Posts Tagged ‘weather’

Proof of concept

Friday, December 4th, 2009

I’ve downloaded the data for station 723150-03812, Asheville Municipal Airport in North Carolina. This is the one the NCDC/NOAA use for their sample data and is, I guess, their local airport.

Throwing the full 1948-2009 data – about 26000 records - into an Excel spreadsheet, using the SUMIF and COUNTIF functions to pull out the relevant days, and sorting the results shows that the ten warmest years in Asheville are (from lowest to highest): 1974, 1980, 2001, 1999, 2002, 1991, 2007, 2009, 1998, 1990

2009 is an aberration as there are a bunch of cold days to come before the end of the year, but it’s clear that there have been equally warm years in recent decades.

This surprised me, I expected at least eight or nine out of the ten to be the warmest if the effect were clear-cut – particularly with all the fuss people have made about airport locations for weather stations, the El Nino in ’98, and the satellite data showing warming throughout the 1990s.

It remains to be seen how typical (or not) Asheville is…

Proposition 02: The warmest ten years

Friday, December 4th, 2009

I was reading the Evening Standard on my way home tonight, and saw the Rt Hon Ed Miliband MP’s op-ed – “Climate change sceptics are today’s flat-earth brigade”. In it he mentions a line I’ve heard numerous times before.

Proposition:

The 10 warmest years on record have all occurred since 1997

Methodology:

It seems to me that this statement is either true or false, and can be checked using temperature records from weather stations – the sort of data I’m collecting already. For any particular station, the statement is either true or false. For the planet as a whole, a clear majority of stations should ‘vote’ for the statement being true.

  • Collect daily station records from the NOAA here
  • Exclude stations which don’t include the years 1985 – 2008 (12 years either side of 1997 – it’s no use using the station if the test is necessarily true because it’s only post-1997 or necessarily false because there aren’t ten years after 1997 to be the warmest)
  • Exclude stations without records for at least 90% of days
  • Calculate the annual mean temperature for each station in each year
  • Sort station data by annual mean temperature to find the ten warmest years, see if they’re all after 1997 or not

I expect there’ll be some stations for which it’s true and some for which it’s false – there will always be random weather effects – but if recent years are the warmest this should be clear from the measurements. It would be strange to imagine a warm year which left most places colder!