Friday, May 23, 2008

Telling Tall Tales With Trends

SEARCH BLOG: GLOBAL WARMING

We've all heard the case made for global warming: the average temperature over the past 100+ years has had an upward trend of 0.6°C.
In actuality, when you look at the data you see a starting point at a relatively low temperature [and starting points are important for trends]. Then you see a few decades going up, then down, then up, then down, then up... and, most recently, down. There may be a trend in there, but the jury is out about how temperatures have been taken during all of that time. Is it one long trend or a cyclical occurrence of short-term trends?

[illustrative data only]

What is the trend of record high state temperatures?


I could make the case that late April is going to be the warmest period in Michigan this year... based on a trend... with a high starting point [oh, those inconvenient starting points]. Just take a look at how May is progressing.

See that gold line? That's the average temperature trend for the month. I'm just going to put that in my model and see where August is going to be. Get the ice skates sharpened.

Just how long do we have to look at temperatures until we see "climate change?" I guess that's dependent upon how you define climate change. Geologically, swings of 5-10°C probably qualify as significant climate changes. In our human experience, we like to narrow that somewhat... maybe 3-5°C... because that's something that affects our comfort levels... but not necessarily our existence.
Then, if you get an 0.6°C change over a century... starting from a cold point... does that mean anything?
Did we just get climate change in May? The temperatures were supposed to go up and they went down!
You'll have to excuse me now, I have to go play with my models... data models!
..