The Call of the Open Sidewalk

For most of my life Environment Canada reported wind chill using units of Watts per square metre. If it was 1500 W/m^2 it was uncomfortable. If it was 1800 W/m^2 you would want to think about covering up exposed skin. If was 2300 W/m^2 then things were sort of life threatening.

At some point it got cold and windy in Southern Ontario. Environment Canada thus reported wind chill values there for first time in a long time. The people of Southern Ontario responded to the use of the units of "Watt" and "metre" with fear and confusion. This generated the political will to come up a new way of expressing wind chill and is the reason this was posted to the political category. What happened next had nothing to do with science.

What was wanted was a way to express wind chill as a temperature. The political people asked the science people to do this. There was a problem. The problem was that expressing wind chill as a temperature makes no sense at all.

People have a body temperature around 37C. If the air temperature is 37C then it does not matter how fast the wind blows, there will be no wind chill. As the temperature decreases then the wind chill will increase (if there is any wind). If the wind then increases there will be more wind chill. This ignores various factors like evaporation but the point is that temperature and wind speed are the factors that cause wind chill. By pretending that wind chill is actually a temperature we are treating one of the things that cause wind chill as the wind chill effect itself. Getting cause and effect backwards is bad. Thinking that cause and effect are the same thing is a lot worse.

So here we have a case where some smart people were asked to do something senseless. In some cultures there would be a fight that might last for years. Canadian culture is famous for compromise. In the case of wind chill I feel we came up with a truly world class compromise.

I like to imagine that the discussion went like this:

INT. MEETING ROOM - DAY

POLITICAL PERSON (brightly) We need to come up with a way to express wind chill as a temperature.

SCIENCE PERSON blinks. S/he looks thoughtful for a moment.

SCIENCE PERSON (resignedly) That does not make sense even at the level of simple logic.

POLITICAL PERSON (brightly) Then how can we make it make sense?

SCIENCE PERSON (slight eye roll) It might be helpful if we knew exactly why do we need to do this?

POLITICAL PERSON (still brightly) We need a way to express wind chill in a way that is more intuitive.

SCIENCE PERSON (incredulous) Expressing it as power per area is too abstract?

POLITICAL PERSON (seriously with firm eye contact) Yes.

SCIENCE PERSON (gazes at a point on the ceiling) OK then. Why do we need to associate this with a temperature? Why not create some sort of unit-less index? For example we could make zero denote the case where there is no windchill and five denote the case where frostbite is possible. Then the index would almost always be a single digit from zero to ten.

POLITICAL PERSON
Good. Good. Just thinking out loud here. What would happen if we made this index negative. You know, it gets smaller as the windchill increases. We could, say, make the index for the frostbyte windchill the same as the temperature that a person would get frostbyte if there was no wind. We could do that for all the other wind chill conditions.

SCIENCE PERSON (irritated) Very funny. The problem with your scheme is that wind chill does not work that way. If there is no wind then there are another set of variables that determine if someone gets frostbite.

POLITICAL PERSON
Great point. Still thinking out loud here... What if the person is walking?

SCIENCE PERSON (confused) What?

POLITICAL PERSON (excited) Yes! Most people do not just stand around outside in the winter. If they are walking then there will always be some wind! So can we do this?

SCIENCE PERSON (peevish) In theory yes but it would still not be a temperature.

POLITICAL PERSON
I can work with that. We will just leave the "C" off the index we provide to the public.

SCIENCE PERSON (desperate) Making a windchill index that looks like a temperature is being deliberately misleading.

POLITICAL PERSON (smug) Perhaps, but that is a judgement call. Report on my desk by Monday. Have a good weekend everyone!

OK, that is almost for sure not how things transpired but the result is accurate. In Canada the number given for the wind chill represents the temperature that would produce the same cooling for a person walking at 4.8 km/h (3 miles per hour) in calm conditions. The people involved even put people on a treadmill in a cold room with a fan. By measuring skin temperatures they came up with an impressive looking formula that takes temperature and wind speed as parameters and produces a negative number that looks like a temperature. This ended up being sciency enough that the United States has now adopted this formula for their weather reports.

Interestingly enough the hack involving the walking speed is no longer mentioned on the explanation page for the wind chill value on the Environment Canada web site. They now claim that it represents the "feels as cold as" temperature for the zero wind condition. ... which of course makes no sense. In time everyone will forget the thinking behind the formula and it will become a legend.

An interesting thing here is that people in places that actually experienced a significant amount of wind chill (Canadian Prairies) generally preferred the straightforward W/m^2 way of expressing windchill. My theory is that people in those places have a better intuitive understanding of the difference between temperature and wind chill. I personally still prefer the wind chill expressed as W/m^2 even after 9 years of the new system.

posted at: 19:44 | path: /politics | permanent link to this entry