only in Spain. In
most of the cases, these disasters are quite dangerous
to populations. It is well known that available
resources to extinguish fires are limited and demand
large budget provisions: Therefore it is important to
optimize their distribution and use.
There are many allies to fight together: fire:
pre
vention, modern resources, new materials and
simulators. All of them are useful and helpful tools
to work as defence against forest fires. Usually,
these technologies are associated to the strong
development that has taken place in the last years in
the field of the information and telecommunication
technologies.
The logical introduction of t
hese technologies in
the fight against forest fires needs of modern
operation stations where to process and store data
(digital cartography and exchange of digital
information need huge media storage) and run
applications. The background provided by these
technologies constitute an invaluable help for an
improved decision taking. Some related technology-
specific issues are:
Technical Applications of GPS
Localization and Tracking of ground and air
resources
Tracking down mobile platforms
Image transmission in real time
Infrared fire detection
Fire simulation, as already different applications
implement such as CARDIN, FARSITE,
FIREFOC, and FEOT (Lymberopoulos,
184
Yagüe A., P. Alarcón P., Garbajosa J., Lisounkin A. and Schreck G. (2004).
WEB-BASED TRAINING SYSTEM FOR FOREST FIRE OFFICE STAFF.
In Proceedings of the Sixth International Conference on Enterprise Information Systems, pages 184-191
DOI: 10.5220/0002650601840191
Copyright
c
SciTePress
Papadopoulos, Stefanakis, Pantalos, Lockwood,
1996), (McGrattan, Baum, Rehm, Forney,
Prasad, 2002), and (Molina, Castellnou, Plant,
2001).
Fire prediction Behaviour as in BEHAVE
(Andrews, 1998).
Most of the efforts mentioned above address fire
simulation. However, it is important not to forget the
training of engineers in charge of fire prevention
(Bardaj, Molina, Castellnou, 1998), (Stolk, 2003)
and (McGrattan, Baum, Rehm, Forney, Prasad,
2002). This is the main application topic of the
research work described, that fills this important
gap. The objectives of the investigation is to
simulate fire extinguishing resources allocation
under different conditions, to consider decision
making procedure of a fire office engineer, to design
the system front-end, and, finally, to process the
simulation and obtain evaluation reports. The listed
objectives consider the modelling to be the key for
human knowledge accumulating as well as the basis
for the knowledge use. The youngest time, the
modelling technology profits a lot from advances in
data acquisition and processing methods and from
communication infrastructure as well.
The development of data modelling and
behaviour simulation tools for a fire office was
performed within the European research and
development project XNETMOD (XNETMOD
Consortium, 2003). Moreover, aspects of distributed
architectures as well as web-based services were
considered as well. Within the scope of this project,
modelling technology focused technical systems
with network structure in general. The application
fields cover rather different application domains:
1. Oil/gas/water distribution networks;
2. Data capturing system, focussing on workflow
simulation;
3. Forest fire office management.
Fire prevention and extinguishing tasks provide
an appealing domain to test the technology and, on
the other side, the application elaboration presents
an intrinsic interest. Road network supervision for
prevention and management of forest fire danger is
only one relevant task. The other mentioned task is
management of the fire extinguishing resources
allocation. Simulation of system evolution, including
the fire situation, resources dynamic, social aspects,
is the further exciting task. The performed work was
structured into four phases:
Client 1
Client 2
Client N
Internet
Training System
Clients
Figure 1: System Architecture Overview
1. Problem Analysis intending to increase
experience with respect to the subject and to
elaborate use cases.
2. System architecture definition, with the main goal
of proposing an Internet-based architecture.
3. Modelling aimed to define the Fire Office domain
in terms of an XML language.
4. Simulation procedure design and implementation,
with the objective to implement a model-based
performing of tasks listed above.
This paper is structured into the following
chapters: a brief description of the problem; the
system architecture, presentation of modelling
language, and technology application example for
modelling the fire office resource management task.
Finally, a description of how the system works is
WEB-BASED TRAINING SYSTEM FOR FOREST FIRE OFFICE STAFF
185
introduced, and example of a training session is
provided.
