Monday, August 06, 2007

Surface Temperature

There's been quite a little storm blown up by the project, which aim to document photographically the compliance (or non-compliance) of USHCN climate statements. That an amateur, a retired weatherman, has been able to so easily find sites near asphalt, trees, buildings (even on buildings!) is disturbing. All these microsite issues go against the climate scientists' (such as NOAA) own guidelines for siting. One would think this would be a straitghtforward issue -- if data problems are identified, fix them. The scientists already have all the site data. It would not be a large effort to remove "bad" sites and re-run the data analysis. Of course, until all 2000 sites are documented we won't know the extent of problem sites; it may very well be insignificant.

Yet many global warming proponents are taking a different tack. They are defending the sites and claiming that nearby A/C units, asphalt, and buildings have no significant effect on temperature measurement. Somewhat bizarre given that we're looking for 1/10 degree per decade trends, which works out to 1/100 degree per year. When you're looking for this level of "signal" even a half a degree of "noise" can have a major effect.

This is how science is supposed to be done: Declare an experimental protocol. Follow it, and report on the results. If the protocol was not followed then disregard the results and redo the experiment again.

This is what some AGW proponents are saying: Declare an experimental protocol. Don't follow it, and then use the results anyway.

It's logically inconsistent to say on the one hand that we (the public) should listen to and accept the findings of the "climate professionals", but on the other hand we don't need to listen to the same scientists when they say data quality requires strict siting guidelines.

What's really strange is that it's unlikely to change much. 20th century warming is pretty clearly shown in many ways. There is sea temperature data showing rising temps, and (the gold standard) satellite data showing the same (since 1980). Even if finds 25% of sites are "bad", the remaining sites will most likely still show warming. They may show less warming, and in particular may show current warming is no larger than the peak around 1930. That would make things interesting, but it's not about to disprove global warming.