Skip to contents

Ground plots from a wider area can inform a calibration, but only in proportion to how much their climate resembles the landscape being modelled. This scores each observation individually, which is the point: aggregating plots into map units first and comparing unit means makes the comparison only as reliable as the thinnest unit, and map units carrying a handful of plots get estimates too noisy to rank.

Usage

growth_climatic_distance(climate, target, vars = names(target), scale = NULL)

Arguments

climate

A data frame with one row per observation and one column per climate variable.

target

Named numeric vector giving the target climate.

vars

Character. Variables to compare on; defaults to the names of target.

scale

Optional named numeric vector of per-variable spreads. Defaults to each variable's standard deviation across climate.

Value

A numeric vector of distances, one per row of climate.

Details

Distance is the root-mean-square deviation across variables after standardizing each by its spread, so that a variable measured in millimetres does not swamp one measured in degrees.