fire_exp() returns a SpatRaster of wildfire exposure values calculated
using the input hazard fuel raster for a specified transmission distance.
Usage
fire_exp(hazard, tdist = c("l", "s", "r"), no_burn)Arguments
- hazard
a SpatRaster that represents hazardous fuels for the transmission distance specified in tdist
- tdist
a character vector, can be:
"l"for long-range embers (Default),"s"for short-range embers or,"r"for radiant heat- no_burn
(optional) a SpatRaster that represents the non-burnable landscape. Any cells that cannot receive wildfire (e.g. open water, rock) and any cells that are not natural (e.g. built environment, irrigated agricultural areas) should be of value 1, all other cells should be NODATA. This parameter should be provided if preparing data for
fire_exp_validate()
Details
This function is the primary function in this package, the output from this function serves as the first input to all other functions in this package.
Background
Wildfire exposure was first proposed for community scale assessments by Beverly et al. (2010). Wildfire entry points to the built environment were identified by evaluating the surrounding wildland fuels as potential ignition sources. These methods were adapted and validated by Beverly et al. (2021) for landscape scale assessments. Visual aids to better understand these concepts are available in the Wildfire Exposure Assessment Guidebook (FireSmart Canada 2018) which includes intuitive illustrations.
Three transmission distances were defined in Beverly et al. (2010) for
different scales of wildfire ignition processes. Different fuel types
are capable of transmitting fire at different scales. The transmission
distances define a potential maximum spread distance from an ignition
source. If the default distances do not accurately represent fire behaviour
in your area of interest you can adjust them with fire_exp_adjust().
Radiant heat
The default transmission distance for radiant heat is 30 meters. At this scale of wildfire transmission wildfire can spread from direct flame contact or thermal radiation warming of fuels.
Ember spotting
Embers (AKA firebrands) can be carried by wind beyond the active fire front. To determin two different ember spreading distances Beverly et al. (2010) reviewed fire-spot distance observations and applied predictive models to verify these distances were a reasonable assumption.
Short-range embers
The default transmission distance for short-range embers is 100 meters. Some fuel types have the potential to produce embers over short distances. An example fuel type that has the potential to transmit short-range embers is a deciduous forest. Fuels considered hazardous to short-range ember ignition will also be hazardous to transmit fire via radiant heat or direct flame contact.
Long-range embers
The default transmission distance for long-range embers is 500 meters. Some fuel types have the potential to transmit embers across large distances. This landscape scale process is often a contributor to large, fast-moving fires. Any fuels that are considered hazardous for long-range ember spotting could also transmit fire by short-range ember spotting and radiant heat. An example fuel type that has the potential to transmit long-range embers is a mature pine stand.
Although it is possible for embers from certain fuel types to be carried much farther than the defined 500 meter transmission distance under extreme conditions, these cases are relatively uncommon.
The 500 meter spread distance has been further validated across Alberta (Beverly et al. 2021), in Alaska (Schmidt et al. 2024), across the entire Canadian landbase (manuscript in preparation), and in Portugal (manuscript in preparation).
Technical
How the metric is calculated
The wildfire exposure metric is calculated for each cell individually using a focal window of the surrounding cells within the specified transmission distance. The proportion of cells within the assessment window that are classed as hazardous is returned. The wildfire exposure metric has a range of 0-1. A wildfire exposure value of 0.5 can be interpreted as 50% of cells within the specified transmission distance area have the potential to spread fire to the assessment cell. A value of 0.5 can also be interpreted in the other direction, a fire in the assessment cell could potentially spread to 50% of the surrounding cells within the specified transmission distance.
Input features
An input hazard raster must be prepared by the user in accordance with the
intended use. First, refer to the Get Started vignette by running
vignette("fireexposuR") to determine what the data requirements are for
the intended application. Then refer to the Preparing Input Data vignette by
running vignette("prep-input-data") for lots of recommendations, advice,
and examples.
A separate hazard raster should be prepared for each of the transmission distances of interest. There are also minimum spatial resolution and extent requirements for each transmission distances.
Long-range embers:
minimum raster resolution is 150 meters
The dimensions of the data must be wider/taller than 1000 meters because 500 meters of data will be lost along the perimeter due to edge effects
Short-range embers:
minimum raster resolution is 33 meters
The dimensions of the data must be wider/taller than 200 meters because 100 meters of data will be lost along the perimeter due to edge effects
Radiant heat:
minimum raster resolution is 10 meters
The dimensions of the data must be wider/taller than 60 meters because 30 meters of data will be lost along the perimeter due to edge effects
References
Beverly JL, McLoughlin N, Chapman E (2021) A simple metric of landscape fire exposure. Landscape Ecology 36, 785-801. DOI
Beverly JL, Bothwell P, Conner JCR, Herd EPK (2010) Assessing the exposure of the built environment to potential ignition sources generated from vegetative fuel. International Journal of Wildland Fire 19, 299-313. DOI
FireSmart Canada (2018) Wildfire exposure assessment guidebook. Available here
Hijmans R (2024). terra: Spatial Data Analysis. R package version 1.7-78, CRAN.
Schmidt JI, Ziel RH, Calef MP, Varvak A (2024) Spatial distribution of wildfire threat in the far north: exposure assessment in boreal communities. Natural Hazards 120, 4901-4924. DOI
Examples
# read example hazard data
hazard_file_path <- "extdata/hazard.tif"
hazard <- terra::rast(system.file(hazard_file_path, package = "fireexposuR"))
# compute long range exposure
fire_exp(hazard, tdist = "l")
#> class : SpatRaster
#> dimensions : 321, 321, 1 (nrow, ncol, nlyr)
#> resolution : 100, 100 (x, y)
#> extent : 384000, 416100, 6033900, 6066000 (xmin, xmax, ymin, ymax)
#> coord. ref. : NAD83 / Alberta 10-TM (Forest) (EPSG:3400)
#> source(s) : memory
#> varname : hazard
#> name : exposure
#> min value : 0
#> max value : 1