Integration of Earth Observation and in-situ spatial data for
the
development of a Decision Support Tool for Technological Risk
Management
Foundation for Research and Technology – Hellas, Institute of
Applied Mathematics,
Regional Analysis Division, FORTH - IACM, Vassilika Vouton, P.O.
Box 1527,
GR-71110, Heraklion, Crete, Greece.
(tel. +30 810 391762, fax. +30 810 391761, e-mail:
[email protected])
Abstract. In this study the potential of Earth Observation
techniques for major
technological accidents analysis is presented and a GIS platform is
proposed for
the support of technological risk management. This GIS tool is
based on the
detection and space-time monitoring of the produced plumes by
integrating
moderate and high resolution satellite imagery and in-situ vector
data. The
Advanced Very High Resolution (AVHRR) on board the NOAA satellites
has
been used for the detection of fire as well as for the detection
and monitoring of
plumes caused by major technological accidents. The detection
algorithms have been
presented in previous studies for past accidents in Europe. AVHRR
images, were
adjusted to the broader area of Athens (Greece) in order to develop
a major
technological accident scenario based on real plumes. This scenario
was used to
present the functionality of the developed GIS platform for the
support of decision
making during the crisis, as well as for the assessment of the
accident’s impact to
natural and human environment.
Key Words: Technological Risk Management, Earth Observation, Fire
and Plume detection, Decision Support System, Impact
assessment.
1. Introduction
Major technological accidents at industrial sites impose
considerable dangers
for the environment and public health. A number of accidents have
taken place
in recent years with serious costs to human life as well as with
considerable -
and in many cases irreversible - damages to the natural
environment. In many
cases toxic substances or released airborne material develop into
plumes which
may create high concentrations at ground level and pose dangers to
the human
and natural environment. Damages may thus occur both as an
immediate and
direct consequence of the accident, and subsequently during
propagation and
dispersion of the resulting plume. It should be mentioned that in
several cases,
while exhaustive consideration has been given to the immediate
ground-level
effects in close vicinity to the installation, only limited effort
is usually given to
examining the impacts of the plume in the wider geographic area
during the
course of the hours or days following the accident.
Risk analysis forecasts the likelihood of accidents, estimates the
consequences
of an accident which is likely to occur, works out strategies to
prevent accidents
and assesses the adverse impacts in the event that an accident
occurs.
Technological accidents can be characterized by a number of
different events
and processes, including spillage or sudden release of materials,
fire, or
explosion. The most common effect is the release of gases and
liquids used and
processed in the installations concerned. Fire and explosion are
also common
effects, while a combination of the above is not rare. Releases may
be toxic and
can be either to land and water or to air. Airborne releases
usually develop in
plumes, which can thereafter be monitored either due to their
optical depth or
their temperature difference from the ambient air.
A number of experiments (Davie and Nolan 1993, Lang 1993, Bartelds
et al. 1993,
Atkinson et al. 1994, Miles Cox 1994, Marliere 1996, Grant and
Drysdale 1994,
Martins and Borrego 1994, Porter et at. 1996, Martins et al. 1996,
Cozzani et al. 1996)
involving fires in warehouses and awareness of the hazards of
plumes in a fire
situation have been oriented towards the definition of the
properties and of the
amount of the plume particulates generated by different materials,
including
pesticides, under varying fire conditions. The emission of toxic
gases and
formation of carbon particles which can carry toxic material
absorbed onto their
surface as well as obscuring vision both present hazards in a fire
situation. Over
150 substances or groups of substances are commonly identified in
accidents.
The substances occurring most often include ammonia, chlorine,
hydrogen,
hydrogen chloride, nitrogen oxides and sulphur dioxide, all of
which are bulk
chemicals.
In past studies various numerical models have been developed for
the
simulation of the conditions and the process of technological
accidents (Hanna
et al. 1993, Webber et al. 1993, Andronopoulos et al. 1994,
Kukkonen et al.
1994, Cleaver et al. 1995). Software packages have been also
developed for
technological risk management; some of these software packages are
WHZAN
(Technica 1992), RISKIT (VVT 1993), EFFECTS (TNO 1991), SAVE
(TNO
1992) and MAXCRED (Khan and Abbasi 1999).
