For many years now, human-caused climate change has been viewed as a large and urgent problem. In truth, however, the biggest part of the problem is neither environmental nor scientific, but a self-created political fiasco. Consider the simple fact, drawn from the official temperature records of the Climate Research Unit at the University of East Anglia, that for the years 1998-2005 global average temperature did not increase (there was actually a slight decrease, though not at a rate that differs significantly from zero).
Is he right? Well, take a look at the chart and see for yourself.
His raw claim is (more or less) true. The average temperature for the years 1998-2005 were more or less contant. Is this a reason to believe that global warming has stopped? Absolutely not. To see why, ignore the black line and look just at the raw data (the blue and red bars). As you can see, there is a lot of random noise on top of the signal. Some years the temperature goes up and in other years it goes down. (On two occasions in the last 20 years global temperatures went down two years in a row!) But the general trend is pretty clearly up. Can we actually quantify this so that it's not just a gestalt assessment? Yes, we can.
The tool that science uses to pull a signal out of noisy data like global temperatures is called statistics. The math can get pretty hairy, but the fundamental idea is very simple. The method works like this:
1. Pick some assumptions. (These are called the null hypothesis conditions.)
2. Figure out the probability distribution of some property of the data if those assumptions are true.
3. Compute the probability of the data that you actually observed. If the probability is low then there are only two possibilities: either a very unlikely event happened, or your assumptions are false.
Let's see how we can apply this to the global temperature data.
1. Let us assume that the earth is not warming up.
2. If that is true, then the probability distribution of temperatures should be a normal distribution around the average. In particular, we should see more or less the same number of above average temperatures as below-average temperatures. (We should also see more or less the same number of increases and decreases. There are many different properties of the data we could choose.)
3. When we look at the data we see that over the last 25 years the data are not evenly distributed between above and below average. In fact, every one of the last 25 years has been above average. The probability that this would happen merely by chance (on the assumption that there is no global warming) is 1 in 2 to the 25th power, or about one in 33 million.
So there are only two possibilities: either our assumption is wrong, or we've just happened to hit upon an extremely unlikely set of events.
In fact, the evidence is even more compelling than that. If you take as your baseline temperature the average before 1900, then every year since 1940 has been above average. The odds of that happening in the absence of real underlying global warming is about one in 10,000,000,000,000,000,000 (more or less).
You can do a similar analysis using an assumption that there is global warming and figuring out the probability of hitting an eight-year long spell of more or less constant temperatures. The probability depends on exactly what your assumptions are (mainly the rate of the underlying warming and the magnitude of the noise) and I don't have time to actually do the math at the moment, but it almost certainly will turn out to be a fairly common event. There are actually several multi-year periods in the recent past with no apparent temperature change (e.g. 1975-85, 1988-92).
So Carter's basic observation is correct, but his conclusion is absolutely wrong. In fact, it is so wrong that I wonder how he ever managed to get his Ph.D., let alone a faculty position. His mistake is so fundamental that it is hard to put it in any kind of favorable light. Carter is either disingenuous, or he is ignorant of basic scientific principles. I can't think of any other possibility.
No comments:
Post a Comment