Back in May, an article appeared in the Opinions section of the journal Nature – “How best to log local temperatures?” (Nature 465, 158-159 (13 May 2010) )
Over at WUWT today Anthony Watts has posted Roger Pielke’s Grave Concerns on the paper – the misleading claim by the authors that the data sets are independent when they are not. It is worth going further than that in examining what is said. This is a call for greater effort and more funding for climate change research.
“The climate community must work together to create a single, clean, comprehensive and open repository of detailed temperature data, say Peter A. Stott and Peter W. Thorne.”
Now what is interesting is that the authors clearly consider the current datasets, which “only have resolutions of hundreds of kilometres”, imperfect.
“In many cases a weather station’s instrumentation,
location, time of observation, or environment
(such as proximity to buildings) has
changed. While such changes aren’t of concern
for weather forecasting, which deals with day to-
day temperature swings of several degrees,
they are relevant for monitoring long-term
decadal changes, where fractions of a degree
make a difference. Unfortunately, the timing
and effect of most of these changes have not
But they are not suggesting that the current data sets are the ones that should be improved. They propose to move on, with a much more detailed data set, promising to be able to predict the effects of climate change at local levels – how it would affect transport, agriculture etc., using suitable scares such as severity of heatwaves or droughts and delays to monsoon rains.
“This new challenge requires climate data sets
at daily or even shorter timescales, at a resolution
of a few kilometres at most, to monitor
how weather is changing and to feed local climate
They propose and invite “anyone” to create rules and algorithms for correcting the new databank. And, it seems, this is a WMO-backed initiative, the outcomes of which would be sanctioned by the WMO-backed project though publication in peer-reviewed journals to “provide transparency and an audit trail”. By creating such a suite of independent data sets, apparently, it will become possible to assess the sensitivity of the data to different sorts of corrections.
You what? Not content with the mess the models make currently, they want to work at an even greater level of detail? Hmm. Create the need; justify funding to do the research.