In this study, the design and implementation of a technological
management
support tool is presented. This tool is based on the detection and
space-time
monitoring of the produced plumes by integrating moderate and high
resolution
satellite imagery and vector data in a GIS platform, capable to
support
technological risk management. A major technological accident
scenario has
been also developed using satellite images acquired during real
past events.
This scenario is used to present the functionality of the developed
GIS-based
system for the support of decision making during the crisis, as
well as for the
assessment of the accident’s impact to natural and human
environment.
2. Fire and plume detection methodology
The methodology for the detection of fires caused by major
technological
accidents with the use of AVHRR imagery has been presented in a
past study
(Chrysoulakis and Cartalis 2000). The detection algorithm carries
the
advantages of a multispectral analysis and provides valuable
results for the
detection of fires caused by technological accidents. In practice,
the
pseudochannel image of brightness temperature difference between
AVHRR
channels 3 (3.55 – 3.93 µm) and 4 (10.5-11.3 µm) is created and
filtered for
clouds by applying a cloud-masking algorithm which is based on
the
combination of AVHRR channels 1 (0.58 – 0.68 µm) and 5 (11.5-12.5
µm). In
this filtered image, pixels with brightness values greater than an
experimental
derived temperature threshold correspond to fires produced by
major
technological accidents. Fire detection algorithm has been
programmed as a
stand alone application for the automatic detection of fires caused
by major
technological accidents with use of AVHRR imagery (Chrysoulakis
and
Cartalis 2002a).
The methodology for the automatic detection and monitoring of
plumes caused
by industrial accidents with the use of AVHRR imagery has been also
presented
in a previous study (Chrysoulakis and Cartalis 2002b): A
two-dimensional
feature space image is used in order to discriminate pixels that
contain plumes
from those that may contain clouds or the underlying surface. The
two-
dimensional feature space is generated by combining AVHRR channels
1, 2
(0.72-1.10 µm) and 5.
Both methodologies have been evaluated on the basis of past
major
technological accidents (Thessaloniki, February 24, 1986; Lyon,
June 2, 1987;
Genoa, April 13, 1991; Enschede May 13, 2000). The effectiveness
and
reliability of these algorithms was found to be satisfactory in all
case studies.
AVHRR images acquired over the broader area of Netherlands on May
13,
2000 (14.44 UTC and 17.20 UTC) were used for the development of a
major
technological accident scenario in this study. This date refers to
the massive
explosion in a firework factory in the town of Enschede. The
produced fire was
detected by applying the aforementioned fire detection algorithm to
the
AVHRR images, whereas, in order to simplify the scenario, AVHRR
channel 2
images were used for the monitoring of the produced plumes (figure
1).
3. Design of a GIS based tool for the support of technological
risk
management
The problem of monitoring the atmospheric results of major
industrial accidents
has been addressed by the EC Directive 82/501/EC regarding Major
Accident
Hazards of Certain Industrial Activities. This Directive, which is
often known
as the Seveso Directive, has set the obligation for major
industrial installations
in Member States to develop ‘Emergency Plans for the Control of
Major
Technological Accidents’. Typically, emergency plans provide
exhaustive
consideration of the immediate on-site effects of such accidents,
but give much
more limited attention to the wider and longer-term effects
generated by the
pollutant plume produced by the incident. In the Directive, the
principles for
regulating industrial hazards are laid down so as to prevent major
accidents,
and, should one occur, to limit the consequences for man and the
environment.
In 1995, a fundamental review of this Directive was accepted by the
Council of
Environment Ministers and the European Parliament. The review laid
down a
demand for the establishment of safety systems management
structures.
Measures to be taken included land use planning, the assessment of
substances
on the basis of their toxicity/hazardousness, and evaluation of
potential side
effects. The main factors that need to be covered in the course of
preparing an
Emergency Plan for Major Industrial accidents range from the
definition of land
use/cover and the potentially exposed population to the rate and
direction of
propagation of the produced plume. An analytical description of
these needs is
given in table 1. In this table the potential of Earth Observation
to meet these
needs is also recorded as well as a synopsis of the satellites
which can be used in
conjunction with the requirements.
