For the past decade or so, each January has produced a news headline about whether or not this year has been the hottest ever. It seems to be agreed that 2016 was the hottest ever, but with a statistically insignificant increase over 2015, and only a tiny bit above 1998. In 2015, 2016 and 1998 the spike was due to an el Nino, which subsided quickly. As I have argued before, there is no human being who has ever experienced a global average temperature, unless coincidentally, and for a moment or two. What we want to know is what our own environment has been like, if such weather details are of interest at all.
I have recommended before a most useful weather-data website, climate4you, conducted by a Norwegian scientist, Professor Ole Humlum. What he does is to assemble all the data from all the major climate dataset compilers, and provide a month-by-month summary. Since the focus at this time of the year is always on temperature, I thought it would be useful to look at what his data sources say. Humlum grades the datasets by reliability, and puts the two satellite datasets, UAH-MSU and RSS into his top category. They show December 2016 as much the same as the average up to the beginning of the 2015/2016 el Nino spike (UAH), and about where average temperature was in 1997 (RSS). Humlum uses the 1979-present period as his base, arguing that (i) 1979 was the beginning of the last strong warming period, and (ii) it is also the year when proper global data from satellites appear. He doesn’t like the older datasets because their controllers make too many ‘administrative changes’. What he has in mind is shown in the next graph.
The size of the changes, 0.7 degrees C over ten years is considerable, as are the actual number of changes, most of them slight. Fiddling with the past is all too reminiscent of Orwell’s 1984. Maybe there are good reasons, and maybe if I did a lot of work I would find them. Humlum’s point is that every time a datum is changed, the trend lines from the past are altered too. What is one to think? He argues that these changes do not seem to appear in the satellite data. though there is a new version of UAH-MSU which is eventually to supplant the original one. The changes there seem small to me. Incidentally he notes that since 2003 the average global surface air temperature has steadily drifted away in a positive direction from the average satellite temperature. He doesn’t know why, but for the moment puts the shift down to those ‘administrative changes’.
Anyway, the whole website is most interesting, and is supplemented by discussion. Just about everything you might want to find is there, with contributions, where they exist, from every dataset. It is my go-to source for weather data.
One aspect of global weather which I find particularly valuable on his website is the mapping. What exactly happened to Australia in terms of temperature over 1916? Well. you can find out by looking at Australia on his global map each month. The data come from the Goddard Institute for Space Science (GISS). Below is the world in October 2016. The colours have numerical referents in the bar on the right-hand side. We need to remember that large areas of Australia and much of the rest of the world, especially the ocean expanse, have no thermometers, so the colours in part represent extrapolations from real instruments elsewhere.
What is arresting about this image is the variation across the globe. In October, on average, Central Asia was having a most unpleasant cold spell and Australia was cooler than the ten-year average. So was Canada, but below the border most of the USA was warmer. Yes, of course you can determine an average for the whole world, but surely the spatial variation here is what we really want to know about. How has it been for us? You can go to such a map for each month, and here is my verbal summary for Australia.
January: a little cooler than the ten-year average
February: a little warmer in the northeast; a little cooler in the southwest
March: a little warmer
April: a little warmer
May: a little cooler in the southwest; in the far north, warmer; the rest, a little warmer
June: a little cooler in the southwest; a little warmer in the north; the rest, same as the ten-year average
July: cooler in the southwest, warmer in the northwest and east
August: cooler everywhere
September: cooler everywhere
October: cooler everywhere
November: cooler everywhere
December: a little cooler in the west, a little warmer in the east
On the face of it, there can’t have been much change from 2015 to 2106, and while we await the 2016 summary in climate4you, our BoM has suggested that Australia had its fourth warmest year since 1910, the year with which the BoM starts its own records. Maybe so. None of this should get anyone’s knickers in a twist, since the change is small, and much of it due to the el Nino. As it happens, no la Nina has started yet, but it would be somewhat unlikely if a further el Nino arrived before the next la Nina episode. (For those for whom these are new terms, el Nino and la Nina are weather episodes produced by shifts in water in the Pacific, the former bringing hot and relatively dry weather to eastern Australia, the latter cooler and wetter weather. They are aspects of the El Nino Southern Oscillation.)
One aspect of climate that is missing from the Humlum website is sea level. But there is a fine new website that seems like to fill the gap and, like Professor Humlum, the originator has assembled all the data from the official sources. The website is SeaLevel.info, and it is great fun too. I am only at the beginning of learning how to operate it. The originator, Dave Burton, was an expert reviewer for AR5 WG1, and he credits the well-known site WoodForTrees as the inspiration for his own work. Here is a summary of his views on the whole sea-level brouhaha.
The worst effect of anthropogenic climate change is supposed to be accelerated sea-level rise. But that fear is the product of superstition, not science. The measurements show that anthropogenic GHG emissions have had no detectable effect on the rate of sea-level rise. At some coastal locations, sea-level is rising, and at other locations it is falling, because of vertical land motion. The global average is slightly rising, but only about 1.5 mm/year (six inches per century), and the globally averaged rate of sea-level rise is no greater now, with CO2 over 400 ppmv, than it was 85 years ago, with CO2 under 310 ppmv.
This is a 111 year record of sea-level measurements at one particular location in the Pacific, but it is perfectly typical. The blue trace is sea-level, the green trace is CO2. If you know how to read graphs, then it will be obvious that CO2 is not affecting sea-level:
As time goes on we are likely to see more and more of these datasets that are not presented by government. That is a good thing. I will return to these issues when Professor Humlum, to whom we owe a great debt, has issued his 2016 summary map, and on sea levels, when I have mastered the website!
Postscript: The great virtues of homogenisation have been praised again, so I thought I might just cross-post something I saw this morning. No further comment is needed, from me at any rate.
‘The raw data that is fed to NASA in order to develop the global temperature series is subjected to “homogenization” to ensure that it does not suffer from such things as the changes in the method of measuring the mean temperature, or changes in readings because of changes in location. However, while the process is supposed to be supported by metadata – i.e. the homogenizers are supposed to provide the basis for any modification of the raw data.
For example, the raw data for my home city, Cape Town, goes back to 1880:
The warmest years were in the 1930’s, as they were in many other parts of the globe. There was then a fairly steep decline into the 1970’s before the temperature recovered to today’s levels, close to the hottest years of the 1930’s.
In NASA’s hands, the data pre-1909 was discarded; the 1910 to 1939 data was adjusted downwards by 1.1deg C; the 1940 to 1959 data was adjusted downwards by about 0.8 deg C on average; the 1969 to 1995 data was adjusted upwards by about 0.2 deg C, with the end result that GISS Ver 2 was:-
Being curious, I asked for the metadata. Eventually I got a single line, most of which was obvious, latitude, longitude, height above mean sea level, followed by four or five alphanumerics. This was no basis for the “adjustments” to the raw data.
Which should I believe? The raw data showed a marked drop from the 1940’s to the 1970’s, which echoed similar drops elsewhere… The raw data is probably accurate. The homogenized data is certainly not. It is difficult to avoid the conclusion that “homogenization” means “revise the story line” and “anthropogenic global warming” really means “humans changed the figures”.
Prof Philip Lloyd, Energy Institute, CPUT, SARETEC, Sacks Circle, Bellville
Second postscript: In response to Nga, others have noticed that the link does’t work, but it didn’t take long for me to find a better version, shown below. And if anyone wants to check this one (from the GISS website), they can do what I did. Go the the WUWT version, read the comments and follow the links.