bm_SRE {biomod2}

Surface Range Envelope

Description

This internal biomod2 function allows the user to run a rectilinear surface range envelop (SRE) (equivalent to BIOCLIM) using the extreme percentiles (as recommended by Nix or Busby, see References and Details).

Usage

bm_SRE(
  resp.var = NULL,
  expl.var = NULL,
  new.env = NULL,
  quant = 0.025,
  do.extrem = FALSE
)

Arguments

resp.var

a vector, a SpatVector without associated data (if presence-only), or a SpatVector object containing binary data (0 : absence, 1 : presence, NA : indeterminate) for a single species that will be used to build the species distribution model(s)
Note that old format from sp are still supported such as SpatialPoints (if presence-only) or SpatialPointsDataFrame object containing binary data.

expl.var

a matrix, data.frame, SpatVector or SpatRaster object containing the explanatory variables (in columns or layers) that will be used to build the SRE model
Note that old format from raster and sp are still supported such as RasterStack and SpatialPointsDataFrame objects.

new.env

a matrix, data.frame, SpatVector or SpatRaster object containing the explanatory variables (in columns or layers) that will be used to predict the SRE model
Note that old format from raster and sp are still supported such as RasterStack and SpatialPointsDataFrame objects.

quant

a numeric between 0 and 0.5 defining the half-quantile corresponding to the most extreme value for each variable not to be taken into account for determining the tolerance boundaries of the considered species (see Details)

do.extrem

(optional, default FALSE)
A logical value defining whether a matrix containing extreme conditions supported should be returned or not

Details

Please refer to References to get more information about surface range envelop models.

This method is highly influenced by the extremes of the data input. Whereas a linear model can discriminate the extreme values from the main tendency, the SRE considers them as important as any other data point leading to changes in predictions.

The more (non-colinear) variables, the more restrictive the model will be.

Predictions are returned as binary (0 or 1) values, a site being either potentially suitable for all the variables, or out of bounds for at least one variable and therefore considered unsuitable.

quant determines the threshold from which the data will be taken into account for calibration. The default value of 0.05 induces that the 5% most extreme values will be avoided for each variable on each side of its distribution along the gradient, meaning that a total of 10% of the data will not be considered.

Value

A vector or a SpatRaster object, containing binary (0 or 1) values.

Author(s)

Wilfried Thuiller, Bruno Lafourcade, Damien Georges

References

• Nix, H.A., 1986. A biogeographic analysis of Australian elapid snakes. In: Atlas of Elapid Snakes of Australia. (Ed.) R. Longmore, pp. 4-15. Australian Flora and Fauna Series Number 7. Australian Government Publishing Service: Canberra.

• Busby, Jeremy. BIOCLIM - a bioclimate analysis and prediction system. Plant protection quarterly 6 (1991): 8-9.

See Also

bm_PseudoAbsences, BIOMOD_FormatingData, bm_ModelingOptions, bm_Tuning, bm_RunModelsLoop, BIOMOD_Modeling,

Other Secundary functions: bm_BinaryTransformation(), bm_CrossValidation(), bm_FindOptimStat(), bm_MakeFormula(), bm_ModelingOptions(), bm_PlotEvalBoxplot(), bm_PlotEvalMean(), bm_PlotRangeSize(), bm_PlotResponseCurves(), bm_PlotVarImpBoxplot(), bm_PseudoAbsences(), bm_RunModelsLoop(), bm_SampleBinaryVector(), bm_SampleFactorLevels(), bm_Tuning(), bm_VariablesImportance()

Examples

library(terra)
## Load real data
data(DataSpecies)
myResp.r <- as.numeric(DataSpecies[, 'GuloGulo'])

data(bioclim_current)
myExpl.r <- rast(bioclim_current)

myRespXY <- DataSpecies[which(myResp.r == 1), c('X_WGS84', 'Y_WGS84')]
myResp.v <- classify(subset(myExpl.r, 1), 
                     matrix(c(-Inf, Inf, 0), ncol = 3, byrow = TRUE))
myResp.v[cellFromXY(myResp.v, myRespXY)] <- 1

## Compute SRE for several quantile values
sre.100 <- bm_SRE(resp.var = myResp.v,
                  expl.var = myExpl.r,
                  new.env = myExpl.r,
                  quant = 0)
sre.095 <- bm_SRE(resp.var = myResp.v,
                  expl.var = myExpl.r,
                  new.env = myExpl.r,
                  quant = 0.025)
sre.090 <- bm_SRE(resp.var = myResp.v,
                  expl.var = myExpl.r,
                  new.env = myExpl.r,
                  quant = 0.05)

## Visualize results
res <- c(myResp.v, sre.100, sre.095, sre.090)
names(res) <- c("Original distribution", "Full data calibration"
                , "Over 95 percent", "Over 90 percent")
plot(res)