Figure 2 presents the architecture of the proposed GIS based
platform. Satellite
data are the main information sources in this system. The
aforementioned fire
detection and plume detection and monitoring algorithms are used
for the
analysis of NOAA/AVHRR imagery. These algorithms operate
independently,
but the results of the first algorithm are compared with the
results of the second.
If both results are positive (this means that an industrial fire
and a plume have
been detected in the same area), the system will classify the
corresponded pixels
as ‘potential incident pixels’. Since the AVHRR images are
geometrically
corrected, the coordinations of these pixels are automatically
stored in the
system.
As it can be seen in figure 2, the areas in which a technological
accident is most
possible to occur (industrial installations, warehouses of toxic
substances,
offshore installations, petroleum products storage sites, ports,
airports etc.) have
been mapped with the use of high and very high spatial resolution
satellite
imagery in combination with other sources of information (CORINE
land cover
data, land use maps, ground survey maps etc.). These areas have
been classified
as ‘possible areas of occurrence’.
Therefore, the system examines if the detected ‘potential incident
pixels’ are
located within any possible area of occurrence. A positive result
will verify the
occurrence of a technological accident and will comprise an ‘alarm’
for the
system. This alarm enables the plume monitoring modules as well as
the
modules used by the system for the estimation of the population at
risk. The
latter combine the high spatial resolution satellite derived
information (urban
areas, road network) with in-situ derived vector data (spatial
distribution of
population in the vicinity of the incident).
During the crisis phase, the GIS-based tool has the
potential:
• to support (in terms of input data and information on
boundary
conditions) dispersion models usually employed in Emergency
Plans
(with the use of NOAA imagery);
• to describe and map the characteristics of the natural
environment (e.g.
land cover, land use) and the population, in the area of concern,
thereby
providing a means to assess the environmental and health impacts of
the
plume (with the use of high and very high spatial resolution
satellite
imagery and in-situ data);
• to provide the state of the environment in the vicinity of the
accident and
to feed the aforementioned models with topographic information
(based
on high and very high spatial resolution satellite imagery);
• to verify simulation models results with respect to the position
and the
speed of propagation of the plume in the atmosphere (based on
NOAA
imagery);
• to help identify at-risk populations and environments, which may
need
special protection by providing the land cover and the spatial
distribution
of population along the predictable plume propagation direction and
by
calculating the number of persons expected to be straightway
affected,
according to the plume speed of propagation (with the use of high
and
very high spatial resolution satellite imagery for the depiction of
the
urban areas and in-situ data for the monitoring of the spatial
distribution
of population);
• to support the coordination among involved parties as well as
the
decision making regarding the evacuation of urban areas to avoid
plumes
(with the combined use of high spatial resolution satellite imagery
and
in-situ data for the depiction of road network together with urban
areas
and spatial distribution of population along plume
propagation
direction);
• to support modelling of exposures and health risks through the
improved
monitoring of the environmental and social characteristics of the
area
(with the combined use of high and very high spatial resolution
satellite
imagery and in-situ data);
During the post-crisis assessment phase, the GIS-based tool has the
potential:
• to thematic map disaster areas (based on NOAA and high
spatial
resolution satellite imagery);
• to support the assessment of the impacts to the local
(anthropogenic and
natural) environment due to the produced plume (with the combined
use
of NOAA and high spatial resolution satellite imagery and in-situ
data);
• to help design and plan follow-up strategies to deal with
the
consequences of accidents (e.g. long-term monitoring of human
health or
environmental impacts, health support, environmental
clean-up);
The main advantage of the proposed GIS tool is the wide area
coverage with
very good spatial and temporal resolution. The main disadvantages
are:
a) It has the potential to detect and monitor incidents which cause
fire
and/or explosion, whereas equally common incidents such as release
of
dangerous substances can not be detected with the use of
Earth
Observation techniques. Release of dangerous substances is
often
expected as gases and liquids are used and processed in the
installations
and, in case of damage to the container/reactor, loss of
containment takes
place.
b) The observation of a given area with the use of AVHRR (swath
width
2400 km) does not accomplished in a continuous basis, because
NOAA
are polar orbiting satellites. However, the simultaneous operation
of a
system of two NOAA satellites results in increase of their
temporal
resolution, especially in mid and high latitudes.
