RISICO a system for wide-nation wildfire risk assessment and

management

Paolo Fiorucci, Francesco Gaetani, Riccardo Minciardi

A general architecture for wildfire risk management

Static hazard assessment procedure

Dynamic hazard

assessment procedure Active fire risk

management

Fire detection and fire tracking

procedures

Real-time resource allocation procedure

Preoperational resources management procedures

Urban and forest planning

Direct observation

Climate data

Topographic and territorial

data

Historical forest fires database

Meteo forecast

Real-time weather conditions

Other sensors data

Exposed elements data

delimitation and measurement of the

burned areas

Definition of administrative

measures

In field measurement

Earth Obs. data

Dynamic hazard assessment

The main drivers of wildfire process are:

➲ Vegetation;➲ Meteorological conditions;➲ Topography;

Defining suitable semi-empirical models it is possible to define an estimation of the potential behaviour of a fire eventually ignited in a certain space time window.

Dynamic hazard assessment

Discretizing the considered area it is possible to define for each grid cell:

➲ Vegetation type (dead and live fuel load, live fuel moisture, Higher Heating Value);

➲ Meteorological condition and/or forecast (Air temperature, relative humidity, rainfall, wind speed and direction);

➲ Elevation➲ Slope➲ Aspect

RISICO: a system for dynamic fire risk assessment

RH

Slopeequivalentwind speed

Windspeed

Winddirection

Aspect Slope

Slope

FirelineIntensity

Slope effectson ROS

No windinitial speedon flat terrain

Initial Rateof Spread

LHVlive fuel

LHVdeadfuel

Dead fine Fuelmoisture model

Potential fire spread model

EquilibriumMoisture Contents

RainfallSnowfall

Leaf Growth Model

PhenologicalModel

AirTemperature

Soil MoistureConditions

Fine live

fuel load

X

BiomassMoisture Model

Live fi n

e fuel m

ois tu

re a nd

l oa d

mo

dels

Airtemperature

Duff Moisture Conditions

Windspeed

Airtemperature

Fine deadfuel Load

Biomass LoadModel

Drying/WettingRate

Fine livemoisture

conditions

SoilMoisture

Conditions

Dead Fine Fuel Moisture Conditions

The The meteorological information informationMETEO FORECASTED DATAThe system receives daily the outputs of a meteorological Limited Area Model (LAM), namely COSMO LAMI consisting of a set of data discretized in time steps of three hours, over a time horizon of 72 hours, defined over a regular grid composed by 57200 cells having a side corresponding to 0.05 degrees

k (h) air temperature [K]

k (h) dew point temperature [K]

pk (h) cumulate rainfall (th – th-1 ) [m]

wk (h) wind speed [m s-1]

k (h) wind direction [rad]

METEO OBSERVED DATAEach run of the system is fed by new fresh data obtained from the extended national Meteorological Observation Network. Information relevant to 24 h cumulate precipitation and temperature observed by almost 3000 meteo stations is interpolated to obtain the fields defining the zero state of the daily run.

The land use and vegetation data

Source data: CORINE LAND COVER release 2000 (CLC 2000).16 categories of fuel has been considered and parametrized.

• average seasonal fuel loads [kg m-2];

• average seasonal moisture contents [%] only for live fuels;

• average seasonal Higher Heating Value (HHV) [kJ kg-1].

L0

L1

L2

Live fuels

Dead fine fuels

SNOW COVER map - MODIS acq. in 2001-01-01

NDVI map – MODIS acquired in 2003-07-18

EO products

Remote Sensing data provide valuable information for the characterization of the state of vegetation, mapping of fuel types and vegetation properties at different temporal and spatial scales including the global, regional and landscape levels.

RISICO graphical user interface

RISICO graphical user interface

Towards GRID architecture

➲ Only a high resolution grid allows to represent the vegetation and topographical heterogeneity;

➲ When several fires are active simultaneously civil protection managers need to know which fire determine the highest risk for the people;

➲ GRID architecture allows to introduce new perspective in potential risk assessment by means of propagation models able to give in output the potential burnt area in a given time interval and the potential damage considering the exposed elements.