2 PROBLEM DESCRIPTION
On fire extinguishing site, there are three main
aspects to be considered:
1. Fire behaviour;
2. Resource allocation season plan;
3. Action against fire in a burning area.
The first one is out the scope of this research
work as long as there are many projects oriented to
analyse and simulate fire behaviour, as mentioned in
the Introduction. One of the main tasks for forest
engineers specialized on fire extinguishing is to plan
resources for a summer season. One of the main and
most difficult issues that the forest engineer faces is
to choose place where resources to be allocated.
Depending on personal background, available
heuristics and statistics, the engineer must find the
most appropriate places for each resource. With
respect to this problem, tools for analysis and
assistance by decision taking are welcome – a an
improper resource allocation can lead to huge
disasters.
One of the objectives for the seasonal planning is
to obtain such resource allocation variant in a given
region, which ensures access to fire accident within
a predefined time interval, e.g. less than 30 minutes.
Then engineers, considering different resources
allocation variants and fire accident access times,
associated to the allocation variants, must decide
which of the resource will be moved from its base to
the burning area, and which other resources must be
reallocated to cover the rest of the region with the
GIS server
XNetMod server
Web server
XNetMod
database
Session
database
GIS
Application
Command
Logs
XML
Commands
S
o
c
k
e
t
/
S
O
A
P
c
o
n
n
e
c
t
i
o
n
S
o
c
k
e
t
/
S
O
A
P
c
o
n
n
e
c
t
i
o
n
S
o
c
k
e
t
/
S
O
A
P
c
o
n
n
e
c
t
i
o
n
S
o
c
k
e
t
/
S
O
A
P
c
o
n
n
e
c
t
i
o
n
XNetMod Model Builder
XNetMod Model Validator
XNetMod Model Server
XNetMod Kernel Simulator
XNetMod Identification System
XNetMod Web Server
XNetMod Front End
XNetMod ArcView Front-End
XNetMod Command Interpreter
XNetMod Arcview Kernel Simulator
XML
Simulation
Commands
XML Image
Commands
XML Image
Results
XML
Simulation
Results
XML
Configuration
Results
Figure 2. Three server architecture approach.
ICEIS 2004 - DATABASES AND INFORMATION SYSTEMS INTEGRATION
186
highest protection level. Forest engineers have to
execute two tasks:
1. To establish a planning on resources allocation
over the region. This task must be performed
before the season has begun.
2. To perform resources management, and
disposition in case of fire accident. This task
must be carried out events dependent, during the
season and considering actual forest fires and
forest fire risks.
3 SYSTEM OVERVIEW
To support both main tasks mentioned above, Figure
1 shows a web-based architecture proposed where it
is possible to identify two different components:
web clients from remote places and the training
system; they are linked together using the Internet
connection. Any Internet browser available with
Java-applet support facilities can be used for clients.
The Internet approach is appropriate to this
simulation environment because forest engineers
training process could be done remotely. As in case
of Cuenca province (Spain), resources are limited,
and sometimes it is not possible to find a convenient
place both for instructors and trainees where to
prepare the training/simulation sessions. In this
sense, this solution makes it possible for engineers to
take part in the training process from different parts
of Cuenca.
The system architecture is refined in figure 2. At
the server side, also a number of subcomponents,
namely servers, were identified during the design
phase. The design criteria applied was to achieve a
system with the highest cohesion and lowest
possible coupling. This was important because of the
complexity of the architecture and thinking in terms
of design robustness and maintenability. In this way
each subcomponent (i.e. server) could work in a
non-dependent way from others. At a first sight,
three different servers form this system:
1. Web server, represents the front-end server to
manage web client connections.
2. XNetMod server, represents the server where all
the model information is stored.
3. GIS server, represents the main system source
information.
This multi-server solution proposed allows
implementing independent services to support each
part of the system. XML language has been selected
to exchange information because of its flexibility
and its the standardization.
Each server sends XML files with commands
requesting information to other servers; and XML
files are sent back with results of the execution of
those commands. A local protocol has been
developed to control this information exchange
process using either sockets or SOAP connections.
This multi-server schema has proved as a solution
with high flexibility, scalability and easy to extend.