3. Application: A major technological accident scenario for the
region of
Athens
Athens concentrates about half of the population of Greece. A lot
of refineries,
chemical industries and warehouses are located in the broader
region of Athens
and especially within the industrial zone of ‘Thriasio Pedio’,
about 15 km NW
from the centre of the city. For the application of the proposed
GIS tool, taking
into account that the prevailing winds in the area of concern are
from N – NW
directions, a major technological accident scenario for a refinery
installation at
Thriasio Pedio was generated. In the past, similar accidents took
place in this
area (i.e. Petrola refinery, February 1, 1992), but for the
application of the GIS
based tool in this study, real plumes depicted on AVHRR imagery
were needed.
For this reason, the available AVHRR data of May 13, 2000, were
used (major
technological accident in Enschede). Rectangular portions (50 x 30
km), around
the pixels corresponded to plumes, were extracted from these AVHRR
images
and adjusted to the broader area of Athens. High spatial resolution
satellite
imagery and in-situ vector data have been also used. More
specifically, for the
development and implementation of the scenario the following data
were
integrated:
a. NOAA/AVHRR images (May 13, 2000, at 14.44 UTC and 17.20
UTC).
The fire detection algorithm was applied to these images in order
to
detect the pixels corresponded to the accident site in Enschede.
Since
pixels corresponded to fire caused by this accident were
detected,
AVHRR images were adjusted to the area of Athens by
corresponding
these pixels to the position of the refinery at the Thriasio Pedio.
A
transverse Mercator projection was applied (Projection
System:
Hellenic Geodetic Reference System 87 - HGRS87; Reference
Ellipsoid: GRS80). Figure 3 presents the final position of the
filtered
pseudochannel image of brightness temperature difference
between
AVHRR channels 3 and 4, after the application of the fire
detection
algorithm and the adjustment to the area of Athens. As it can be
seen in
Figure 3, the pixels for which the brightness temperature
difference
between AVHRR channels 3 and 4 peaks (very bright tones,
indicative
for the accident site) are now located over the area of Thriasio
Pedio.
Figure 4 presents the position of the plume at 14.44 UTC
(AVHRR
channel 2), whereas figure 5 presents the position of the plume
about
2.5 hours later (AVHRR channel 2, 17.20 UTC).
b. Landsat Thematic Mapper (TM) image (April 26, 1994).
Ground
Control Points (GCP’s) were used for the geometric correction of
this
image and for its projection to the HGRS87 system. The TM image
was
used for the depiction of urban areas and of the main road network,
as
well as for the depiction of the industrial zones in combination
with
land use maps. It was also used for the estimation of the state of
the
environment in the area of concern. For example, the spatial
distribution
of vegetation was estimated with the use of NDVI (Normalized
Deference Vegetation Index), which is based on the combination of
TM
channels 3 (0.63 –0.69 µm) and 4(0.76 – 0.90 µm).
c. CORINE Land Cover data and land use maps. These data were used
in
combination with Landsat TM imagery for the depiction of areas
in
which a technological accident is most possible to occur (possible
areas
of occurrence).
d. 100 m contours were used for the production of a Digital
Elevation
Model (DEM) for the broader area of Athens. This DEM is used to
feed
the dispersion models with topographic information, as well as
to
produce 3D views of the landscape with the combined use of
Landsat
TM imagery.
e. Vectors of the main and secondary road networks for the broader
area
of Athens. The integration of these vectors with the Landsat TM
image
has the potential to support either the development of emergency
plans
for the area of concern, or the decision making during the
crisis
mitigation phase.
f. Vectors of the spatial distribution of population for the
broader area of
Athens. These vectors were used by the GIS platform in order
to
support the decision making during the crisis mitigation
phase,
especially regarding the evacuation of urban areas to avoid plumes.
The
system has the capability to combine these vectors with the
satellite
derived information in order to estimate the population at
risk.
4. Results
The possible areas of occurrence are presented in red colour in
figure 6. These
areas are superimposed on a pseudocoloured composition (RGB: 3-2-1)
of
Lansat/TM channels 1, 2 and 3. As it has been shown in figure 2,
the system
examines if the detected ‘potential incident pixels’ are located
within any
possible area of occurrence. According to the scenario used in this
study, the
area represented by the detected ‘potential incident pixels’
(pixels in bright
tones in figure 3), is located within a possible area of occurrence
(area within
the yellow polygon in figure 6).
