toxicIntendity function wrapping dispersal and exposure
toxicIntensity(
object,
sf,
size_raster = 2^10,
tolerance_square = 0.1,
kernel = "NIG",
kernel.options = list(a1 = 0.2073, a2 = 0.2073, b1 = 0.3971, b2 = 0.3971, b3 =
0.0649, theta = 0),
loss = NULL,
beta = 0.4,
nbr_cores = 1,
squared_frame = NULL,
quiet = FALSE
)sf or SpatialPolygonsDataFrame. A simple feature of class sf or SpatialPolygonsDataFrame
sf. And object of class `sf` on which exposure is computed from the previous list of raster by patch `RasterStack_dispersal`. See sf for details.
integer. Raster size (default = 2^10)
numeric. Tolerance rate to test if an sf set is squared
string. Dispersion kernel, function name (default = NIG)
list. Parameters list for the kernel function
numeric. Numeric vector to applied a loss on exposure cells.
numeric. toxic adherence parameter between 0 and 1 (default = 0.4).
integer. Parameters for parallel computing: the
number of cores to use, i.e. at most how many child processes
will be run simultaneously. Default is 1 (non parallel).
sf. Select the sf to be considered as frame to rasterized. Default is `NULL`, and `object` is used.
boolean. Set `TRUE` to remove progress bar.
The dispersal of contaminants is implemented by rastering the landscape and by computing the convolution between sources emissions and a dispersal kernel.
The dispersion kernel by default is Normal Inverse Gaussian kernel ("NIG" function).
Currently, two others are implemented "geometric" (with parameter a) and "2Dt" kernels
(with parameters a, b, c1, c2).
Local intensity depends of beta and alpha parameters. Beta represents the toxic adherence between [0,1].
Alpha represents a list of parameters of the lost of toxic particules due to covariates (precipitation).
There are two configurations to integrate the loss in the function :
(i) simulating covariate (simulate=TRUE) or (ii) uploading covariate (simulate=FALSE).
The covariate is linked to the loss by a linear regression with paramaters minalpha, maxalpha, covariate_threshold.