EUFIRELAB

EVR1-CT-2002-40028

D-03-06

http://www.eufirelab.org

EUFIRELAB:
Euro-Mediterranean Wildland Fire Laboratory,
a “wall-less” Laboratory
for Wildland Fire Sciences and Technologies
in the Euro-Mediterranean Region

Deliverable D-03-06

Behaviour models of wildland fire:

third versions of the codes

Collective work moderated by Josep PIÑOL

May 2006

The views expressed are purely those of the writers and may not, in any circumstances, be regarded as stating an official
position of the European Commission
EUFIRELAB

CONTENT LIST

1. SimpleFire.................................................................................................................................................................1
2. FireRegime 1.0 .........................................................................................................................................................3
3. FocGest 1.0 ..............................................................................................................................................................5
4. Fire Rotation Model (FRM ........................................................................................................................................7
5. FireAnalytic 1.1 .........................................................................................................................................................9
6. TECNOMA-FSE (Fire Spread Engine) ...................................................................................................................11
7. AirFire .....................................................................................................................................................................13
8. DisperFireStation ....................................................................................................................................................16

d-03-06.doc
EUFIRELAB

SUMMARY
This document is the update of Deliverable D-03-04 and describes eight fire behaviour models.
The main model characteristics summarized are:
- The main physical models included in the code;
- The numerical resolution methods and physical models;
- The physical inputs;
- The physical outputs;
- The development tools and platforms;
- The persons to contact;
- A simulation example.

d-03-06.doc
EUFIRELAB

1 SimpleFire.

1.1 WILDFIRE MODELLING CLASSIFICATION

Cellular automata model of fire regime

1.2 MAIN PHYSICAL MODELS INCLUDED IN THE CODE

Vegetation dynamics is described by the simplest possible model: a pixel can be occupied by vegetation or not.
Transition from non-vegetated to vegetated is probabilistic; the opposite transition only occurs by burning
Fire spread: probabilistic, influenced by the existence or absence of fuel in a cell and by meteorology
Fire management: fire suppression and prescribed fire
Spatial resolution: 1 ha
Time resolution: 1 year

1.3 NUMERICAL RESOLUTION METHODS & PHYSICAL MODELS

It is a model mainly used for teaching and demonstration purposes. It lacks parameter optimization routines as
the more complex models FIREREGIME and FOCGEST.

1.4 PHYSICAL INPUTS

None. The model is conceptual.

1.5 PHYSICAL OUTPUTS

Number of fires per year (total and per fire size class)
Area burnt per year (total and per fire size class)
Area burnt per year in prescribed fires

1.6 DEVELOPMENT TOOLS AND PLATFORMS

Java
Windows, MacOs, Linux
Freely available at http://www.EcologiaConNumeros.uab.es . SimpleFire is the applet 1.1 of the collection of 24
Java applets used in the teaching of Ecology at Univ. Autònoma of Barcelona and included in the book Ecología
con Números.

1.7 CONTACT
PINOL Josep
CREAF
Universitat Autònoma de Barcelona
08193 Bellaterra
Spain
Email : Josep.Pinol@uab.es

1.8 REFERENCES
PIÑOL, J. & MARTINEZ-VILALTA, J.M. (2006) – Ecología con Números. Una introducción a la Ecología con
Problemas y Ejercicios de Simulación. Ed. Lynx, Barcelona. 440 pp.

d-03-06.doc 1
EUFIRELAB

Figure 1-1: Example of simulation: The area covers ca. 40,000 ha of flat terrain. Green (yellow) colour indicates
vegetated (unvegetated) pixels; red colour indicates fires produced in the last year. The output is in English,
Spanish or Catalan depending on the computer configuration.

d-03-06.doc 2
EUFIRELAB

2 FireRegime-1.0.

2.1 WILDFIRE CLASSIFICATION MODEL

Cellular automata model of fire regime

2.2 MAIN PHYSICAL MODELS INCLUDED IN THE CODE

Fire spread: probabilistic, influenced by the fuel load of the cell and meteorology
Vegetation dynamics: deterministic linear increase of fuel load until a maximum is reached
Fire management: fire suppression and prescribed fire
Spatial resolution: 1 ha
Time resolution: 1 year

