PEThorn: Thornthwaite's Potential Evapotranspiration
In jarioksa/biogeo: Tools for Biogeography Lectures

Description Usage Arguments Details Value Author(s) References See Also Examples

Function estimates the Potential Evapotranspiration (PET) from monthly mean temperatures following Thornthwaite (1948). The original equations were calibrated for U.S.A., and were the first attempt at finding PET from readily available meteorological data.

1 2	PEThorn(temp, lat, monthly = FALSE, heatlimitUSA = TRUE, daylimit50 = TRUE)

`temp`	Vector of 12 values of monthly mean temperatures in degrees C.
`lat`	Latitude in degrees.
`monthly`	(logical) Return monthly values instead of annual total.
`heatlimitUSA`	(logical) Limit heat index to calibrated values within the U.S.A.
`daylimit50`	(logical) Do not use latitudes >50 in daylength calculation.

Thornthwaite (1948) used different models for hot and cool months with limit of 26.5 degrees C. Below 26.5 C, same monthly mean temperature generated lower PET in hot than in cool climate, and this was accommodated for by a ‘heat index’. The monthly contribution to the heat index is I = (T/5)^1.514, where T is the monthly mean temperature, and the heat index is the sum of monthly values. Below 26.5 C the monthly PET for is calculated as 16*(10*T/I)^A, where the exponent A is defined so that with all values of I the monthly PET will be 135 mm at 26.5 degrees C. Thornthwaite (1948) gives a third degree polynomial to find A from I, and this works fairly well within the range I = 20 … 140 which covers the U.S.A. except Alaska and Northwestern high mountains, but fails badly in hotter and cooler climates. The original polynomial model restricted to the valid range is used with option heatlimitUSA=TRUE. If this argument is FALSE, we use instead a more accurate derivation of exponent A that is also valid outside the U.S.A. range, and gives similar results within U.S.A. (this is unpublished and developed here: see the code). Above 26.5 C, Thornthwaite (1948) tabulated PET against monthly mean. The tabulated values seem to be derived from a parabolic model, and we use coefficients of a second degree polynomial fitted to the tabulated values. The monthly PET values are for 12 hours day length, and they are adjusted to the actual day length of the latitude. If daylimit50=TRUE, we use day length at 50 degrees latitude for all more extreme locations, like suggested by Thornthwaite (1948).

Function thornthwaite (package SPEI) provides an alternative implementation, but it uses the same basic equations also in hot climates (montly mean > 26.5 degrees), and does not restrict the range of accepted ‘heat index’. Small numerical differences are also caused by the algorithms of estimating the day length.

Annual potential evapotranspiration in mm, or monthly values if monthly=TRUE.

Jari Oksanen

Thornthwaite C.W. (1948) An approach toward a rational classification of climate. Geographical Review 38, 55–94.

Package SPEI has a wider choice of indices of PET, among them Hargreaves index and its modification (hargreaves) which is also suitable for standard meteorological data, and has been calibrated globally.