4 XNETMOD MODELLING
LANGUAGE
Once the architecture has been defined, the next step
is to present how the system is modelled and the
simulation performed. Roadmap networks is the
base concept for define the model and to describe
roadmap networks requires the following issues:
To describe nodes of the network
To describe relations between nodes
Each node represents some kind of information
that networks could manage, like topological nodes
or elements allocated to topological nodes. For
example: in a roadmap network, topological nodes
could be, for example populations, or crossroads
(Molina, Castellnou, Plant, 2001). But in most of the
cases, it is also interesting to describe all the
elements that could be allocated to nodes like
hospitals, hotels, or gas stations. The second type of
elements are links. Each link represents network
relations. On one hand links could be topological
relations between nodes to describe the network
topology. On the other hand, links could represent
how some resources are allocated to nodes.
P2 P1PinC2 PinC1 PinC3
A1
V1
CR AB
Figure 3. Network components sample
WEB-BASED TRAINING SYSTEM FOR FOREST FIRE OFFICE STAFF
187
In figure 3 a simple scheme for a road
representation is provided with the following
elements:
Four topological nodes: P1 (city of Cuenca) , P2
(village of Villalba de la Sierra), CR1 (a
crossroad) and AB1 (airport at a crossroad).
Two resource nodes: A1 (aircraft 1) and V1 (big
truck 1)
Two resource allocation links:
1. V1 to P1
2. A1 to AB1
Three topological links:
1. PinC1: P1 to CR1 (highway 1)
2. PinC2: P2 to CR1 (local –path- 2)
3. PinC3: P1 to AB1 (local –path- 3)
In most of cases, each type of link has their own
regularization rules. These rules describe all the
constraints that each connection must check to be
valid. That is the reason why links could be
represented as a component with two elements:
a) Pin connections, to define the existing element-
to-element relations within an application model.
b) Rules, to define regularization rules with respect
to a connection type and within an instance of
Links.
The instances of class Rule play an important
role for the model structure verification
(XNETMOD Consortium, 2003). From a formal
point of view the rule based verification is founded
on the work presented in (Stefanescu, 2000). The
model is interpreted with respect to existing
connections defined and considering these rules.
Thus, when the model checker is running using the
interpreter, the defined rules will be applied to the
model structure.
Some topological rules could be:
1. Populations must be linked together with roads
(Rule1)
1. Populations could be linked with crossroads
(Rule 2)
2. Some resource allocation (allocation “pins”)
rules could be:
Hospitals must be allocated into
populations
Gas stations could be allocated everywhere
A brief example of a model is:
<Model xmlns:xsi =
"http://www.w3.org/2001/XMLSchema-
instance" xsi:schemaLocation =
"http://www.w3.org/2001/XMLSchema/Model
Model.xsd" Name="Model">
<Node NodeId="A1" description="Aircraft
1" cost="3.14159" xsi:type="Node:Air">
<Node:speedAvg>500</Node:speedAvg>
</Node>
<Node NodeId="V1" description="Big
Truck 1" cost="1000"
xsi:type="Node:Vehicle">
<Node:speedAvg>100</Node:speedAvg>
<Node:costKM>10.3</Node:costKM>
</Node>
<Node NodeId="P1" description="Cuenca"
x1="200" y1="200"
xsi:type="Node:Population">
<Node:habitNumber>50000
</Node:habitNumber>
</Node>
<Node NodeId="P2" description="Villalba
Sierra" x1="125" y1="250"
xsi:type="Node:Population">
<Node:habitNumber>5000
</Node:habitNumber>
</Node>
<Node NodeId="CR1" description="String"
x1="100" y1="150"
xsi:type="Node:Crossroad">
<Node:Risk>100</Node:Risk>
</Node>
<Node NodeId="AB1"
description="Airport" x1="200" y1="250"
xsi:type="Node:Crossroad">
<Node:Risk>100</Node:Risk>
</Node>
<Link linkId="topology"
Name="topology">
<Pin PinId="PinC1" description="Highway
1" xsi:type="Pin:Highway" from="P1"
to="CR1">
<Pin:limit>100 120</Pin:limit>
<Pin:minSpeed>60</Pin:minSpeed>
<Pin:lines>2</Pin:lines>