2.3 NUMERICAL RESOLUTION METHODS & PHYSICAL MODELS

Montecarlo search to find behavioural parameter sets, that is, those who conform with fire regime data (n fires
per year, annual area burnt, and distribution of number of fires and area burnt among fire size classes)

2.4 PHYSICAL INPUTS

None. The model is conceptual. Nevertheless inputs reproduced likely scenarios for meteorology and vegetation
dynamics in Mediterranean regions

2.5 PHYSICAL OUTPUTS

Number of fires per year (total and per fire size class)
Area burnt per year (total and per fire size class)
Area burnt per year in prescribed fires

2.6 DEVELOPMENT TOOLS AND PLATFORMS

Java
Windows, MacOs, Linux

2.7 CONTACT
PINOL Josep
CREAF
Universitat Autònoma de Barcelona
08193 Bellaterra
Spain
Email : Josep.Pinol@uab.es

2.8 REFERENCES
PIÑOL, J., BEVEN, K.J. & VIEGAS, D.X. (2005) – Modelling the effect of fire-exclusion and prescribed fire on
wildfire size in Mediterranean ecosystems. Ecological Modelling 183: 397-409

d-03-06.doc 3
EUFIRELAB

Figure 2-1: Example of simulation: The area covers ca. 100,000 ha of flat terrain.
Colours indicate the age of vegetation (and its fuel load); red colour indicates fires produced in the last year

d-03-06.doc 4
EUFIRELAB

3 FocGest 1.0

3.1 WILDFIRE MODELLING CLASSIFICATION

Cellular automata model of fire regime on a realistic terrain.

3.2 MAIN PHYSICAL MODELS INCLUDED IN THE CODE

Fire spread: probabilistic, influenced by the fuel load of the burning surrounding cells, meteorology and slope.
The model distinguishes surface- and canopy- fires and it includes spotting.
Vegetation dynamics: two different layers (understory, trees), deterministic sigmoidal increase of fuel load until a
maximum is reached for each of the layers
Fire management: prescribed fire and mechanical outtake of fuel (manual & automatised)
Spatial resolution: 1/3 ha
Time resolution: 1 year

3.3 NUMERICAL RESOLUTION METHODS & PHYSICAL MODELS

Monte Carlo search to find behavioural parameter sets, that is, those who conform with fire regime data (n fires
per year, annual area burnt, and distribution of number of fires and area burnt among fire size classes)

3.4 PHYSICAL INPUTS

Topography of the terrain.
Land use

3.5 PHYSICAL OUTPUTS

Number of fires per year (total and per fire size class)
Area burnt per year (total and per fire size class)
Area burnt per year in prescribed fires

3.6 DEVELOPMENT TOOLS AND PLATFORMS

Java
Windows, MacOs, Linux

3.7 CONTACT
LOEPFE Lasse
CREAF
Universitat Autònoma de Barcelona
08193 Bellaterra
Spain
Email : putolasse@gmx.net

d-03-06.doc 5
EUFIRELAB

Figure 3-1: Example of simulation: The area covers ca. 90,000 ha of terrain.
Colours indicate the different land uses and recent fires (surface fires in yellow and canopy fires in red).
The area is Tarragona (NE Spain)

d-03-06.doc 6
EUFIRELAB

4 FIRE LINE ROTATION MODEL (FRM)

4.1 WILDFIRE MODELLING CLASSIFICATION

Semi-empirical.

4.2 MAIN PHYSICAL MODELS

Heat transfer due to convection; feedback effect on the chemical reaction.

4.3 NUMERICAL METHODS AND PHYSICAL MODELS

The evolution of the fire line shape is determined accordingly to:
- VIEGAS, D.X., “Fire line rotation as a mechanism for fire spread on a uniform slope”, International Journal of
Wildland Fire. Vol.11 Nº1 (2002), p.11-23.

The rate of spread grows with time (dynamic behavior) and is determined accordingly to:
- VIEGAS, D.X, “A mathematical model for forest fires blow-up”, International Journal of Wildland Fire. Vol.13
Nº1 (2004), p.27-51.

