A Few Late-Evening ICCC Items

Just a quick wrap-up, since I’ve done fairly well today with keeping up with things, particularly in the morning.

The two late-afternoon presentations I went to involved metrology and methods. They were marvelous.

Anthony Watts (of “Watts Up With That?” fame) gave a presentation on an unusual project he has been running. Either on his own or with the help of a few volunteers, he has been “auditing” all of the observation stations used by the National Weather Service to collect temperature data. So far, this audit has reached 75% of the reporting stations, and you can find more about the project here.

The most interesting part of this quest is the way that this effort has found all sorts of amazing (and surprising) environmental details that just make a mockery of basic metrology; you can find a partial summary here, and there are a variety of problems – stations in parking lots, stations near air conditioning vents, stations on rooftops (and whiskers on kittens), a station whose enclosure had been painted dark brown, etc.

While this was humorous in some ways, it raised a critical point – one I’ve been asking in different ways for a number of years. The whole “AGW” notion depends on parsing the measurability of temperature down to hundredths of a degree; yet the goofinesses in these stations would seem to introduce systematic errors on the order of multiple degrees. There simply is no way to pull a fraction-of-a-degree “global warming” signature out of such a completely messed-up system of measurement stations.

Anthony’s friend and colleague Steve McIntyre gave a more complicated talk examining the background of many of the data collection methods that were used. Stever is the gent who smoked out that infamous NASA data “Y2K” bug (sic?) that was (before he got them to fix it) adding 0.15 degrees C to all data from 2000 and later.

He noted that the Mann “hockey stick” arises from one single data collection method (western bristlecone pines), yet when all of the other non-hockey stick data is glommed together, the hockey stick appears. All of the data handling appears to disappear behind clouds of Principle Component Analysis (yes, I know what this is) and then strange things appear. He also noted excuses about difficulties with revisiting old data sites (the “hockey stick” was based on data that had nothing newer than the mid-1980s – it was an extrapolation from that) – due to difficulty moving “heavy equipment” to remote locations. He showed how his team went to Colorado and carried the “heavy equipment” (a hand-crank large-bit drill) up to the trees – many of which had first been measured in the original studies. He also noted that others have revisited the same tree sites for further analysis, and found no “hockey stick” stuff.

Well, the final event was a very important meeting. Your humble correspondent, Moe, blackhedd, Crank, and MarkI got together for dinner. We went to the Heartland Brewery (appropriate given that the Heartland Institute is sponsoring the conference), quaffed Brooklyn-brewed microbrews, chomped burgers, and plotted the counter-revolution.


Tomorrow is the final day, and is basically (mercifully) a half-day. But once again we start bright and early – with a 7am breakfast session that features former New Hampshire governor (and former WH CoS) John Sununu (who also holds a Ph.D. in Mechanical Engineering).