THE PROJECT

PREVINCAT is a web server aiming at providing the required information to plan fuel and forest management actions to reduce the risk of large forest fires in Catalonia.

METHODOLOGY

In order to generate the information and landscape maps, that can be found and downloaded from PREVINCAT, we compiled spatially explicit data on topography and vegetation characteristics. From the initial data, different variables related to potential fire behavior were estimated.

Initial data

To generate the cartography on vegetation (Module 1) we combine existing information on vegetation composition (MCSC-4, 2009; Vericat et al., 2010; Piqué et al., 2011b), a digital terrain model and vegetation variables defining vegetation structure were obtained from LiDAR related data (mapes de variables biofísiques, CREAF (2016)).
When defining the meteorological scenarios (Module 2) a total number of 15 regions were identified for all Catalonia, with at least one reference meteorological station within the regions limits. Those stations having at least 20 years of recorded data. The methodology applied to define days of high risk of fire was the same as the one used by the fire prevention service (SPIF). Such methodology aims at identifying days with a combination of meteorological variables values (high temperatures, low relative humidity and high wind speeds, based on percentile assessments) that facilitates the development of a large forest fires.

Obtained data

Based on the initial data different variables related to fire behaviour were estimated, such as: bush coverage, branch insertion height and canopy bulk density. These new variables have being estimated through published statistical models (Coll et al., 2011; Ruiz-Peinado et al., 2011 and 2012) and new allometric relations developed from the 4rd National Forest Inventory. Additionally, fuel models (Scott & Burgan 2005) were defined for the whole Catalonia by combining initial and derived data, taking also into account specific adjustments for those areas affected by fires after 2009.
From the meteorological data, recorded on each reference station, different days were risky conditions were maintained at least during 4 hours were selected. These selected days were grouped according to well-known synoptic conditions (South entrance, ponentada, etc.) and their occurrence probability estimated. For most of the 15 regions data on meteorological scenarios leading to historic fires was also recorded (e.g. Bages 1998, la Jonquera l 2012 etc.).

PRODUCTS

Module 1: Vegetation cartography server
The information offered by PREVINCAT to generate a. LCP file for any zone with homogeneous fire regime (ZHR) is:

• Digital Terrain Model (m)
• Slope (%)
• Aspect (degrees)
• Canopy cover (%)
• Canopy height (m)
• Crown base height (m)
• Canopy bulk density (Kg/m3)
• Fuel model (according to Scott & Burgan 2005)

A composite landscape file (.LCP) consists on a multi-layer raster, containing descriptive information about the terrain and fuels, that can be used as a base to run spatial explicit fire simulation. These type of files are the required input for simulators like FARSITE, FlamMap, WildfireAnalyst, etc.

Landscape file (LCP) description (Finney, 2005)

Module 2: Climatic data server
The PREVINCAT server offers the meteorological scenarios in two different formats:

• Static meteorological scenarios: as an average of the hourly data from the days identified as risky. These data can be used on static fire spread simulators (Flammap, BehavePlus, etc.) with the objective of foreseen the potential fire across the whole landscape.
• Dynamic meteorological scenarios: hour by hour data from the risky days. These data can be used on dynamic fire spread simulators (FARSITE, WFA, etc.) with the objective of predict the fire spread pattern given a single or multiple fire initiation points.

Data description
All the cartography can be downloaded in ASCII format (layers with a 20x20m resolution and UTM 31N ETRS89 Datum coordinate system)
Additionally, for each ZHR, meteorological information on reference fires and on those conditions identified from stations data as potential precursors of large fires was recorded and uploaded on PREVINCAT. These meteorological files (.WTR) and wind files (.WND) are ready on the server to be used to run FARSITE and FlamMap simulations.