4.4 PHYSICAL INPUTS

Terrain slope or reference wind flow velocity; basic rate of spread of the fuel bed, experimentally determined
parameters.

4.5 PHYSICAL OUTPUTS

Evolution of a fire line in a slope originated by a point ignition.

4.6 DEVELOPMENT TOOLS AND PLATFORMS

Written in Visual Basic under a Microsof Office Excel® calculus worksheet. Runs in a personal Computer
(INTEL, AMD processors).

4.7 CONTACT
Domingos Xavier VIEGAS
ADAI – Associação para o Desenvolvimento da Aerodinâmica Industrial
CEIF – Centro de Estudos Sobre Incêndios Florestais
Rua Pedro Hispano, Nº12
Apartado 10131
3031-601 Coimbra
Telf: 00351 239 708 580
Fax: 00351 239 708 589
E-mail: xavier.viegas@dem.uc.pt

d-03-06.doc 7
EUFIRELAB

Figure 4-1: Contours of the fire front shape evolution with 30s intervals (distances in m)
for a Pinus pinaster dead needles litter, fuel load of 1kg/m2 and 30º slope

d-03-06.doc 8
EUFIRELAB

5 FIREANALYTIC 1.1

5.1 WILDFIRE MODELLING CLASSIFICATION

A 3D physical real-time model of Wildland fire

5.2 MAIN PHYSICAL MODELS INCLUDED IN THE CODE

Energy equation for the fuel layer
Radiant and convective heat transfers
Gas flow in the fuel layer

5.3 NUMERICAL RESOLUTION METHODS

3D description
Laboratory and field scale
Analytic equations: no discretization needed

5.4 PHYSICAL INPUTS

Spatial distribution of vegetation, moisture content
External flow conditions (wind speed before fire ignition, ambient air temperature …)
Slope
Fuel load

5.5 PHYSICAL OUTPUTS

Temperature of the fuel layer
Rate Of Spread (ROS) evaluated from the time evolution of the position of the pyrolysis front,
Fire front geometry (height, tilt angle, width…) obtained by physical considerations (in Fireanalytic 1.0, they
were determined from empirical laws as Mc Caffrey correlations).

5.6 DEVELOPMENT TOOLS AND PLATFORMS

Mathematica Package: “Statistics NonlinearFit”

5.7 CONTACT
BALBI Jacques-Henri ROSSI Jean-Louis
UMR CNRS 6134 – SPE UMR CNRS 6134 - SPE
Campus Grossetti Campus Grossetti
BP 52 BP 52
20250 CORTE 20250 CORTE
Email: balbi@univ-corse.fr Email: rossi@univ-corse.fr

MARCELLI Thierry
UMR CNRS 6134 – SPE
Campus Grossetti
BP 52
20250 CORTE
Email: marcelli@univ-corse.fr

d-03-06.doc 9
EUFIRELAB

Figure 5-1: Predicted and observed fire front shapes over time for 30° up-slope fire spread across a 0.3 kg/m2 load
of Pinus pinaster needles with no wind. Slope gradient is shown in the figure.

Fig. 5-2. Predicted and observed (in Viegas

D.X, Int. J. of Wildland Fire, 2004, 13:143-156)
fire front shapes over time for 30° up-slope fire
spread across a 1 kg/m2 load of Pinus pinaster
needles and for a 5 m/s wind velocity. Wind
and slope gradient are shown in the figure.

d-03-06.doc 10
EUFIRELAB

6. TECNOMA-FSE (FIRE SPREAD ENGINE)

6.1 WILDFIRE MODELLING CLASSIFICATION

Surface Forest Fire propagation – semi empirical approach.

6.2 MAIN PHYSICAL MODELS INCLUDED IN THE CODE

Fire rate of spread: based on Rothermel’s surface fire spread (ROTHERMEL, 1972).
Fire shape: Alternatively using the simple ellipse model (ALEXANDER, 1985) or the double ellipse (ANDERSON,
1983), selected by user in the command line
Fire growth: raster approximation. Fire growth is a process of contagion between burning and non-burning cells.
In this geometry, errors associated to grid taxicab metrics are considered and solved.

