Skip to contents

Adjust modelled AET by land cover, as Chapman et al. (2018) did

Source: R/wet_aet_landcover.R
wet_aet_landcover.Rd

Chapman, Kerr & Wilford (2012, p. 83) multiply the CGIAR AET by a per-class ratio of measured AET for the cover to the modelled (agronomic-crop) AET, drawing the measured values from the table published as Table 3 of Chapman et al. (2018, p. 8). The ratios they used were tuned in calibration and never published, so here each class's ratio is fixed without fitting:

Usage

wet_aet_landcover(
  aet,
  frac,
  et_class = NULL,
  domain = NULL,
  min_cells = 1,
  clamp = c(0.25, 4)
)

Arguments

aet

One-layer SpatRaster of modelled annual AET (mm). The ratio denominators are taken over its non-NA cells.

frac

SpatRaster of cover fractions (0-1), one layer per class, named by class, on the grid of aet; e.g. wet_landcover_nrcan().

et_class

Named numeric: measured AET (mm) per class, with a value for every layer of frac. Defaults to wet_chapman_table3().

domain

Optional one-layer SpatRaster on the grid of aet: only cells where it is TRUE (non-zero) count towards the ratio denominators. The adjustment is still applied everywhere.

min_cells

Minimum majority cells for a class to get its own ratio.

clamp

Lower and upper limits of a class ratio.

Value

list(aet, ratio): the adjusted AET (SpatRaster, mm) and a data.frame(class, et_mm, aet_mean, n_cells, ratio).

Details

ratio_c = et_class[c] / mean(aet over the cells where c is the majority)

and the multiplier of each cell is sum_c frac_c * ratio_c + (1 - sum_c frac_c). Ground no class covers (outside the land-cover product) keeps ratio 1, i.e. the input AET. A cell votes for a majority class only when more than half of it is covered and it lies in domain. A class with fewer than min_cells majority cells gets ratio 1, and every ratio is clamped to clamp, so a class whose modelled AET is near 0 (snow and ice) cannot scale its cells without limit.

Examples

r <- terra::rast(nrows = 1, ncols = 3, xmin = 0, xmax = 3, ymin = 0, ymax = 1)
aet <- terra::setValues(r, c(350, 330, 240))
frac <- c(terra::setValues(r, c(1, 1, 0)), terra::setValues(r, c(0, 0, 1)))
names(frac) <- c("coniferous", "grass")
out <- wet_aet_landcover(aet, frac)
out$ratio
#>        class et_mm aet_mean n_cells     ratio
#> 1 coniferous   276      340       2 0.8117647
#> 2      grass   275      240       1 1.1458333
terra::values(out$aet, mat = FALSE)
#> [1] 284.1176 267.8824 275.0000