date: Wed, 3 Jun 2009 15:07:25 +0100 from: "Parker, David" subject: RE: Tom's thoughts on urban errors ... to: "Thorne, Peter (Climate Research)" , "Phil Jones" , "Stott, Peter" Peter Thanks. I agree we should ask Tom for the actual numeric metadata or its on-line address. David David Parker, Climate Research scientist Met Office Hadley Centre FitzRoy Road Exeter Devon EX1 3PB United Kingdom Tel: +44 (0)1392 886649 Fax: +44 (0)1392 885681 Email: david.parker@metoffice.gov.uk Website: www.metoffice.gov.uk See our guide to climate change at http://www.metoffice.gov.uk/climatechange/guide/ -----Original Message----- From: Thorne, Peter (Climate Research) Sent: Wednesday, June 03, 2009 2:34 PM To: Phil Jones; Parker, David; Stott, Peter Subject: Tom's thoughts on urban errors ... Hi, Peter, The attached paper describes the rural/urban metadata we have for most stations in GHCN (at least as of 1999). We should be able to easily provide you with both the map and night lights metadata, as I believe they are on-line with GHCN. My personal view is: I heard the world's foremost urban meteorology expert, Tim Oke (Canada), say that urbanization takes place on three scales: micro (on the scale of a garden), local (on a scale of a park) and meso (on the scale of the city). Of the three, he said the mirco is most important and the meso is least. I have seen a lot of lousy urban heat analysis, such as papers which looked at the temperature difference between rural and urban but used the temperature difference itself as the metric for which was the most urban and most rural. Of the analyses that seemed robust, they tend to find little or no uhi bias in homogeneity adjusted data. My view is that this is because the adjustments to account for sudden changes in the micro and local scales which, by the very nature of comparison to neighboring stations, adjusts for subtle meso scale biases. Everybody wants to add an estimate of what UHI bias might be into their error bars, but it seems to me that rather than trust folk lore that there is a uhi bias, they first need to find one systematically in the network. Until they do that, the former is just hand waving to appease the know-littles. Jim Hansen adjusts his urban stations (based on night-lights) to nearby rural stations, but if I recall correctly (I'll send that paper shortly), he warms the trend in 42 percent of the urban stations indicating that nearly half have an urban cold bias. Yet error analyzers want to add a one sided extra error bar for uhi..... Regards, Tom -- Peter Thorne Climate Research Scientist Met Office Hadley Centre, FitzRoy Road, Exeter, EX1 3PB tel. +44 1392 886552 fax +44 1392 885681 www.metoffice.gov.uk/hadobs