6.3 NUMERICAL RESOLUTION METHODS AND PHYSICAL MODELS

This fire spread simulation application makes use of square-cell grid maps .
The local fire spread law is obtained in each cell, after which the cellular automata calculates the progression of
fire.
Initially the resolution is not limited, but the model is suggesting to use spatial resolution of 20 m. for forest fuels,
100 m. for slope and aspect and 500 m. for wind direction and speed, to obtain best results.
The temporal resolution is of 1 minute.
Total spatial extension of the analysed area is unlimited, but it is suggested to cover 10x10 km, and simulation
time less than 3 hours of projection, although longer periods are allowed.
TECNOMA-FSE application looks into the quasi-erratic behaviour of the wind vector direction (WDIR) by
applying a wind vector aperture (AA) expressed degrees to the existing vector in each cell.
This aperture defines a minimum (WDIR-AAº) and a maximum (WDIR +AA) value between which the wind
vector direction oscillates randomly, according to Pasquill table of atmosphere stability
Some of the most relevant problems associated to taxicab geometry present in grid-based automatas (Krause,
1987) are addressed and solved, which were first implemented in the previous versions of the application, i.e.
CARDIN system, (Martínez-Millán et Al., 1991), FOMFIS (Caballero et Al., 1999) and E-FIS (Caballero et Al.,
2001).
This simulation engine makes use of the concept of pseudo-starting points (focoids) in the borders of fire spread
conditions change, applying Huygens-principle in such points to generate a common propagation wave.

6.4 PHYSICAL INPUTS

Inputs are provided in ASCII-GRASS raster format in separate files: terrain slope in percentage, terrain aspect
in degrees (downhill), fuel models, wind speed in km/h, wind direction in degrees, live fuel moisture content in
percentage, 1HR, 10HR and 100HR dead fuel moisture content in percentage, list of starting points (number and
then X, Y and appearance time in minutes), forest fuel models parameters, the same as described and used in
BEHAVE system (fuel load, surface to volume ratio, fuel depth, energy release, fuel extinction moisture etc.)

6.5 PHYSICAL OUTPUTS

All raster maps in ASCII-GRASS format, output provided in separate files: time (in minutes) invested by fire in
accessing each cell from the starting point, fire linear intensity, fire maximum propagation rate, fire maximum
propagation direction, fire heat per unit area.

6.6 DEVELOPMENT TOOLS AND PLATFORMS

Developed in Real Basic 5.0 and compiled for Linux (bin) and Windows (exe) environments as stand-alone
executable in the command line with modifiers.
Required maps are generated in ASCII format in the GRASS environment.
Currently implemented in a proprietary on-line map server (SIGYM2) based on GRASS at operational level in
Spain, through the private firm Meteologica Ltd.

d-03-06.doc 11
EUFIRELAB

6.7 CONTACT
David Caballero
TECNOMA S.A. Department of Forest Ecosystems
Isla Graciosa 1, 28700 S. Sebastián de los Reyes, Madrid, Spain
Phone: +34 91 6586636
FAX: +34 91 6519490
email: davidcaballero@tecnoma.es
url: http://www.tecnoma.es
http://www.davidcaballero.com

6.8 REFERENCES
CABALLERO, D. VIEGAS, D.X. XANTHOPOULOS, G. (2001) "E-FIS: an electronic on-line decision support system for
forest fires". In proc. Of the International Workshop on Improving dispatching for forest fire control.
Mediterranean Agronomic Institute of Chania, Crete, Greece. December 6-8
CABALLERO, D. XANTHOPOULOS, G. KALLIDROMITOU, D. LYRINTZIS, G. BONAZOUNTAS, M. PAPACHRISTOU, P. PACIOS,
O. (1999). "FOMFIS: Forest Fire Management and Fire Prevention System". In proc. of DELFI Intl. Symp.
Forest Fires Needs and Innovations. Pp. 93-98. Athens, Greece, 18-19 November.
CABALLERO, D. MARTINEZ-MILLAN, F.J. MARTOS, J. VIGNOTE, S. (1994). "CARDIN 3.0, A Model for Forest Fire
Spread and Fire Fighting Simulation". Vol.1: p.501. In proc. of 2nd Intl. Conf. on forest Fire Research. Coimbra,
Portugal.
KRAUSE, E. F. (1987). “Taxicab Geometry: An Adventure in Non-Euclidean Geometry”. Dover Publications,
February.
MARTINEZ-MILLAN, F.J. MARTOS, J. VIGNOTE, S. CABALLERO, D. (1991). "CARDIN, Un Sistema para la Simulación de
la Propagación de Incendios Forestales" MAPA-INIA. Investigación Agraria Serie Recursos Naturales Vol. 0
December
Figure 6-1: Fire front evolution

