Calculating wind u- and v-components from direction and speed

[mpvginde]

Hi,
I think there is a mistake in the calculation of the wind u- and v-components from wind speed and direction:

meteogrid/R/geowind.R

Lines 49 to 63 in 9da2345

	wind.uv <- function(wspeed, wdir, fieldname=c("U","V"), rad=FALSE){
	if (missing(wdir) && is.list(wspeed)){
	wdir <- wspeed$wdir
	wspeed <- wspeed$wspeed
	}
	if (!rad) wdir <- wdir*pi/180.
	u <- wspeed*cos(wdir)
	v <- wspeed*sin(wdir)
	if (is.geofield(wdir)) {
	attr(u, "info")$name <- fieldname[1]
	attr(v, "info")$name <- fieldname[2]
	}
	if (is.vector(wspeed)) return(data.frame(U=u,V=v))
	else return(list(U=u,V=v))
	}

The cos and sin directly use the direction as angle.
But this is the direction the wind is blowing from (meteorological convection).
The true angle to be used for the cos and sin function should be:
270 - direction (rotate wind 180 degrees + 90 degrees to account for the fact that the N/S axis is used as reference in stead of E/W axis)

As it is now, if you apply wind.dirspeed and then again wind.uv the original u an v are swapped and have the opposite sign.

Michiel

[mpvginde]

@adeckmyn

[adeckmyn]

Indeed, that is an error (or at the very least an inconsistency). The function is rarely used (I think), so it has been there for many years. Which of course means that changing this "long time behaviour" may actually cause a problem for a few users...