</Pin>
<Pin PinId="PinC2" description="Local
2" xsi:type="Pin:Local" from="P2"
to="CR1">
<Pin:limit>80 100</Pin:limit>
<Pin:status>Good</Pin:status>
</Pin>
<Link:Pin PinId="PinC3"
description="Local 3"
xsi:type="Pin:Local" from="P1"
to="AB1">
<Pin:limit>80 100</Pin:limit>
<Pin:status>Good</Pin:status>
</Link:Pin>
<Link:Rule RuleId="Rul1"
description="Rule 1"
ICEIS 2004 - DATABASES AND INFORMATION SYSTEMS INTEGRATION
188
xsi:type="Rule:Instance"
from="Population" to="Population"/>
<Link:Rule RuleId="Rul2"
description="Rule 2"
xsi:type="Rule:Instance"
from="Population" to="Crossroad"/>
<Link:Rule RuleId="Rul3"
description="Rule 3"
xsi:type="Rule:Instance"
from="Crossroad" to="Crossroad"/>
</Link>
<Link linkId="Allocation"
Name="Allocation">
<Link:Pin PinId="All1"
description="A1toAB1"
xsi:type="Pin:Member" from="A1"
to="AB1"/>
<Link:Pin PinId="All8"
description="V1toP1"
xsi:type="Pin:Member" from="V1"
to="P1"/>
<Link:Rule RuleId="Rul11"
description="Rule 11"
xsi:type="Rule:Instance"
from="Resource" to="Base"/>
</Link>
</Model>
5 INTEGRATED SYSTEM AT
WORK
Our main goal in this sense is to obtain the most
flexible and application independent system that is
the reason why we decided to develop this GIS
server. Most of GIS system has defined internal
programming languages to develop scripts to
automate some operations. Many scripts have been
developed to deal main simulation operations over
Arcview® (ESRI, 2003) and to export Arcview®
information to the model in XML format.
So, if GIS system changes, it is only necessary
write scripts in the language maintaining the same
command files format. In our particular application,
we used Arcview® and Avenue® (script language)
as GIS application. The following program code is
an example of an Avenue script to create a new
Arcview theme to represent air resources (aviones in
Spanish). Required fields are identification,
maximum speed, water resources, staff, efficacy
rate, node and node layer, that in the script are in
Spanish.
'***************************************
**
'--- Script to build air resources theme
'***************************************
**
aView = av.FindDoc("MONTES Y VIAS
PECUARIAS")
theTheme=aView.FindTheme("Aviones.shp")
if (theTheme<>nil) then
aView.DeleteTheme(theTheme)
end
aFile =
"C:\XNetMod\Acebo\Capas\Salida\Aviones"
.AsFileName
aNewFtab = FTab.MakeNew (aFile,
Polygon)
'--- add required fields ---'
aNewShapeField =
aNewFtab.FindField("Shape")
aNewField = Field.Make ("Id",
#FIELD_SHORT , 20, 0)
aNewField1 = Field.Make ("Vel_Max",
#FIELD_SHORT , 20, 1)
aNewField2 = Field.Make ("Cap_Agua",
#FIELD_SHORT , 20, 0)
aNewField3 = Field.Make ("Personal",
#FIELD_CHAR , 80, 0)
aNewField4 = Field.Make ("Coef_Efec",
#FIELD_SHORT ,20, 0)
aNewField5 = Field.Make ("Node",
#FIELD_LONG , 12, 0)
aNewField6 = Field.Make ("Capa_Node",
#FIELD_CHAR , 20, 0)
aNewFtab.AddFields({aNewField,aNewField
1,
aNewField2,aNewField3,aNewField4,aNewFi
eld5, aNewField6})
aNewFtab.Flush
aNewTheme = Theme.Make(
aNewFtab.GetSrcName)
aView.AddTheme(aNewTheme)
'--- turn on the theme ---'
aNewTheme.SetVisible(True)
'--- update the screen ---'
aView.GetDisplay.Flush
To achieve an effective interaction with the GIS
server has been a complex task. Each request
containing an action is specified by two different
messages in XML, the first goes from any other
server to the GIS server and represents the action
itself, and the second goes from the GIS server to the
server from where the request was sent and contains
WEB-BASED TRAINING SYSTEM FOR FOREST FIRE OFFICE STAFF
189
the requested information. Some actions
implemented are related to start and stop the GIS
application, get actual system views, move resources
from source to target. It includes ground vehicles
(coches terrestres in Spanish), and intersections
(intersecciones). A command example is:
<?xml version="1.0" encoding="iso-8859-
1"?>
<commands>
<command name="move" mode="internal">
<parameter name="resource"
type="coches terrestre.shp" value="3"/>
<parameter name="from"