d-03-06.doc 12
EUFIRELAB

7 AIRFIRE

7.1 WILDFIRE MODELLING CLASSIFICATION

Smoke dispersion model - Physical Approach

7.2 MAIN PHYSICAL MODELS INCLUDED IN THE CODE

7.2.1 MEMO meteorological model
Three dimensional Eulerian non-hydrostatic prognostic mesoscale model;
Numerical solution of the conservation equations for mass, momentum and transport for scalar quantities
(potential temperature, turbulent kinetic energy and specific humidity) in terrain-following coordinates;
Nesting facility, one-way interactive nesting scheme is implemented
Modified sub-routine for energy fluxes and atmospheric heating rate allows the introduction of the fire as a heat
source.
As a recent development in AIRIFIRE modelling system a PM10 dispersion and deposition module was included
in MEMO model.
Surface deposition of particles occurs via diffusion, impactation, and/or gravitational settling.
Particle size is the dominant variable controlling these processes.
The approach suggested by VENKATRAM (VENKATRAM and PLEIM, 1999), has been adopted in MEMO. Particle
deposition velocity for a given aerosol size is calculated using the following equation:

vs
vd =
1 − exp(− rv s )

where vs is the gravitational settling (or sedimentation velocity) and r the resistance to transport.
7.2.2 Fire model
Based on Rothermel model (spread model of low intensity surface forest fires)
Fire growth simulation: deterministic model based on Huygen’s principle, which uses an elliptical spread at each
point of the fire front.
7.2.3 MARS photochemical model
Three dimensional Eulerian model;
Numerically simulates photo-oxidants formation considering the chemical formation process and its transport in
the atmospheric boundary layer;
Solves the differential concentration transport equation system in terrain following co-ordinates, with the
meteorological variables calculated by MEMO;
Fire emissions model integrated.

7.3 NUMERICAL RESOLUTION METHODS AND PHYSICAL MODELS

7.3.1 MEMO meteorological model
Turbulence: K-theory (0-, 1- or 2-equation turbulence model);
Radiative heating/cooling rate and radiative fluxes : based on emissivity method for long wave radiation and an
implicit multilayer method for short wave radiation.
7.3.2 MARS photochemical model
Estimation of fire emissions according to Ward and Radke;
Two chemical mechanisms: EMEP (66 species, 139 photochemical reactions) and KOREM (20 species, 39
photochemical reactions).

d-03-06.doc 13
EUFIRELAB

Figure 7-1: Structure of the model

ROTHERMEL
EMISSIONS PROGRESSION
MARS
INTERACTION
FIRE / ATMOSPHERIC FLOW

ATMOSPHERIC FLOW DISPERSION+CHEMISTRY

MEMO MARS

AIR QUALITY

7.4 PHYSICAL INPUTS

7.4.1 MEMO meteorological model
Topography and surface type for each grid location;
Meteorological data: initial state (surface measuring and upper air soundings) and time-dependant boundary
conditions.
7.4.2 Fire model
Mid-flame wind speed (given by MEMO)
Type of fuel and fuel moisture.
7.4.3 MARS photochemical model
Topography and surface type for each grid location;
Meteorological data: wind speed and direction, turbulent kinetic energy, Monin-Obukhov length and friction
velocity (given by MEMO);
Emission data (VOC, NOx and CO): forest fire and a diurnal cycle of all emitters (traffic, industry, biogenics).
Initial and boundary conditions.

7.5 PHYSICAL OUTPUTS

7.5.1 MEMO meteorological model
Three dimensional fields: wind speed and direction, turbulent kinetic energy, Monin-Obukhov length and friction
velocity.
7.5.2 Fire model
Spread rate in the maximum spread direction, fire size and shape.
7.5.3 MARS photochemical model
Three dimensional concentration fields: O3, NOx and CO.

