Skip to contents

Annual AET and precipitation from the TerraClimate 1981-2010 climatology

Source: R/wet_terraclimate_aet.R
wet_terraclimate_aet.Rd

Downloads the TerraClimate monthly climatology files (Abatzoglou et al. 2018, Scientific Data 5, 170191; CC0) for period, sums the 12 months into annual totals, and resamples them bilinearly from TerraClimate's 1/24-degree lattice onto grid. GDAL's bilinear reweights around missing neighbours, so only cells whose own TerraClimate cell is empty (sea, and some large lakes) come out NA; they are left for the caller to fill.

Usage

wet_terraclimate_aet(
  grid,
  period = "19812010",
  vars = c("aet", "ppt"),
  dir = "data/terraclimate",
  overwrite = FALSE,
  timeout = 1800
)

Arguments

grid

Target grid: a SpatRaster or a path to one, in EPSG:4326.

period

Climatology period as in the file names: "19812010", "19611990" or "19912020".

vars

TerraClimate variables to return, as annual totals.

dir

Cache directory.

overwrite

Logical. Rebuild the resampled grid even when cached.

timeout

Seconds allowed for each download.

Value

Path to a GeoTIFF with one layer per variable, named <var>_tc (mm per year).

Details

TerraClimate's AET comes from a Thornthwaite-Mather soil-water balance on its own precipitation, so it carries that precipitation's biases. Its precipitation layer is returned alongside so they can be seen.

The whole global file is downloaded (about 95 MB for AET, 147 MB for precipitation), because the server's subset service returned empty bodies when this was written (2026-09-27). The option wet.terraclimate_url only sets where missing files are fetched from; the cache key is the md5 of the files actually read.

Examples

if (FALSE) { # interactive()
g <- terra::rast(xmin = -120, xmax = -119, ymin = 49, ymax = 50, res = 1 / 120)
tc <- terra::rast(wet_terraclimate_aet(g))
terra::global(tc, "mean", na.rm = TRUE)
}