Lucky, thanks to the BOM, that we have such high quality data to understand the Australian climate. Without it, we would never have found out that the hottest day ever recorded in Australia appears to be the 8th of February, 1933 in, wait for it… Albany, in the far cool south of Western Australia.
Chris Gillham emails:
“Guess where and when was the hottest day ever recorded in Australia? 50.7C at Oodnadatta on 2 Jan 1960? No way! Mardie at 50.5C on 19 Feb 1998? Get out of here! It was at Albany on 8 February 1933, that historic day when this normally chilly town on WA’s southern coast was razed with a temperature of 51.2C. Don’t believe me? You can’t question the accuracy of ACORN, a temperature network that shapes economic policy, and the screenshot from last night’s ACORN download for Albany max proves it …”
The all new marvellous ACORN dataset has been “expert peer reviewed”, it “employs the latest analysis techniques “ and it is “a complete re-analysis of the Australian homogenised temperature database”. Phew. The BOM lists their hottest days here, but they don’t seem to have noticed the hotter than hot Albany record yet.
It gives you a lot of confidence in the accuracy of all the other daily temps, doesn’t it? There doesn’t appear to be any sinister design here, just plain incompetence.
Australians need a full independent audit on our temperature records.
Here’s the month of February in Albany in 1933 according to ACORN.
DATE Temp (c) 19330201 25.5 19330202 29.6 19330203 30.4 19330204 35.2 19330205 23.0 19330206 33.3 19330207 32.8 19330208 51.2 19330209 38.8 19330210 30.8 19330211 25.5 19330212 24.2 19330213 27.0 19330214 30.8 19330215 28.9 19330216 28.2 19330217 28.0 19330218 26.5 19330219 27.0 19330220 29.6 19330221 28.5 19330222 23.1 19330223 23.9 19330224 25.0 19330225 28.5 19330226 25.2 19330227 23.0 19330228 23.2
The original raw data for Albany shows that Feb 8th was closer to 44.8C, still a very hot day. But as Chris Gillham points out, bumping just one day up to 51.2 changed the monthly average by 0.2C. How many times have other errors shifted averages and we don’t know about it?
“You wouldn’t expect one incorrect day to influence overall monthly or annual trends but February 1933 in Albany with current temps including the 51.2C was 28.8C in ACORN. February 8 1933 was actually 44.8C according to CDO raw at Albany 9500, the hottest February day ever recorded in the town. If that’s substituted, February 1933 had a monthly average of 28.6C. A 0.2C shift is significant from a single day error. If you check the ACORN monthly averages for Albany 9741, Feb 1933 has an average of 28.8C, so the 51.2C day has infiltrated and corrupted all ACORN calculations. Feb 1933 averaged 25.8C in CDO raw 9500 and 28.8C in ACORN 9741 (which starts 1965 in CDO).”
The ACORN dataset is supposedly double checked and expert yet no one has done the most basic of quality control. You’d think experts would have put in some basic checks and error messages. eg IF TMax > Highest on record THEN look again!
Thanks to Chris Gillham for doing the checking that our paid experts forgot to do. His site is waclimate.net