7.6 DEVELOPMENT TOOLS AND PLATFORMS

UNIX

7.7 CONTACT
MIRANDA Ana Isabel
Department of Environment and Planning
University of Aveiro
3810-193 Aveiro, Portugal
Email : aicm@dao.ua.pt

d-03-06.doc 14
EUFIRELAB

7.8 EXAMPLE OF SIMULATION

Figure 7-2: Horizontal ground pattern of wind and CO concentrations

11000

10000

9000
South/North (m)

8000

7000

6000

5000

6000 7000 8000 9000 10000 11000 12000 13000 14000 15000
West/East (m)
g/m3 CO 5 m/s
0 10000 20000 30000 40000

Figure 7-3: Horizontal ground pattern of wind and PM10 concentrations

0 0 0
0 5 0 0 0 0 0 0 0 0
5 2 0 0 0 0 0 0 0 0 0 0 0
1 1 1 8 6 5 4 3 2 1 8 6 3 0
PM10
(μ g.m-3) 4m.s-1
200000

180000

Santarem
160000

140000

120000

Lisboa
100000

80000 Setúbal Évora

60000
Oc
ea
no
40000 At
la
nt
ic
o
20000

0
0 20000 40000 60000 80000 100000 120000 140000 160000 180000 200000

7.9 REFERENCES
VENKATRAM, A. and PLEIM, J. (1999) The electrical analogy does not apply to modelling dry deposition of particles,
Atmospheric Environment, 33, pp. 3075-3076.

d-03-06.doc 15
EUFIRELAB

8 DISPERFIRESTATION
(This model resulted from the integration of DISPERFIRE in Fire Station – both described in D-03-02)

8.1 WILDFIRE MODELLING CLASSIFICATION

Semi Empirical Approach

8.2 MAIN PHYSICAL MODELS INCLUDED IN THE CODE

8.2.1 Modified NUATMOS meteorological model
Three dimensional wind field diagnostic model (ROSS, 1988) with the calculation of effective viscosity (MIRANDA,
1998).
8.2.2 Fire rate of spread
Based on Rothermel’s surface fire spread (ROTHERMEL, 1972).
8.2.3 Fire shape
Simple ellipse model (ALEXANDER, 1985) or double ellipse (ANDERSON, 1983) if wind speed at mid flame is lower
than 0.2 m/s.
8.2.4 Fire growth
Raster approximation.
Fire growth is a process of contagion between burning and non-burning cells.
The algorithm followed is based on Dijkstra’s dynamic programming algorithm.
8.2.5 Smoke dispersion model
Three dimensional Lagrangian stochastic model.
8.2.6 Visibility module
Visibility estimated using air pollutants concentration on air.
8.2.7 Fire Weather Index
The system allows to (1) have a broad assessment of large-scale fire potential through the evaluation of the
daily and spatial variation of the fire danger index and (2) estimate the moisture content of dead and live fine fuels
through empirical relationships.

8.3 NUMERICAL RESOLUTION METHODS & PHYSICAL MODELS

8.3.1 NUATMOS meteorological model
Interpolation of meteorological observations throughout the domain satisfying conservation equation for mass, in
terrain-following coordinates and variable vertical grid spacing.
8.3.2 Fire models
Fire propagation and wind simulation are limited to the resolution (cell size) of the input raster maps (DTM and
vegetation).
8.3.3 Smoke dispersion model
3D Lagrangian approach.
The dispersion of the pollutant is simulated by markers whose displacement reproduces the statistics of the
considered turbulent transport.
The dispersion grid is in terrain-following coordinates and variable vertical grid spacing.
Estimation of fire emissions according to MIRANDA (1998);
Plume rise calculation according to BRIGGS (1969) formulas.

d-03-06.doc 16
EUFIRELAB

8.4 PHYSICAL INPUTS

8.4.1 NUATMOS meteorological model
Topography for the simulation domain; meteorological data: surface measurements and upper air soundings (if
available)
8.4.2 Fire propagation
Terrain characteristics (Digital Terrain Model); wind data (speed and direction) and fuel characteristics (physical
characteristics such as load by classes, s/v ratio, heat content, extinction moisture, fuel depth, fuel moisture)
8.4.3 Smoke dispersion model
Topography for the simulation domain;
Meteorological data: wind speed and direction and effective viscosity (given by NUATMOS);
Emission data (NOx and PM10 and PM2.5)
8.4.4 Visibility model
Extinction efficiencies.