Figure 7 presents the physical and artificial characteristics of
the area located
around the accident’s site. Three sources of information have been
used: a)
NOAA/AVHRR imagery for the detection of the exact position of
the
technological accident (yellow hatched area); b) Landsat/TM imagery
for the
monitoring of the area located around the accident’s site; c)
vectors of the main
and secondary road network. Figure 7 can be used to inform decision
making
Authorities about the accessibility of the area of interest, as
well as about its
artificial characteristics, which may be used during the crisis
mitigation phase
(i.e. airports).
Figure 8 presents a product of the ‘monitoring module’ of the
proposed GIS
platform. The selected GIS layers in figure 8 present:
• the urban areas (Landsat/TM pseudocoloured image 3-2-1);
• the spatial distribution of vegetation, in green tones
(Landsat/TM,
NDVI);
• the possible areas of occurrence, in red colour (from Landsat/TM,
land
use maps and CORINE land cover data);
• the position of the accident’s site, within the yellow polygon
(from
NOAA/AVHRR channels 3 and 4);
• the position of the produced plume at 14.44 UTC, within the
vertically
hatched black polygon (from NOAA/AVHRR channel 2);
• the position of the produced plume at 17.20 UTC, within the
horizontally
hatched black polygon (from NOAA/AVHRR channel 2);
This product may be used for the monitoring of urban and natural
disaster areas
as well as for the location of areas where high ground level
concentrations of
toxic substances are expected. It can be also used for the
estimation of the
horizontal propagation velocity of the produced plume. For the
scenario used in
this study, the analysis of this product indicated that, the mean
plume
propagation velocity along the main propagation direction (NW to
SE) was
about 3.5 Km/h, whereas its mean diffusion velocity in
perpendicular direction
was about 0.8 Km/h. Therefore, this product has the potential to
offer valuable
information to the decision making Authorities, either during the
crisis
mitigation phase, or during the post crisis assessment. However,
for the
estimation of the population at risk during the propagation of the
toxic plume,
the spatial distribution of population was needed.
The proposed GIS tool has the capability to integrate moderate
spatial
resolution satellite data with the spatial distribution of
population for the area
located around the accident’s site in order to estimate the
population at risk.
Figure 9 presents a product of the ‘population distribution module’
of the GIS
platform. The selected GIS layers in figure 9 present:
• The spatial distribution of population between 5 and 10 Km
(blue
circles) from the accident’s site. This vector has been produced
using in-
situ spatial data. Inhabited areas are presented in different
colours,
according to their population density, as it is shown at the
legend.
• The position of the produced plume at 14.44 UTC, within the
green
polygon (from NOAA/AVHRR channel 2).
• The position of the produced plume at 17.20 UTC, within the
red
polygon (from NOAA/AVHRR channel 2).
The population distribution module has been designed for the
automatic
estimation of the population at risk. In terms of the scenario used
in this study,
the number of potentially affected inhabitants at the area within a
circle centred
at the accident’s site with 5 km radius, was found to be at the
order of
magnitude of 15000 persons, whereas the number of potentially
affected
inhabitants at the area between 5 and 10 Km from the accident’s
site (figure 9),
was found to be at the order of magnitude of 300000 persons.
Therefore, the
integration of the spatial distribution of population with moderate
spatial
resolution satellite data has the potential to estimate the
population at risk
during the dispersion of a toxic plume, as well as to support (in
combination
with other elements of the proposed GIS platform such as the road
network) the
decision making regarding the evacuation of the disaster
area.
5. Conclusions
In this study the potential of Earth Observation techniques for
major
technological accidents analysis was presented and a GIS platform
was
designed for the support of technological risk management. This GIS
based
decision support tool carries the advantages of the detection and
space-time
monitoring of the produced toxic plumes, by integrating moderate
and high
spatial resolution satellite imagery and vector data in order to
estimate the
population at risk during an emergency, as well as to support the
assessment of
impacts of these plumes to natural and human environment.
Satellite data are the main information sources in this GIS system.
