Mean and average temperature ?

[nitrx]

I saw cumulus use the term mean temperaure and avarage, mean is measured in two ways :

Exmaple: Summary for oktober 2010

Temperature (°C):
Mean (1 minute) 14,1
Mean (min+max) 14,7
Mean Minimum 9,9
etc

I couldn't find much about the way how an average temperature is measured, or if there is an official rule, I could find that the average temperature measured over a day in The Netherlands (official KNMI) is as follows:

each hour the temperature is measured so 24 times a day the sum is divided by 24 makes the avg for the day I think personally this isn't the right way but a couple of decades ago there was no way to measure the temprature in milli-nano or whatever-seconds. Anobody knows the 'official world-norm' ?

[steve]

I don't think there's a standard, it's just a question of how many samples you have available. I would imagine that for stations where manual readings are taken, the (min+max)/2 value is used, as the min and max are normally the only two readings you have each day. For an automatic weather station, of course, you have many more readings, so I chose 1 minute as a reasonable sample interval.

[nitrx]

I agree but if there's no standard how can they come to conclusions of global heating andso-on , for me the question is more curiosity and I want to see my values in a perspective to official meteodations.

[DaveNZ]

Our official Metservice in NZ use (min+max)/2. I tend to pay more attention to the 1 min average though.

[serowe]

The problem is you can only work with what you have.

AWS's can possibly provide minute-by-minute readings BUT most likely only report the normal 3 hourly figures. This is even more so with manual reporting stations - 9am and 3pm readings and nothing more (which makes an average from these stations more suspect as, in reality, they are only reporting an average figure for a 6 hour period).

It's all relative really - you believe what you want to believe (not putting anyone down here btw) - but you have to keep in mind WHEN the temperatures being reported were taken so they can be put into perspective.

[nitrx]

Thanks for the comments and your point of view

[beteljuice]

9am and 3pm readings and nothing more (which makes an average from these stations more suspect as .....

They do have Maxima / Minima thermometers - that's why the original / 2 rule applied !

[serowe]

9am and 3pm readings and nothing more (which makes an average from these stations more suspect as .....

They do have Maxima / Minima thermometers - that's why the original / 2 rule applied !

As I said though - you need to put the figures into perspective. A mean or average of Maximum and Minimum is a 24 hour average - it represents the daily mean/average. The mean/average of 9am and 3pm figures, in perspective, is more indiciative of a real 'day' temperature (except where the maximum emperatures are usually reached after 3pm daily!

It is like average summer or winter temperatures - theres are averages based on human defined boundaries.

[steve]

A mean or average of Maximum and Minimum is a 24 hour average - it represents the daily mean/average. The mean/average of 9am and 3pm figures, in perspective, is more indiciative of a real 'day' temperature (except where the maximum emperatures are usually reached after 3pm daily!

Beteljuice's point is that they won't be quoting a 24-hour average temperature based on the temperatures at 9am and 3pm, but on the minimum and maximum temperatures between those times (i.e. 9am to 3pm and 3pm to 9am). Or perhaps, like stations which are just read once a day at 9am, they just use the 9am to 9am min and max. It does give pretty good results.

[serowe]

I understand about that - just pointing out that HOW they actually arrive at these averages and means should be taken into account. And, from experience, the max/min thermometers weren't always highly reliable especially before the mid 1980's... More so wit4h stations that used the old mercury type compared to an analogue recorder style (paper disc)