mical
saster
of the
ration
| is to
(geo)
cation
export
aware
ntext-
text,
s with
, GIS,
inition,
ed on
that is
ions of
several
ely not
lack of
current
l's) at
E, etc.)
, W3C)
n most
on, ie.
ailable.
f many
nnected
‘ordable
capable
'ence to
S (with
esponse
ropriate
| to the
rent and
n detail.
arch and
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
2. THE USERS AND THE SYSTEM
Disaster management is an application that involves a very
wide group of users. Figure 1 shows a categorisation of the
users according to the environments they are working in
(related to their tasks in the emergency operations). Two
general categories of users can be distinguished, i.e. teams
working in wireless environments, on the field (indoor or
outdoor) and users working in wired environments (indoor) in
management centres and related institutions (Zlatanova and
Holweg, 2004). The users in wired environments can be
subdivided further into users working in VR environments
(controlling, analysing and managing), Desktop environments
(advising on particular situations and occasions) and accessing
information through the Web (wide audience, press, etc.). In the
centres for coordination, various pieces of information have to
be assembled for decision-making. In the field, the workers
need information about the current situation and prognosis for
the immediate development in their area. Moreover the field
workers can collect information to be returned to the central for
analysis and redistribution. These generalised activities impose
a variety of requirements such as consistent information in any
environment, search, analysis and processing of information on
the Internet and distributed databases, real-time data update,
routing outdoor working teams, individual and intuitive
visualisation on different devices to support decision-taking.
Wireless Nelwork
Internet&Intranet
Figure 1: The users in a disaster management situation
3. GEO-DBMS FOR DISASTER MANAGEMENT
Most of the important data and information necessary for the
support of such a system are spatially related; a geo-component
is of special relevance. Amongst all the systems dealing with
geo-information, DBMS are the fundamental componet.
c
3.1 State of the art in Geo-DBMS
The integrated architecture of storing geometric data and
relationships together with administrative data in DBMS’s is
now getting mature. The importance of the integrated
architecture was recognised by the industry and the OpenGIS
consortium standardised the basic spatial types and functions
(i.e. Simple Feature Specification, SFS) (OGC, 1999). ISO (ISO
TC211, 2003) and OpenGIS agreed to harmonize their geo-
information standards and specifications. Several commercial
DBMSs are available with support for spatial data type: Ingres,
Oracle, Informix or IBM DB2. In addition several heavily used
non-commercial DBMS have geo-information support:
PostgreSQL (with roots more than two decades ago) and
629
MySQL (since the most recent version in 2004). Also more and
more originally commercial CAD and GIS packages support the
integrated architecture: ESRI, MapInfo, Intergraph Bentley or
AutoDesk. Even one of the DBMSs (Oracle Spatial) has started
supporting 2D topology in the most recent version (10g).
Currently, the main attention is on 2D spatial data types, but 3D
geometric objects can be maintained as well. Research on 3D
has resulted in defining a 3D geometry data type (Stoter & van
Oosterom, 2002, Arens et al., 2003). Recent experiments and
benchmarking have clearly shown a significant progress in
DBMS's performance (van Oosterom et al., 2002). Loading and
querying spatial information is still more elaborate than
semantic (attribute) data but the response time is compatible
and can be tuned to meet requirements of disaster management
where the response time is of critical importance (Zlatanova et
al, 2003).
3.2 Current and future developments
Despite the progress shown within DBMSs developments, still
a number of generic issues needs to be addressed in order to
provide service to multi-risk management:
e Extending the management functionality into the third
dimension is a research question of critical importance for
fighting crisis situations in urban areas. For example, true
three-dimensional data (instead of 2D maps), supplied
directly to a fire brigade working in a dense built-up area
will increase the possibilities for orientation and reduce
time, which significantly improves the effectiveness and
safety of rescue teams.
e Developing 3D models and frameworks for management
of different topologies at database level and corresponding
operations. This will allow extension of the spatial
functionality to be able to perform 3D routing,
generalisation, and adaptation of different types of data,
and consistent field update of data. Some explicit examples
of implementations related to mobile environments are
map orientation, map generalisation, bounding box, route
maps, map styles, and colouring. Some of this
functionality can be provided as a generic set of operations
at a database level.
e Frameworks for describing multidimensional spatial
relationships, structures for maintenance of
multiresolution, multidimensional and historic geo-data
need further research. Investigating and developing
multidimensional and multimedia data models for efficient
organisation of large urban models, utilising and
elaborating different international specifications including
ISO, OpenGIS etc.
e Data update with newly collected data from the field can
be very critical for both a) monitoring the disaster event
and b) giving instructions to the involved people. From a
database point of view, this process requires strict
consistency rules for integration with existing models and
immediate propagation of the information to all the users.
In this respect, extended models for maintenance of
historic information (to be used also for prognosis and
future scenarios) are becoming especially desirable.
e A major issue in disaster management is the presentation
of the information. It has to be prepared in the best
appropriate form. For example, navigation instructions to
the closest exit can be presented as voice, text, simple
graphics, animation or even 3D graphics. The time for
delivering such formats differs significantly. The system
has to be able to understand the context (type device,