The areas in
which a technological accident is most possible to occur were
mapped with the
combined use of high and spatial resolution satellite imagery and
in-situ spatial
data. Special fire detection and plume detection and monitoring
algorithms
were used for the analysis of NOAA/AVHRR imagery. A major
technological
accident scenario was developed in order to present the
functionality of the GIS
based tool for the support of decision making during the crisis, as
well as for
the support of the assessment of the accident’s impact to natural
and human
environment.
The main advantage of the proposed GIS platform is the wide area
coverage
with very good spatial and temporal resolution. The main
disadvantages are: a)
It is able to detect the technological accidents which cause fire
and/or
explosion, whereas incidents such as release of dangerous
substances can not be
detected. b) The observation of a given area with the use of AVHRR
does not
accomplished in a continuous basis, because NOAA are polar
orbiting
satellites.
The proposed GIS based tool can be used to support the development
of
emergency plans for a given area, the decision making during the
crisis
mitigation phase, as well the post crisis assessments. Therefore,
it has the
potential to support the activities of involved parties,
including:
• Civil Protection Authorities in charge of management and
mitigation of both
natural hazards and technological accidents.
• Local or Regional Administrations responsible for the control and
audit of
industrial installations such as refineries, power stations,
natural gas
industries etc.
• Environmental Agencies.
Finally, it may be used to assist the activities of researchers
(e.g.
epidemiologists, environmental scientists) and agencies responsible
for
following up the effects of industrial accidents on the population
and the
environment.
Acknowledgements
The author is grateful to Dr. P. Pastacos (Scientific Director of
the Infocarta
Ltd., Science & Technology Park of Crete), for the supply of
road network and
population distribution vectors as well as for his support for the
GIS
integration. The author is also grateful to Assist. Prof. C.
Cartalis (Director of
the Remote Sensing Laboratory, University of Athens, Dept. of
Applied
Physics) for the supply of Landsat/TM and NOAA/AVHRR images used in
this
study, as well as for his valuable comments and suggestions.
References
ASIMAKOPOULOS, D., 1994, Modelling the effects of obstacles on
the
dispersion of denser than air gases. Journal of Hazardous
Materials, 37,
327 – 352.
ATKINSON, G. T., and JAGGER, S. F., 1996, Plume seeding mechanisms
in
warehouse fires. Industrial Fires II Workshop Proceedings,
Cadarache,
France, 17-18 May 1994, EUR 15967 EN.
BANDINELLI, G., and CARLÁ, R., 1993, Fire Detection by Analysis
of
Infrared Satellite Imagery. In Advanced Infrared Technology
and
Applications, edited by G. M. GARLOMAGNO and C. CORSI
(Firenze:
Instituto di Ricerca sulle Onde Elettromagnetiche (I.R.O.E.) -
C.N.R.), pp.
415-428.
BARTELDS, H., SMIT, E., and MOLAG, M., 1993, Emissions from
medium
scale fires of chemicals. Industrial Fires Workshop
Proceedings,
Apeldoorn, The Netherlands, 11-12 March 1993, EUR 15340 EN.
DAVIE, F. M., and NOLAN, P. F., 1993, Combustion of chemical
substances
and the impact on the environment of the fire products.
Particle
characterisation. Industrial Fires Workshop Proceedings, Apeldoorn,
The
Netherlands, 11-12 March 1993, EUR 15340 EN.
DOZIER, J., 1981, A method of Satellite Identification of Surface
Temperature
fields of subpixel resolution. Remote Sensing of Environment, 11,
221 -
229.
CHRYSOULAKIS, N., and CARTALIS, C., 2000, A new approach for
the
detection of industrial accidents, using NOAA/AVHRR imagery.
International Journal of Remote Sensing, 21, 1743 - 1748.
CHRYSOULAKIS, N., and CARTALIS, C., 2002a, TAD - a new
Satellite
Images Analysis Software Tool for the Detection of Major Fires
caused by
Technological Accidents. International Journal of Remote Sensing
(in
press).
CHRYSOULAKIS, N., and CARTALIS, C., 2002b, A new algorithm for
the
detection of plumes caused by industrial accidents, on the basis
of
NOAA/AVHRR imagery. International Journal of Remote Sensing
(in
press).
CLEAVER, R. P., COOPER, M.G., and HALFORD, A. R., 1995,
Further
model development of a model for dense gas dispersion over real
terrain.
