cc: trenbert@ucar.edu, "Thomas C Peterson" , "Parker, David (Met Office)" date: Thu May 26 09:13:41 2005 from: Phil Jones subject: Re: FW: World's First Global Thermometer to: Mike MacCracken Mike, Thanks for emailing this. Not seen it. The original is junk science at its best ! It seems that they have little idea of the literature. I recall reviewing a paper on global temperatures in the mid-1980s, where a group at a US University (don't recall which) who just took the NCAR dataset and averaged all the surafce temps together. No anomalies, no allowance for coverage, no SSTs. Claimed there were massive jumps at certain times of several degrees ! This seems more of the same sort of awful science. HadCRU, NCDC and GISS all get figures out within a few weeks of the end of the month and all agree. A paper in submission to GRL saying this - I know it's all been said before. My main reason for replying though is that soon (hopefully) you'll be able to point out that satellites too show warming. Look out for the CCSP report on vertical temperature trends when it's available for open review sometime from late August. Also the hockey stick has been reproduced by Eugene Wahl and Caspar Amman (Alfred Uni and NCAR) and they go on to reproduce what McIntyre/McKitrick got wrong. They have two papers in submission and there are at least 3 others showing much the same. Ch 6 is likely to conclude much the same as the TAR wrt the last 1000 years. It also seems that borehole evidence has been reconciled. I reckon you'll have at lot more articles like this to respond to in the coming months, but a lot more material to work with. IPCC Chapters will be available for review from September. Cheers Phil At 19:31 25/05/2005, you wrote: Hi Phil--I hope I got the response right--and thought you might be interested in the original article, if you have not already seen it. Mike Subject: Re: World's First Global Thermometer Hi Jane--Well, JunkScience.com is just that--junk science. Certainly, he identifies a number of the challenges First, the scientific results are actually looking to estimate the "change' in global average surface temeprature and not its value, per se--he is right that a global average temperature is an artifact. Becasue most of the stations are likely in the Northern Hemisphere, it is likely that the summer month average temeprature he calculates will be greater than the winter average temperature--for the change in value, however, one needs to get the departure of the June temperature from the average June temperature, etc.--just the kind of seasonal adjustment done for the economic and job indicators (and that can take a good bit of time to work out accurately). Near as I can tell from his writeup, this is neglected even though it will surely be a much larger cycling than the amount of change we are looking for. In addition, each year is distinct, and so there will be variations from year-to-year, and not comparing to a long-term, carefully checked average is essential. To get at a valid estimate of the average of the local changes in surface temperature around the Earth, great care has to be taken to make sure one has good and self-consistent baseline temperatures to work with that are not contaminated by station moves, changes in instruments and measuring technique, development around stations, and many other factors that this JunkScience effort calls "statistical mysteries"--well, maybe they are mysteries to the writer, but they are necessary and why experts work so hard at it. In economics, the equivalent type of adjustment is, for example, adjustments across brands and models of cars or gasoline, etc. , and there are all sorts of other adjustments that it often takes them months to make sure they have right. This is all really necessary to be doing to get a reliable picture from the first estimate. Then, there is the issue of coverage--on land going from an irregular array of stations to a useful measure (so combine, carefully accounting for location and area represented--more "statistical mysteries" roughly equivalent to accounting in the inflation index for how much of each thing a person buys). There are, for example, a great many more stations in Europe than in Africa, so one does have to be careful in averaging, etc. Since the ocean covers about 70% of the Earth's surface, its changes do have to be included, and this has to also be done carefully using the right type of measurement to get an equivalently representative and consistent indication of temperature (and because ocean measurements come via different techniques, one has to be especially careful). Satellite estimates of surface temeprature, which are very good in a relative sense (one value compared to anohter) can be used to help interpolate between surface measurements that are more accurately done in a particular location, but an evaluation needs to be done about how they compare with what one might get elsewhere. So, what JunkScience has is a number that comes out in real-time--well, congratulations, but it is roughly equivalent on the stock market to having the average of the prices of all the stocks that trade any given day without accounting for how many shares of which stock traded that day, whether any splits occurred, and so on--this is just why there are normalized stock averages that are carefully controlled (those 'statistical mysteries"). Mike World's First Global Thermometer Friday, May 20, 2005 By Steven Milloy As the Northern Hemisphere enters the summer season and natural global warming occurs, itıs a good time to consider the concept of global temperature perhaps the most talked about, but least understood, component of the global warming controversy. Since 1988 when National