type="intersecciones.shp" value="269"/>
<parameter name="to"
type="intersecciones.shp" value="273"/>
</command>
</commands>
And finally, an example of some information
exported from Arcview in XML format, including
information on villages (pueblos), and the village of
Albalate de Noguera:
<node nodeId="80"
xsi:type="Node:pueblos">
<pueblos_id>494</pueblos_id>
<area>150928</area>
<perimeter>1961.7</perimeter>
<pueblos_>490</pueblos_>
<nombre_pue>Albalate de
Nogueras</nombre_pue>
</node>
<link linkId="NodesTopology"
type="Topology">
<pin pinId="0"
xsi:type="pin:SimpleTopological">
<pin:description></pin:description>
<pin:bidirectional>true</pin:bidirectio
nal>
<pin:from>45</pin:from>
<pin:to>360</pin:to>
<pin:roadId>2622</pin:roadId>
<pin:length>8635.13</pin:length>
<pin:roadType>5</pin:roadType>
</pin>
</link>
6 TRAINING SESSION
Before the training session can be started, the system
administrator must define the simulation framework.
This framework is formed by a roadmap network
and lists of available resources (one list for each
resource type).
Once the system has been configured, the
training session is ready. The first part of the
simulation has the objective to prepare seasonal
planning. In this part, forest engineers must decide
where to allocate available resources in different
nodes of the network. At the same time as the
resources are being allocated; the system determines
the fire accident access time limit for the resources
depending on geographical situation. In parallel, an
optimal resources distribution will be obtained by
the system. Once the initial resources allocation
process is finished, the forest engineer can compare
this distribution with the optimal to check
similarities and disparities.
After the planning phase, the system can
simulate the appearance of a fire at random in some
parts of the region. Then, the engineer has care on
the resources movement towards the burning area,
and reallocate free resources to cover unprotected
areas. Finally, once the fire is controlled and
extinguished the system generates a report archiving
the actions provided by the engineer.
7 RELATED WORK
The project Gamma-EC (Dirk, 2003), is oriented
towards developing computer-aided tools to improve
the education and training disaster managers. This
system focuses on crisis management in general
terms, even when forest fire is one of the topics
dealt. NIST Fire Dynamics Simulator and
Smokeview, developed at NIST (NIST, 2003) is a
computational fluid dynamics model of fire-driven
fluid flow. Future of Fire Simulation, related to
show different works about fire modelling and the
nature of the fires in the Building and Fire Research
Lab at NIST. The application of a GIS system for
similar purposes that those described within this
paper can be found in (Lymberopoulos,
Papadopoulos, Stefanakis, Pantalos, Lockwood,
1996) and (Stefanakis, E., 2000). However the
degree of integration with the rest of the application
and functionalities achieved are higher in our
approach.
Concerning implementation of the client server
architecture, similar approaches can be found in
(Alarcón, Garbajosa, Yagüe, Garcia, 2002),
ICEIS 2004 - DATABASES AND INFORMATION SYSTEMS INTEGRATION
190
(Altendorf, Hohman, Zabicki, 2002) and
(Milenkovic, Robinson et al., 2003).
8 CONCLUSION
The developed architecture was applied for the
solution of an actual fire extinguishing problem in
Cuenca province (Spain). This system will be
included as powerful tool into Cuenca forest
engineers training process.
The developed system has proved its flexibility,
platform independency and scalability.
Consequently, the developed XML-based language
was applied for the modelling of road network, fire
behaviour, and physical resources (human teams and
vehicles) distribution related to the road network.
Herewith, the access of small and medium-sized
enterprises (SMEs) to the advanced modelling
technologies will sustain and strengthen their
business fields as well as guarantee further
diversification of intelligent IT-based services.