Journal of Hazardous Materials, 40, 85 – 108.
COZZANI, V., AMENDOLA, A., and ZANELLI, S., 1996, Hazardous
substances formed as a consequence of industrial fires. Industrial
Fires III
Workshop Proceedings, Riso, Denmark, 17-18 September 1996,
EUR
17477 EN.
GRANT., G., and DRYSDALE, D., 1994, Flame spread modelling and
fire
growth in warehouses. Industrial Fires II Workshop
Proceedings,
Cadarache, France, 17-18 May 1994, EUR 15967 EN.
HANNA, S. R., CHANG, J., and STRIMAITIS D. G., 1993, Hazardous
gas
model evaluation with field observations. Atmospheric Environment,
27A,
2265 – 2285.
KHAN, F. I. and ABBASI, S. A., 1999, MAXCRED - a new software
package
for rapid risk assessment in chemical process industries.
Environmental
Modelling & Software, 14, 11 - 25.
KUKKONEN, J., KULMALA, M., NIKMO, J., VESALA, T., WEBBER, D.
M., and WREN, T., 1994, the homogeneous equilibrium approximation
in
models of aerosol cloud dispersion. Atmospheric Environment, 28,
2763 –
2776.
LANG, D., 1993, Development of a methodology for the prediction of
the
formation of toxic species in warehouse fires. Industrial Fires
Workshop
Proceedings, Apeldoorn, The Netherlands, 11-12 March 1993,
EUR
15340 EN.
MARLIERE F., 1996, Acute toxicity of smoke: Study of gaseous
interactions.
Industrial Fires III Workshop Proceedings, Riso, Denmark,
17-18
September 1996, EUR 17477 EN.
MARTINS, J. M., and BORREGO, C., 1994, Simulation of the
atmospheric
dispersion of gases and aerosols. Industrial Fires II
Workshop
Proceedings, Cadarache, France, 17-18 May 1994, EUR 15967 EN.
MARTINS, J. M., BORREGO, C., HOLDIN, A., and MEJEAN, P.,
1996,
Comparison of the real scale, wind tunnel and numerical model
results of
atmospheric dispersion in the vicinity of a building. Industrial
Fires III
Workshop Proceedings, Riso, Denmark, 17-18 September 1996,
EUR
17477 EN.
MATSON, M., and DOZIER, J., 1981, Identification of subresolution
high
temperature sources using a thermal IR sensor.
Photogrammetric
Engineering and Remote Sensing, 47, 1311 - 1318.
MATSON, M., SCHNEIDER, S. R., ALDRIDGE, B. and SATCHWELL, B.,
1984, Fire Detection Using the NOAA-Series Satellites. NOAA
Technical
Report NESDIS 7.
MATSON, M., STEPHENS, G., and ROBINSON, J., 1987, Fire
detection
using data from the NOAA - N satellites. International Journal of
Remote
Sensing, 8, 961 - 970.
MILES, S. D. and COX, G., 1994, Transient computational fluid
dynamics
model to predict fire product emissions from warehouses. Industrial
Fires
II Workshop Proceedings, Cadarache, France, 17-18 May 1994,
EUR
15967 EN.
MUIRHEAD, K., and CRACKNELL, A. P., 1984, Identification of gas
flares
in the North Sea using Satellite data. International Journal of
Remote
Sensing, 5, 199 - 212.
MUIRHEAD, K., and CRACKNELL, A. P., 1985, Straw burning over
Great
Britain detected by AVHRR. International Journal of Remote Sensing,
6,
827 - 833.
PORTER, S. R., HALL, D. J., and CARRUTHERS, D. J., 1996, Physical
and
computer modelling of dispersion of toxic smoke from warehouse
fires.
Industrial Fires III Workshop Proceedings, Riso, Denmark,
17-18
September 1996, EUR 17477 EN.
TECHNICA, 1992, WHAZAN: A Software for Hazard Assessment.
Technica,
London.
TNO, 1991, EFFECTS: A software for Hazard Assessment. TNO
Prins
Mauritis Research Laboratory, The Netherlands.
TNO, 1992, SAVE: A software for Hazard Assessment. TNO Prins
Mauritis
Research Laboratory, The Netherlands.
