I've written about this disability before - here and here, which featured Bob Tisdale too.
When I read some of the comments to my last article it prompted me to go and re-read Tamino's article as well as Bob's article. Veritas6053 asked a question of another commenter and answered it:
why are you not asking Tamino the same question when he detrended the data?So I went back to Tamino's article to see why he detrended the data. It was for the same reason as Bob gave in his article (archived here). They both detrended the data to determine whether or not 1990 was especially hot.
it looks as though Tisdale used detrended data because Tamino used detrended data.
btw, detrending is done all of the time. NOAA's AMO index is detrended N. Atlantic SSTa data.
Here is Tamino's chart using just GISTemp data. He's circled 1990 in red:
|Source: Tamino's Open Mind|
|Source: Bob Tisdale at WUWT|
Both charts show that 1990 lay above the trendline. That's not exceptional of itself. One would expect that roughly half the data points would be above the trendline.
However Bob's chart shows that 1990 was fairly hot, even after removing the trend. In Bob's chart there were only eight years where the anomaly was higher, out of the 34 data points he showed.
Tamino's chart also shows 1990 was hot even when the trend is removed. Tamino stated that 1990 was "10th-hottest departure from the smooth trend in the 133-year data record from NASA".
Here again is the chart I originally created when Bob first claimed that 1990 was not an especially hot year:
|Data Source: NASA and WUWT!|
In other words, all charts show that 1990 was a very hot year for the time.
My chart above shows that 1990 was at the time an especially hot year in absolute terms. It was the hottest year on record at the time and the 1990 record didn't get broken for another five years.
The charts from Tamino and Bob Tisdale show that it was also an especially hot year even if the warming trend is removed - about equal 9th highest above the trend since 1979 using five data sets (Bob Tisdale) and tenth highest above the trend in 133 years of data (Tamino).
And since Bob Tisdale made much of the fact that there was no ENSO event that year, 1990 can be considered an especially hot year with no help from any El Nino.
Here is how Bob explains his stance in response to someone who pointed out the obvious - excerpt with my bold italics (archived here):
We appear to have different definitions of “especially hot”, joeldshore. Looking at my Figure 8, there are 19 years “hotter than the trendline”. The average positive deviation of those 19 years is 0.078 deg C. For 1990, the difference was 0.068 deg C. Being below the average of the years with the positive deviations, doesn’t make 1990 “especially hot” in my book. If you’re looking for an example of “especially hot”, that would be 1998. October 28, 2013 at 1:04 amBob notes that there were 19 years where the temperature was above his trendline. This is roughly to be expected, given there are 34 years in his series. Yet he is still trying to argue that despite the fact that 1990 was not just above his trendline but exceeded by only eight other data points of those 34 years, it wasn't "especially" hot.
Bob's argument is that using his averaged data it was "below the average of the years with the positive deviations". But look at his series again. There is one year that dragged up that average by a heap. That was, you guessed it, 1998. In fact, I think it would be fair to conclude that in Bob's mind, only the very hottest year was "especially hot"! And this was eight years after 1990 so really, should it even be considered?
Now all that is leaving aside that the world is warming. The trend matters! It's getting hotter. 1990 was at the time the hottest year on record and retained that distinction for another five years. I think Bob Tisdale would like to ignore the trend altogether.
I've got to say that IMO Bob's mammoth and repeated efforts to deny the fact that 1990 was hot seem pathetic, to say the least.
Additional observation: Bob's argument is not dissimilar to fake sceptics arguing that if it's happened before it's not extreme. A record doesn't have to be broken to rate as "extreme". All that needs to happen is for the data to be at close to the upper or lower extremities of a range. Of course if a record is broken it would normally be considered extreme. Like when 1990 broke all records. It was not just especially hot, it was an extremely hot year!