Fire probability weights for point locations

Compute fire probability weights for point locations.

fire_weights(r, xy, buffer = 20000, weight = "linear",
  lonlat_to_akalbers = TRUE, veg = NULL, age = NULL)

Arguments

r

raster layer or path to file, an ALFRESCO fire scar output geotiff.
xy

a single location, in a 2-column data frame or matrix of longitude and latitude.
buffer

radius of buffer around xy.
weight

inverse distance weighting, "linear" or "quadratic".
lonlat_to_akalbers

if xy is unprojected (lon/lat), reproject to NAD83 Alaska Albers equal area conic projection.
veg

same asr but for accompanying vegetation class ALFRESCO output.
age

same asr but for accompanying vegetation age ALFRESCO output. See details.

Value

a data frame containing a logical column indicating burn or no burn, a column of distances from point location, and a column of weights.

Details

This function applies an inverse distance weighting to grid cells within a radial buffer of a center grid cell. The function is used for extracting values within a buffer around a given point from ALFRESCO fire scar output geotiffs. Points outside the boundary defined by buffer are excluded and have zero influence on estimated probability of fire at the center point. Fire scars that intersect the radial buffer zone contribute more to the estimated likelihood of fire for the point location the closer the fire is to the point.

It is recommended to not use age because ALFRESCO has a longstanding bug involving geographic areas of uninitialized ages, resulting in extreme negative age values that ruin analyses unless the user first implements their own fix to any age geotiffs before using them.

Examples

# NOT RUN {
scar <- "FireScar_0_2004.tif"
veg <- "Veg_0_2004.tif"
age <- "Age_0_2004.tif"
coords <- data.frame(lon = -147.7164, lat = 64.8378)
fire_weights(r = scar, xy = coords, veg = veg, age = age)
# }