I look at this and first think… wow that is cold. Then I look at it again and wonder “how would I figure out an average temperature” of the US and Canada from a series of maps like this. How would I figure out if it was getting warmer or colder. It isn’t obvious and lots of problems immediately crop up.
One example being… it was 21 degrees when I woke up this morning, and it is now 46 degrees. Which temperature should I use? Should I average the temperature taken each hour (minute, second) over 24 hours and use it for the day? Another, is I have two weather stations at my house… one said 21, the other 28. Two very accurate digital yielded very different readings 50 yards apart. I have two because I updated to a new one and never took down the old one.
This is why I don’t trust any climate science paper that will not share data and exact methods. The methods, instruments, station location, and what data is used can make a big difference!
January 4th, 2010 at 7:21 pm
The map is wind chill factor! If you try to get an average temperature from that even if you are a genius with a supercomputer, I still wouldn’t believe it.
January 4th, 2010 at 9:13 pm
If two thermometers 50 yards apart measure temperatures seven degrees apart, look either for different conditions at the two points or suspect one is not accurate. Wind chill does not affect thermometers; wind chill is about loss of heat from the human body. Solar radiation does affect measurements. Also remember the law of large numbers, that the larger the sample size, the closer the sample mean will be to the true mean. Even if the individual thermometers have a decent error band, unless there is a systemic bias in the measurements, the sample mean will approach the true mean. As for the intra-day variation, I’m sure the data processing takes that into account since the mechanism is well known. It could for example use only the maximum and minimum temperatures of each day.
If one doesn’t know how all those temperature measurements are processed into an “average,” and how that average is used to judge overall secular trends, then perhaps it’s safest not to have too firm an opinion about the accuracy of trend assessments.
January 4th, 2010 at 10:16 pm
I know why the thermometers are different. One is in an ideal spot to not be affected by shade or house effects. The other isn’t.
I do know a fair amount about how the land temperature measurements are done, and they are way to subjective to be trusted. The stations are inaccurate, their sites not suitable for consistent measurement, the station locations change, the stations have outages and other incomplete data. There is no way it can be accurate.
I’m have more confidence in satellite data, for instance, which I consider to be accurate and less subject to scientist bias. It also is consistently made public because it is publicly funded.