Wednesday, November 29, 2006

Does Global Warming Stop when the Sun Goes Down?

So is the sun a bigger player in global warming than we think? We showed previously that when the maximum temperatures had increased (last 5 years), the increase appeared more so during the day only. Likewise when the yearly maximum temperature was less than that of norm, is was mainly during the heat of the day in which it occurred.

But lets look a little closer. Instead of summing up the data by looking at years, we decided to look at months. In fact a total of 83,000 of them. We decided to look at the months in particular which recorded a maximum temperature of say 5 degrees above the norm. For these months, what was the temperature like during the course of the day?

The graph to the left may look very confusing but let me explain. Basically, it shows the average deviation in temperature at certain times of the day, given that month showed a maximum temperature at a certain deviation from the average. Ouch, now that’s something to get your head around. Don’t stress, I shall explain. Lets look the top line, the bright red one. The legend on the right says that this is the average deviation from the norm at certain times of the day when the maximum for the month is 5 degrees above normal. In fact as we go down from 4 degrees above normal, to 3 to 2 and down to -5 we see a trend. The temperature is in fact becoming less. But this of course makes perfect sense.

A particularly hot month will of course have hotter times throughout the day. But what is interesting is that when looking at the graph, is that the greatest deviation occurs from between noon and 6pm when the maximum would occur. However, the deviation at say 3am-6am is nowhere as extreme when the sun goes down, especially for temperatures of negative monthly maximum deviations. Of course one could argue that the temperatures at night do not vary by as much, and hence it would be better looking at standard errors from the norm instead of temperature deviations and this is true, but this is just the first step in looking at time based data during the day.

This analysis alone proves that the current analysis being done by climate scientists based on purely maximum and minimum temperatures is hardly in depth enough – especially if we plan on spending billions of dollars on it. It still seems crazy to me, that the entire global warming science is based on analysis on simple statistics of which even this blog in a matter of several weeks has more than outdone the scientific literature on the matter.

What conclusion can we make to this research in this post? When analyzing a month that has an increased or decreased maximum temperature from the norm, the same deviation doesn’t occur throughout the day, and is significantly reduced when the sun is on the other side of the world.

6 comments:

Anonymous said...

Do you honestly believe this has not been done before?

Anonymous said...

Since the extrema occur at similar times of day, perhaps looking at maximum temperature, minimum temperature and temperature anomaly isn't so crazy after all...

What is interesting is that the variation in maximum anomaly is much higher than the variation in minimum temperature anomaly. What are the distributions of the (daily) temperature anomaly for max and min? Are they Gaussian? Symmetric? And have the distributions changed over the years?

There's currently very little meaning (from a climate science point-of-view) in the data you're presenting right now. Interesting as statistics for the sake of statistics, but the way temperature anomaly varies would be much more interesting. I think you've shown nicely that using maxima and minima is fine and the next step is to examine how they vary in a thorough way. If you could show some distributions...

Jonathan Lowe said...

thanks steve, this is only an initial look at the data as I have said, and george - yes - practically all temperature analysis is done solely on maximum and minimum temperatures.

Anonymous said...

Okay, so lets for a moment assume that you are the only person to have done this.

I think you graph is extremely confusing for one. I don't think it is clear what you are trying to show with it at all.

So for the 5deg line you are using all the months where the maximum temp in that month was 5 degs above the norm? Is that norm the monthly norm? the yearly norm or the entire data norm?

You need to make your point much clearer.


"A particularly hot month will of course have hotter times throughout the day. But what is interesting is that when looking at the graph, is that the greatest deviation occurs from between noon and 6pm when the maximum would occur. However, the deviation at say 3am-6am is nowhere as extreme when the sun goes down, especially for temperatures of negative monthly maximum deviations."


Are you trying to say that when its abnormally hot its not as abnormally hot at night?

While interesting, relating that to climate change would be tenuous

Jonathan Lowe said...

no worries george, this piece wasn't really trying to solve anything or state anything massive, it is really just a piece of interesting information that basically says as you say "Are you trying to say that when its abnormally hot its not as abnormally hot at night?"

Read the previous posts for more in depth information.

John M Reynolds said...

Jonathan, your blog needs an index. You invited George to read your previous posts, but that is not possible. Currently, I can only see 2 and a half posts. The Katrina post is cut part way through the word snowfall.

About this post, does it say that the variance from the norm is greater during the day than at night?