Judith Curry tries (unsuccessfully) to walk back some of the heat she’s unleashed on Richard Muller.
On her blog, Judith Curry writes,
… Showing preliminary results is of course fine, but overselling them at this point was a mistake IMO…
I also suggested a FAQ on [Muller’s] “end of skepticism” claim… “Our study addressed only one area of the concerns: was the temperature rise on land improperly affected by the four key biases (station quality, homogenization, urban heat island, and station selection)? The answer turned out to be no – but they were questions worthy of investigation. Berkeley Earth has not addressed issues of the tree ring and proxy data, climate model accuracy, or human attribution.” This is a reasonable statement, but comes across very differently from the WSJ editorial.
Curry also says,
And finally, this is NOT a new scandal. An important new data set has been released. Some new papers have been posted for comments, which are not surprisingly drawing criticism and controversy. The main issue seems to be Richard Muller’s public statements. All this does not constitute a new scientific scandal in any way.
The only way that Mullergate is not a new scandal is if one considers global warming alarmism to be one giant, ongoing scandal.
Dr. Curry is merely not trying to not let the abyss look into her. She got ticked off by Muller’s unprofessional and unscientific accusations and pronouncements and used some similarly unscientific (though, from the details I have seen, perhaps appropriate) words to describe his lack of quality control. Now that she’s calmed down, Dr Curry is trying to back off and approach it in a more rational and measured manner.
What she fails to mention also is that two of those tests (station quality and UHI) have been challenged on very sound grounds.
The station quality information only extends back to 1979, yet this study relates to temperature data to 1950. That means any statistical tests on the whole data set are invalid for considering station quality.
The urban-heat-island (UHI) study compared rises in rural and urban stations. Because rural stations showed slightly greater rise in temperature (on the edge of statistical significance) they concluded the UHI effect did not exist. However the whole point is that urbanisation has occurred in all but completely isolated locations, and few stations are in rural settings away from any buildings at all. The UHI effect is likely to be very close to logarithmic measured against overall development, so if we assume increased urbanisation at the same rate in all inhabited, commercial and industrial areas we would expect those that start out least developed to show the greatest increase in UHI effect.
In other words, if you assume that everywhere there is anyone has developed (not an assumption that can be summarily ruled out) then they have reversed the logic, and there proof there is no UHI effect is weak proof of UHI effect.
So while no good scientist would say they proved UHI effect, they have not ruled it out.
Of course the best test for UHI effect is the one that anyone with a modern car can do. Drive into the centre of London on a cloudy night, then repeat over a few nights to confirm. On cloudy nights the temperature changes slowly, so the increase in temperature seen must be from UHI effect.
The UHI effect is not very dynamic – it is a change in temperature with location not with time. It should be tested using a series of readings of temperatures at the same in different areas (wild, agricultural, dormitory, commercial and industrial of various intensities) over the period of a year. It should not be tested by variation over the years, due to the huge difficulty in documenting changes of urbanisation level over time given changes in land use, changes in technology for everything from heating to insulation, decline in manufacturing in the west … how do we define the change in the “Urban” bit of UHI?