Repost: Putting the "Urban Heat Islands" issue to bed
Preface on reposting policy I and others have been discussing concerns about the ephemerality of blogging. Given that blogs are a searchable database, there’s no real reason why people should look only at the current pages. But in my experience, comment threads tend to die off after a few days. However, Aaron Oakley has just put in a comment on a post from December 2002, and this gives me a chance to announce my new policy. If anyone comments on a post that has been archived (more than 10 days) I will do my best to repost it and thereby reopen the debate.
Preface to the repost Now here’s the reposted piece. I think the article quoted in the paper refutes Aaron in advance, but just to be clear I’ll restate my point. I agree with Aaron that urban heat islands are real even in small towns (in fact, I’m glad to see Aaron endorsing the reality of human-induced climate change). I also agree that estimates of climate change need to be checked using only rural stations. But, as the cited article says this has already been done, and it makes no significant difference to estimates of global warming. Note that Aaron himself recommended this article.
Reposted article begins
Bizarre Science points to this study confirming the IPCC contention that Urban Heat Islands, while a real phenomenon, are not important in assessing estimates of the rate of global warming C.J.G. (Jon) Morris of the School of Earth Sciences, The University of Melbourne, reports
Whilst climatologists now think that the warming in the temperature record from some small urban areas is partly the result of the UHI, this is not evidence that Australia’s climate has remained unchanged rather than warmed over the past 100 years. Average minimum temperatures from many stations over most of Australia have shown an increase of between 0.1 deg C and 0.3 deg C per decade since 1951. Whilst some temperature records from small towns do not represent the large scale climate, it is unlikely to have any major impact upon our estimates of temperature warming over Australia. This is because there are numerious other weather stations located in remote areas such as lighthouses and regions far removed from urban areas that still indicate a warming temperature trend.
Thanks for this useful link!
Update While I’m at it, I also appreciate this post, in which BS reader Reader George Bogg points out that, given the number of points in Alan McCallum’s scatterplot, the trend he finds is almost certainly statistically significant (thereby resolving the main remaining point of dispute). However, Bogg misses the point that this is a panel data set, consisting of observations from many different stations over time. An analysis taking this into account would yield a much stronger correlation. Of course, as Bogg points out the fact that the world is getting hotter doesn’t prove anything about the causes. But at least agreement on the facts is a start.