ACKNOWLEDGEMENTS
The work described is being developed in the
framework of a European research and development
project XNetMod "XML Based Modelling
Language for Simulation of Technical Networks"
(CRAFT-IST Contract No. 2001-52057). The
consortium includes two RTD performers – the
Fraunhofer IPK (Berlin / Germany) and the
Universidad Politécnica de Madrid, Escuela
Universitaria de Informática (Madrid / Spain). The
first users of the technology are four industrial
companies – ELPRO Prozessindustrie- und
Energieanlagen GmbH (Berlin / Germany),
SimPower Simulator Systeme GmbH (Karlstein am
Main / Germany), Investigacion y Programas S.A.
(Madrid / Spain), and ACEBO, C.B. (Arcas de Villar
- Cuenca / Spain).
REFERENCES:
Alarcón, P.P., Garbajosa, J., Yagüe, A., Garcia, C. (2002).
Data Sources Server: An Approach to Heterogeneous
Data Integration. ICEIS 2002: 3-10
Altendorf, E., Hohman M., and Zabicki, R., 2002. Using
J2EE on a Large, Web-Based Project. IEEE Software,
Marc h / A p r i l, IEEE Press.
Andrews, P. (1998). Update and Expansion of the
BEHAVE Fire Behavior Prediction System BEHAVE.
http://www.frames.gov/tools//BEHAVE/behave_singl
e.html. Avalible in 2003.
Bardaj, M, Molina, D.M. &. Castellnou, M. (1998).
Probability of large fires: structural & meteorological
components In Proc. Joint conferences: 'Third
International Conference on Forest Fire Research' &
'14th Fire and Forest Meteorology Conference',
(Viegas, D.X. ed.) ADAI, University of Coimbra,
November 16-20, 1998, p. 957-974.
ESRI (2003).
http://www.esri.com/. Available in 2003.
Guardia Civil de España, (2003).
http://www.guardiacivil.org/00prensa/actividades/ince
ndios2002/estadisticas.asp. Available in 2003
Instituto Nacional de Estadística de España, (2003).
http://www.ine.es/inebase/cgi/um?M=%2Ft26%2Fa015%
2Fa1998%2F&O=pcaxis&N=&L=0. Available in
2003
Ministerio de Medio Ambiente de España, (2003)
http://www.incendiosforestales.org/estadisticas.htm.
Available in 2003
Lymberopoulos, N., Papadopoulos, C., Stefanakis, E.,
Pantalos, N., Lockwood, F. (1996). A GIS-based
forest fire management information system. EARSEL
Journal, 4(4), pp. 68-75.
McGrattan, K. B., Baum, H. R., Rehm, R. G., Forney, G.
P., Prasad, K. (2002). Future of Fire Simulation. Fire
Protection Engineering.
http://europa.eu.int/comm/research/leaflets/disasters/e
n/forest.html
Milenkovic, M., Robinson, S.H., Knauerhase, R.C.,
Barkai, D., Garg, S., Tewari, V., Anderson, T.A:,
Bowman, M., (2003). Toward Internet Distributed
Computing, Computer, May, IEEE Press
Molina, D., Castellnou, M., Plant R. (2001). MARF -
Qualitative simulation model to automatically refresh
the forest fuel type layer (to be use in Farsite) under
both temporal and spatial changes.. Proc. of the
Workshop on Tools and methodologies for fire danger
mapping (IEFC and UTAD). Vila Real, Portugal.
NIST (2003).
http://www.nist.gov. Available in 2003
Stefanescu G., (2000). Network Algebra. Springer-Verlag.
Berlin Heidelberg
Stolk, D. (2003). GAMMA-EC. Computer assisted
education and training in disaster management.
http://www.tno.nl/instit/fel/gamma_ec/. Available in
2003
Stefanakis, E. (2000). The Synergy of GIS with Other
Systems. IAPRS, Vol. XXXIII, Amsterdam, 2000
XNETMOD Consortium (2003). XML Based Modelling
Language for Simulation of Technical Networks. Final
Project Report. CRAFT-IST Contract No. 2001-
52057.
WEB-BASED TRAINING SYSTEM FOR FOREST FIRE OFFICE STAFF
191