8.5 PHYSICAL OUTPUTS

8.5.1 NUATMOS meteorological model
Three dimensional fields: wind speed and direction and effective viscosity.
8.5.2 Fire spread
Simulation: characteristics (graphical and numerical) in each cell, namely fire rate of spread, linear intensity,
flame length, energy released.
8.5.3 Smoke dispersion
Three dimensional concentration fields: NOx and PM10 and PM2.5.
8.5.4 Visibility
Two dimensional visibility fields in dV.
8.5.5 Fire Risk Mapping
FWI and its components, namely FFMC, DMC, DC, ISI and BUI.

8.6 DEVELOPMENT TOOLS AND PLATFORMS

Written in MDL, a specific C language of Microstation® (Bentley CAD software) that has built-in subroutines for
the design of window-based interfaces, generation of visualization elements in the 3D space, on top of the usual
mathematical capabilities of the C language.
The wind and smoke dispersion models are self-contained Fortran codes, which run as external programs.
Requires MicroStation® to run.

8.7 CONTACT
MIRANDA Ana Isabel
Department of Environment and Planning
University of Aveiro
3810-193 Aveiro, Portugal
email : aicm@dao.ua.pt
url: http://www.dao.ua.pt/gemac
GAMEIRO LOPES António Manuel
ADAI / Department of Mechanical Engineering
FCTUC – University of Coimbra
3030 Coimbra
Tel: 351 239 790773
Fax: 351 239 790771
email: antonio.gameiro@dem.uc.pt
url: http://www2.dem.uc.pt/antonio.gameiro/amglopes.htm

d-03-06.doc 17
EUFIRELAB

8.8 EXAMPLE OF SIMULATION

Figure 8-1: CO concentration (μg.m-3) and wind fields
obtained for the burning of a experimental fire during GESTOSA 2004 experiments.

8.9 REFERENCES
ALEXANDER, M.E. (1985) - "Estimating the length-to-breadth ratio of elliptical forest fire patterns", in Proceedings of
the eighth conference on fire and forest meteorology. Soc. Am. For., pp. 287-304, Bethesda, Maryland.
ANDERSON, H.E. (1983) - "Predicting Wind-Driven Fire Size and Shape", USDA-FS, Ogden UT, Research Paper
INT-305.
BRIGGS, G.A. (1969) - Plume Rise. Air Resource Atmospheric Turbulence and Diffusion. Lab., Environ. Sci. Serv.
Adm., U.S. Atomic Energy Comm., Oak Ridge, Tenn., USA. 81 p.
LOPES, A.M.G., SOUSA, A.C.M., VIEGAS, D.X. (1995) - "Numerical Simulation of Turbulent Flow and Fire
Propagation in Complex Terrain", Numerical Heat Transfer, Part A, N. 27, pp. 229-253, 1995.
MIRANDA, A.I. (1998) - Efeito dos Incêndios Florestais na Qualidade do Ar (only Portuguese version available;
English translation of the original title: “Forest Fire Effects on the Air Quality). PhD Thesis, Department of
Environment and Planning, University of Aveiro, Portugal.
ROSS, D.; SMITH, I.; MANINS, P. and FOX, D. (1988) - Diagnostic Wind Field Modelling for Complex Terrain: Model
Development and Testing. Journal of Applied Meteorology, Vol. 27. pp. 785-796.
ROSS, D.G., SMITH, I.N., MNINS, P.C. and FOX, D.G. (1988) - "Diagnostic wind field modelling for complex terrain:
Model development and testing", Journal of Applied Meteorology, Vol. 27, pp. 785-796.
ROTHERMEL, R.C. (1972) - "A Mathematical Model For Predicting Fire Spread in Wildland Fuels", USDA Forest
Service Research Paper, INT-115, Ogden UT.

d-03-06.doc 18