Aeronautics and Space Administration (NASA) researcher James Hansen launched global warming alarmism with his congressional testimony that manmade emissions of greenhouse gases were warming the Earthıs atmosphere, global warming has been a hot topic. The controversy only heightened with the advent of the so-called ³hockey stick² graph that purports to show a dramatic rise in global temperature during the 20th century. At JunkScience.com, weıre trying to shed light on the problem of relying on global temperature as an indicator of global warming by developing and displaying the worldıs first (almost) real-time global thermometer. We gather temperature readings from about 1,000 surface-based temperature stations around the globe, calculating an average temperature, which we call the ³global mean temperature² (GMT). We use ³raw² temperature data that isnıt statistically massaged to account for seasonal variation or for the urban heat island effect the phenomenon caused by the heat-retaining properties of concrete and asphalt in urban areas that is known to artificially increase local temperatures. We display the current GMT and maintain old GMTs to track weekly, monthly and, eventually, annual trends. From what we can tell, our data track pretty well with the temperature estimates published by other climate researchers, which are available only weeks to months after the data are collected. At the time of this column, the GMT according to our calculations is roughly 62 degrees Fahrenheit. So what does that mean exactly? Weıre not really sure. First, global temperature is a contrived concept. There is no magical point in the Earthıs atmosphere to place a thermometer and take the planetıs temperature. Moreover, if you live in a polar or tropical region (or almost anywhere for that matter), a GMT of 62 degrees F is patently meaningless what matters is whatıs going on outside where you are. Our GMT is based on surface records. But if you look at a map of weather stations around the globe, youıll readily see the built-in bias of temperature readings from surface-based weather stations. The overwhelming majority of surface-based weather stations are land-based relatively few temperature readings come from ocean-based facilities, resulting in a major upward bias in available temperature data since about 75 percent of the Earthıs surface is water. An additional bias arises from the fact that there is more land mass and, therefore, more surface temperature stations in the Northern Hemisphere than in the Southern Hemisphere. Thereıs an even further bias introduced by the tendency of land-based weather stations to be located in more heavily populated areas, which are subject to the urban heat island effect. Relatively speaking, not many temperature readings come from the wilds of northern and central Asia or eastern Africa, for example. There are alternatives to the JunkScience.com-calculated GMT none, however, are available in real-time. The National Climactic Data Center collects temperature data from about 3,000 surface-based weather stations. But researchers often try to statistically adjust these data to account for the urban heat island effect, which produces results that are more statistical mysteries than true averages of global surface temperature readings. Other researchers calculate GMTs from data collected by satellites and weather balloons. These data measure atmospheric temperatures from all around the Earth and donıt suffer from the same biases as the surface temperature data. Itıs important to note that without the upward bias inherent to the surface temperature data, the satellite/balloon temperature measurements show no significant increase since data collection began 30 years ago. Global warming alarmism is largely based on the notion that global temperatures have increased since the 19th century industrial revolution due to manmade emissions of greenhouse gases, primarily carbon dioxide. The infamous hockey stick graph tries to dramatize the alleged increase in temperature by going back 1,000 years. But the pre-20th century GMTs in the hockey stick graph for the most part donıt come from thermometer readings. Instead they are guesstimates of GMTS based on geographically and temporally scattered data scavenged from tree rings, ice cores and other dubious proxies for thermometers. Whether calculated in real-time or two months after-the-fact, surface-based calculations of GMT are inherently and impossibly biased. In this light, the hockey stickıs GMTs over the last 1,000 years are near worthless yet it is this very data that are being used to drive global warming hysteria. We hope that the JunkScience.com global thermometer will help demystify the flawed science that has led to the present state of climate clamoring. Remember, just 30 years ago, early climate alarmists were actually fretting about global cooling. Itıs shocking that our government may commit us to potentially harmful energy and policies like the international global warming treaty known as the Kyoto Protocol or the legislation introduced in the U.S. Senate by Sen. John McCain, R-Ariz., and Sen. Joe Lieberman, D-Conn., based on such an elusive, if not meaningless, concept as global temperature. Steven Milloy publishes JunkScience.com and CSRwatch.com, is adjunct scholar at the Competitive Enterprise Institute, and is the author of Junk Science Judo: Self-defense Against Health Scares and Scams (Cato Institute, 2001). Prof. Phil Jones Climatic Research Unit Telephone +44 (0) 1603 592090 School of Environmental Sciences Fax +44 (0) 1603 507784 University of East Anglia Norwich Email p.jones@uea.ac.uk NR4 7TJ UK ----------------------------------------------------------------------------