prepraster: Prepare the structural raster

Description

This function prepares the structural raster for the follow-up analyses. The size and extent of the structural raster defines the resolution at which the isoscapes and the assignments are defined.

Usage

prepraster(
  raster,
  isofit = NULL,
  margin_pct = 5,
  aggregation_factor = 0L,
  aggregation_fn = mean,
  manual_crop = NULL,
  values_to_zero = c(-Inf, 0),
  verbose = interactive()
)

Value

The prepared structural raster of class SpatRaster

Arguments

raster: The structural raster (SpatRaster)
isofit: The fitted isoscape model returned by the function isofit
margin_pct: The percentage representing by how much the area should extend outside the area used for cropping (default = 5, corresponding to 5%). Set to 0 if you want exact cropping.
aggregation_factor: The number of neighbouring cells (integer) to merge during aggregation
aggregation_fn: The function used to aggregate cells
manual_crop: A vector of four coordinates (numeric) for manual cropping, e.g. the spatial extent
values_to_zero: A numeric vector of length two specifying the range of values for the structural raster that must be turned into 0. Default is c(-Inf, 0) which for an elevation raster brings all seas to an elevation of zero. For using IsoriX for marine organisms, you should use c(0, Inf) instead.
verbose: A logical indicating whether information about the progress of the procedure should be displayed or not while the function is running. By default verbose is TRUE if users use an interactive R session, and FALSE otherwise.

Details

This functions allows the user to crop a raster according to either the extent of the isoscape or manually. If a fitted isoscape object is provided (see isofit), the function extracts the observed locations of isotopic sources from the model object and crops the structural raster accordingly. Alternatively, manual_crop allows you to crop the structural raster to a desired extent. If no model and no coordinates for manual cropping are provided, no crop will be performed. Importantly, cropping is recommended as it prevents extrapolations outside the latitude/longitude range of the source data. Predicting outside the range of the source data may lead to highly unreliable predictions.

Aggregation changes the spatial resolution of the raster, making computation faster and using less memory (this can affect the assignment; see note below). An aggregation factor of zero (or one) keeps the resolution constant (default).

This function relies on calls to the functions terra::aggregate and terra::crop from the package terra. It thus share the limitations of these functions. In particular, terra::crop expects extents with increasing longitudes and latitudes. We have tried to partially relax this constrains for longitude and you can use the argument manual_crop to provide longitudes in decreasing order, which is useful to centre a isoscape around the pacific for instance. But this fix does not solve all the limitations as plotting polygons or points on top of that remains problematic (see example bellow). We will work on this on the future but we have other priorities for now (let us know if you really need this feature).

Examples

Run this code


## The examples below will only be run if sufficient time is allowed
## You can change that by typing e.g. options_IsoriX(example_maxtime = XX)
## if you want to allow for examples taking up to ca. XX seconds to run
## (so don't write XX but put a number instead!)

if (getOption_IsoriX("example_maxtime") > 30) {
  ## We fit the models for Germany
  GNIPDataDEagg <- prepsources(data = GNIPDataDE)

  GermanFit <- isofit(
    data = GNIPDataDEagg,
    mean_model_fix = list(elev = TRUE, lat_abs = TRUE)
  )

  ### Let's explore the difference between aggregation schemes

  ## We aggregate and crop using different settings
  ElevationRaster1 <- prepraster(
    raster = ElevRasterDE,
    isofit = GermanFit,
    margin_pct = 0,
    aggregation_factor = 0
  )

  ElevationRaster2 <- prepraster(
    raster = ElevRasterDE,
    isofit = GermanFit,
    margin_pct = 5,
    aggregation_factor = 5
  )

  ElevationRaster3 <- prepraster(
    raster = ElevRasterDE,
    isofit = GermanFit,
    margin_pct = 10,
    aggregation_factor = 5, aggregation_fn = max
  )

  ## We plot the outcome of the 3 different aggregation schemes using terra

  oripar <- par(mfrow = c(1, 3)) ## display 3 plots side-by-side

  plot(ElevationRaster1, main = "Original small raster")
  polys(CountryBorders)
  polys(OceanMask, col = "blue")

  plot(ElevationRaster2, main = "Small raster aggregated (by mean)")
  polys(CountryBorders)
  polys(OceanMask, col = "blue")

  plot(ElevationRaster3, main = "Small raster aggregated (by max)")
  polys(CountryBorders)
  polys(OceanMask, col = "blue")

  par(oripar) ## restore graphical settings
}

## The examples below will only be run if sufficient time is allowed
## You can change that by typing e.g. options_IsoriX(example_maxtime = XX)
## if you want to allow for examples taking up to ca. XX seconds to run
## (so don't write XX but put a number instead!)

if (getOption_IsoriX("example_maxtime") > 10) {
  ### Let's create a raster centered around the pacific

  ## We first create an empty raster
  EmptyRaster <- rast(matrix(0, ncol = 360, nrow = 180))
  ext(EmptyRaster) <- c(-180, 180, -90, 90)
  crs(EmptyRaster) <- "+proj=longlat +datum=WGS84"

  ## We crop it around the pacific
  PacificA <- prepraster(EmptyRaster, manual_crop = c(110, -70, -90, 90))
  ext(PacificA) # note that the extent has changed!

  ## We plot (note the use of the function shift()!)
  plot(PacificA, col = "blue", legend = FALSE)
  polys(CountryBorders, col = "black")
  polys(shift(CountryBorders, dx = 360), col = "black")
}