scenfire (version 0.1.0)

calc_abp: Calculate annual burned probability (BP) from perimeters

Description

This function calculates the burned probability raster from a set of simulated fire perimeters. It first converts the perimeters into a `terra` vector object, ensures their validity, and then rasterizes them onto a template raster. The resulting raster indicates the number of times each pixel has been burned, which is then normalized by a scaling factor derived from the total simulated area and a reference surface.

Usage

calc_abp(template_raster, candidate_surfaces, reference_surface)

Value

A `terra::SpatRaster` object representing the burned probability. Each cell value indicates the proportion of times that pixel was burned, normalized by the ratio of total simulated area to the reference surface.

Arguments

template_raster

A `terra::SpatRaster` object that defines the extent, resolution, and CRS for the output burned probability raster.

candidate_surfaces

An `sf` object (simple features) representing the simulated fire perimeters. It is assumed to have a column named `size` containing the area of each perimeter.

reference_surface

Numeric value. A reference total surface area used for normalization. This typically represents the target total burned area (e.g., historical average annual burned area).

Examples

# Create a dummy template raster
library(terra)
template_raster_ex <- rast(nrows=10, ncols=10, xmin=0, xmax=100, ymin=0, ymax=100,
                           crs="EPSG:25830")

# Create dummy candidate_surfaces (sf object with polygons)
# Ensure 'size' column exists
library(sf)
poly1 <- st_polygon(list(cbind(c(10,10,30,30,10), c(10,30,30,10,10))))
poly2 <- st_polygon(list(cbind(c(50,50,70,70,50), c(50,70,70,50,50))))
poly3 <- st_polygon(list(cbind(c(20,20,40,40,20), c(20,40,40,20,20))))

candidate_surfaces_ex <- st_sf(
  id = 1:3,
  size = c(400, 400, 400), # Dummy sizes for example
  geometry = st_sfc(poly1, poly2, poly3),
  crs = 25830
)

# Example reference surface
reference_surface_ex <- 1000

# Calculate burned probability
bp_raster <- calc_abp(template_raster = template_raster_ex,
                     candidate_surfaces = candidate_surfaces_ex,
                     reference_surface = reference_surface_ex)

plot(bp_raster, main = "Burned Probability Raster")