# Prevailing Winds: Is the speed relevant?

The Wikipedia entry for prevailing winds notes (in my understanding) that the prevailing wind direction is determined solely by the mode of the directions. They note another qualifier called 'dominant wind' which should give the direction of the strongest wind.

However, it feels a bit counter-intuitive that if we had (an example) over 8h a wind from the east at 2m/s and after that a 10m/s wind from the north for 4h, that the prevailing wind direction would still be east.

Based on the above example, would the prevailing wind direction be east?

Is the term 'dominant wind' (with a northerly direction) used in common meteorology, and would that be (in the above example) northerly?

• You might want to think about why the concept of "prevailing wind" is useful. I've seen it most often in describing rain shadows and other things linked to precipitation and moisture transport. Those phenomena are more linked with wind direction than with strength, especially since weak wind at the surface might still have strong winds higher up. Aug 4 '17 at 7:16

Part of the complication goes to two separate definitions for the word:

prevailing:

To be frequent (Miriam-Webster)

The most frequent, the most occurring wind direction in your example, even if weak, is east. So that would be the prevailing wind, as that's the definition we generally use in meteorology.

So don't think of it as

Having superior power or influence (Dictionary.com)

as indeed, that'd become a bit more subjective of a concept.

Sometimes you may hear about the average direction (which would be east-northeast in your example). So perhaps that's more what you're wishing for? Of course if you have an east wind for half the day and a west wind for half the day, the average wind direction would be either north or south. Or, if instead of treating direction and speed differently, you took the vector average... you'd have a mean windspeed of nothing!

Looks like this MathSE question fairly discusses some of the benefits and troubles with different methods of summarizing wind direction... which line up closely with the more general disparities of each of mean, median, and mode.

Alternatively, wind roses can be much more informative and enticing. They give both an indication of the directions (by the length of the bars in each direction) and the intensity (color coded). You can find things, such as: in Colorado Springs in May (as shown below), winds aren't from the southwest very often... but when they are, they tend to be fairly strong (whereas when winds are from about east, as is more common, they're typically quite light):

Looks like you can make some pretty snazzy/customizable recent wind rose plots of locations around the world at the Iowa Environmental Mesonet site.

But the take home is: in meteorology we typically define prevailing wind as simply the most common direction, with no consideration of the speed distribution.

• Thanks for the detailed answer, and for tying the question into wind roses. I was actually investigating it because I was looking into how to code a wind rose implementation from speed and bearing data, and I wanted to make sure the principles were solid before looking into the actual coding bit. Your comments on averaging wind speed/direction are eminently sensible as losing the directionality of the data also gets rid of the major value in it. Jun 29 '17 at 16:45

I think when determining the dominant or prevailing wind direction it is mainly used to describe a longer term than a day (perhaps a month, a year or the climatology in that spot). In my opinion the prevailing wind in the sense you challenge it might be a bit subjective to the one who is assessing it or coding it from a time series. By looking at a wind rose (a figure for example in here), that has the wind speeds and directions as a function of occurrence in the radial scale, you could interpret that the prevailing wind could be to the direction that has the most occurrence or another person might say that it is the other maximum that is slightly wider but not that sharp and could even have stronger winds.

For example from continuous wind data lets say for wind direction resolution 0.5 degrees there might be a lone spike that has the most occurrence, but judging by the eye another maximum is much wider and clearer. Most wind roses for readability the wind directions are even binned (south, south-east..) that flattens the lone spikes.

To conclude: different people may interpret the data differently when dealing with borderline cases.

Prevailing Wind is predominantly related to the most common wind direction and how long in a period of time the wind is blowing in this direction.