VVT, 1993, RISKIT: A Software for Risk Assessment. The Technical
Research
Centre of Finland, Tempere, Finland.
WEBBER, D. M., JONES, S. J., MARTIN, D., 1993, A model of the
motion of
a heavy gas cloud released on a uniform slope. Journal of
Hazardous
Materials, 33, 101 – 122.
Captions
Table 1. Potential of Earth Observation to support the requirements
of an
emergency plan and examples of satellites which can be used.
Figure 1. NOAA/AVHRR channel 2 image of May 13, 2000 (14.44
UTC)
calibrated and geometrically corrected. The country borders have
been also
superimposed. The white arrow shows the position of the plume
caused by the
massive explosion in a firework factory in the town of
Enschede.
Figure 2. Architecture of the GIS based technological risk
management support
tool.
Figure 3. Pseudochannel image of brightness temperature difference
between
NOAA/AVHRR channels 3 and 4 (May 13, 2000; 14.44 UTC) adjusted to
the
broader area of Athens. Bright tones are indicative for pixels with
maximum
brightness temperature difference values. These pixels, which are
located over
Thriasio Pedio area, correspond to the scenario accident’s
site.
Figure 4. NOAA/AVHRR channel 2 image depicting the position of the
plume
at 14.44 UTC, according to the technological accident
scenario.
Figure 5. NOAA/AVHRR channel 2 image depicting the position of the
plume
at 17.20 UTC, according to the technological accident scenario. The
plume in
this case has been diffused over the city of Athens.
Figure 6. Landsat/TM pseudocoloured composition 3-2-1 (April 26,
1994). The
possible areas of technological accident occurrence have been
superimposed
(areas in red colour). The position of the scenario accident’s site
(derived from
AVHRR imagery) has been also superimposed (area within the
yellow
polygon).
Figure 7. Landsat/TM pseudocoloured composition 3-2-1 (April 26,
1994).
Physical and artificial characteristics of the area located around
the accident’s
site are presented. The road network has been sumperimposed. The
position of
the accident’s site (derived from AVHRR imagery) has been also
superimposed
(yellow hatched area).
Figure 8. Integration of Landsat, NOAA, CORINE and in-situ data for
the
monitoring of urban areas (TM pseudocoloured composition 3-2-1); of
the
spatial distribution of vegetation (in green tones); of the
possible areas of
occurrence (in red colour); of the position of the accident’s site
(yellow
polygon); of the position of the produced plume at 14.44 UTC
(smaller black
polygon) and 17.20 UTC (greater black polygon).
Figure 9. Integration of NOAA and in-situ data for the estimation
of population
at risk during the propagation of the toxic plume over the city of
Athens. The
spatial distribution of population between 5 and 10 Km (blue
circles) from the
accident’s site is monitored. Inhabited areas are presented in
different colours
according to their population density. The positions of the plume
at 14.44 UTC
(green polygon) and 17.20 UTC (red polygon) are also
presented.
Requirement for an Emergency Plan
Potential of Earth Observation to support requirements
Satellites
Assessment of land use High Landsat, SPOT, IRS, JERS, IKONOS
Assessment of land cover High Landsat , SPOT, NOAA, IRS, JERS
Depiction of urban areas and major installations High Landsat ,
SPOT, IRS, JERS, IKONOS Depiction of road network High Landsat,
SPOT, IRS, JERS, IKONOS Detection of the fire High NOAA Detection
of the plume High Landsat, SPOT, NOAA, IRS, JERS Definition of the
size of the plume High Landsat, SPOT, NOAA, IRS, JERS Definition of
speed and direction of propagation of the plume High NOAA
Definition of the chemical composition of the plume Not possible -
Definition of the opacity of the plume High Landsat, SPOT, IRS,
JERS, NOAA
Co-ordination among involved parties Indirect (through the
production of EO based thematic maps)
Landsat, SPOT, NOAA, IRS, JERS, IKONOS
Decision making regarding the evacuation of urban areas to avoid
plumes
High
Decisions regarding long-term follow-up of the effects (e.g.
monitoring, health protection)
High (e.g. by integrating above information)
Landsat, SPOT, NOAA, IRS, JERS, IKONOS