ensemble.envirem: Calculate bioclimatic variables via the 'envirem' package for...

In BiodiversityR: Package for Community Ecology and Suitability Analysis

Calculate bioclimatic variables via the envirem package for data.frames.

Description

Function generateEnvirem uses RasterStack (stack) objects as input and also generates outputs in the same format. The functions described here can be used to generate the bioclimatic variables for data.frames while using envirem functions internally. This feature can be useful in situations where models are calibrated with higher resolution data, but where maps will only be generated in lower resolutions, thus avoiding the need to generate the higher resolution envirem layers first.

Usage

ensemble.envirem.masterstack(
    x,
    precipstack, 
    tmaxstack, tminstack, 
    tmeanstack = NULL,
    envirem3 = TRUE)

ensemble.envirem.solradstack(
    x, solrad,
    envirem3 = TRUE)

ensemble.envirem.run(
    masterstack, solradstack,
    var = "all", ...)

Arguments

x	Point locations provided in 2-column (eg, LON-LAT) format.
precipstack	RasterStack object (stack) or SpatRaster object (rast) containing monthly precipitation data.
tmaxstack	RasterStack object (stack) or SpatRaster object (rast) containing monthly maximum temperature data.
tminstack	RasterStack object (stack) or SpatRaster object (rast) containing monthly minimum temperature data.
tmeanstack	RasterStack object (stack) or SpatRaster object (rast) containing monthly average temperature data.
envirem3	generate a SpatRaster object (rast) as used by envirem 3.
solrad	RasterStack object (stack) or SpatRaster object (rast) containing monthly extrasolar radiation data.
masterstack	RasterStack object (stack) expected to have been created via ensemble.envirem.masterstack.
solradstack	RasterStack object (stack) expected to have been created via ensemble.envirem.solradstack.
var	Names of bioclimatic variables to be created; see: generateEnvirem.
...	Other arguments for generateEnvirem, dealing with the scale of temperature or precipitation data.

Details

The objective of these functions is to expand a data.frame of explanatory variables that will be used for calibrating species distribution models with bioclimatic variables that are generated by the envirem package (See details in generateEnvirem).

It is important that monthly values are sorted sequentially (January - December) as the ensemble.envirem.masterstack and ensemble.envirem.solradstack functions expect the inputs to be sorted in this order.

Function ensemble.envirem.solradstack requires monthly extraterrestrial solar radiation layers at the same resolution as the climatic layers. It is possible, however, to also calculate these values directly for point observation data as shown below in the examples.

Value

Function ensemble.envirem.run returns a data.frame with bioclimatic variables for each point location.

Author(s)

Roeland Kindt (World Agroforestry Centre)

References

Title P.O., Bemmels J.B. 2018. ENVIREM: An expanded set of bioclimatic and topographic variables increases flexibility and improves performance of ecological niche modeling. Ecography 41: 291-307.

Kindt R. 2023. TreeGOER: A database with globally observed environmental ranges for 48,129 tree species. Global Change Biology. \Sexpr[results=rd]{tools:::Rd_expr_doi("10.1111/gcb.16914")}

See Also

generateEnvirem, ensemble.calibrate.models, ensemble.calibrate.weights

Examples

## Not run: 
# Based on examples in envirem package for envirem::generateEnvirem
# Modified Sep-2023 due to change in function name in envirem

library(terra)
library(envirem)

# Find example rasters
rasterFiles <- list.files(system.file('extdata', package='envirem'), 
                          full.names=TRUE)

precip.files <- rasterFiles[grepl(pattern="prec_", 
                                x=rasterFiles)]
precip.files <- precip.files[c(1, 5:12, 2:4)]
precip.stack <- terra::rast(precip.files)
precip.stack
names(precip.stack)

tmin.files <- rasterFiles[grepl(pattern="tmin_", 
                                x=rasterFiles)]
tmin.files <- tmin.files[c(1, 5:12, 2:4)]
tmin.stack <- terra::rast(tmin.files)
tmin.stack
names(tmin.stack)

tmax.files <- rasterFiles[grepl(pattern="tmax_", 
                                x=rasterFiles)]
tmax.files <- tmax.files[c(1, 5:12, 2:4)]
tmax.stack <- terra::rast(tmax.files)
tmax.stack
names(tmax.stack)

tmean.files <- rasterFiles[grepl(pattern="tmean_", 
                                x=rasterFiles)]
tmean.files <- tmean.files[c(1, 5:12, 2:4)]
tmean.stack <- terra::rast(tmean.files)
tmean.stack
names(tmean.stack)

# Random locations
locs <- dismo::randomPoints(raster::stack(precip.stack[[1]]), n=50)

# Climatic data
master.input <- ensemble.envirem.masterstack(x=locs,
                             precipstack=precip.stack,
                             tmaxstack=tmax.stack,
                             tminstack=tmin.stack,
                             tmeanstack=tmean.stack)

# Calculate solar radiation for 1975
# (Use other midpoint for the 1970-2000 WorldClim 2.1 baseline)
solrad.stack <- ETsolradRasters(precip.stack[[1]], 
                             year = 1975-1950, 
                             outputDir = NULL)

solrad.input <- ensemble.envirem.solradstack(x=locs,
                             solrad=solrad.stack)

# Obtain the envirem bioclimatic data

envirem.data1 <- ensemble.envirem.run(masterstack=master.input,
                             solradstack=solrad.input,
                             tempScale=10)

# Generate all the envirem layers, then extract
# See envirem package for envirem::generateEnvirem

worldclim <- rast(c(precip.files, tmax.files, tmin.files, tmean.files))
names(worldclim)

assignNames(precip = 'prec_##')

# generate all possible envirem variables
envirem.stack <- generateEnvirem(worldclim, solrad.stack, var='all', tempScale = 10)

# set back to defaults
assignNames(reset = TRUE)

envirem.data2 <- extract(envirem.stack, y=locs)

# compare
envirem.data1 - envirem.data2

# Calculate extraterrestrial solar radiation for point observations
solrad1 <- extract(solrad.stack, y=locs)
solrad2 <- array(dim=c(nrow(locs), 12))
for (i in 1:nrow(locs)) {
  lat.i <- locs[i, 2]
  for (m in 1:12) {
    solrad2[i, m] <- envirem:::calcSolRad(year=1975-1950,
                                          lat=lat.i,
                                          month=m)
  }
}

solrad1 - solrad2


## End(Not run)

BiodiversityR documentation built on June 22, 2024, 11:57 a.m.