Erskine Childers’ “Riddle of the Sands”
Recently, every time I have looked at the GISS Station Data, I have found something different, a subtle change. It would be surprising if this were not the case, due to the process of continuous updating, correction and improvement. Maybe my brain is just a bit strange in retaining such detail that I see the small changes – a bit like some people seem to remember phone numbers effortlessly, or are able to recite Pi to 25 decimal places.
First there was the big change to the US data in November (2009) – using USHCNv2, and now since January the use of satellite-observed nightlight radiance across the world, not just in the contiguous United States. According to Hansen et al. in a forthcoming publication
“We use satellite nightlight measurements to identify measurement stations located in extreme darkness. Independently, we expect nightlight intensity to be a better indication than population of urban heat generation.”
“Based on the 1900-2009 period, that change reduces it by about 0.005 °C per century.”
The nightlights thing is potentially an important change and has huge implications for the data, those changes I’ve already noticed, and many more to come. The paper goes on to ask:
“….is it possible that our ‘rural’ stations themselves contain substantial human-made warming?”
It goes on to explain how this was examined in the US is –
“…a good place to search for greater urban effects, because of its high energy use and a consequent expectation of large urban effects.”
“The standard nightlight adjustment removes slightly more warming than does the population adjustment. Theimportant conclusion is that the strict nightlight adjustment has no significant additional effect, compared with the standard nightlight adjustment.”
OK, but what about the rest of the world? I guess that will take time to work through and we’re only seeing the start of the changes.
First off, I noticed a few weeks ago that the three stations in Guam are no longer available in the GISS adjusted version of the data. The image on the left from Google Earth with a Earth City Lights overlay shows Guam. This apparently has better spatial resolution than the TIFF file used by GISS. From the GISS v2.inv file, the three Guam stations have radiance values as follows:
- Nwso Agana G – 43
- Wsmo Agana G – 23
- Andersen Afb – 22
Saipan (radiance 18) just to the north of Guam, follows the same fate.
At the moment the sites are still labelled as rural (in the v2.inv file), but seem now to be treated as urban or peri-urban. This has consequences for their requirement for adjustment.
“If an urban station has at least three rural stations within 500 km, all of these are used for the adjustment with closer stations receiving greater weight, as described above. If there are not three stations within 500 km, but there are three or more stations within 1000 km, these stations are used for the adjustment.”
Only Falalop Island and Yap Caroline Is. (both radiance 0) are within 1000km of Guam; Truk (radiance 10) is just over at 1021km away. Appyling the distance rule means that only two rural stations are available to correct those on Guam, so it seems the Guam stations join a list of many not used in the final output from GIStemp for this reason.
There are other changes too. Nassau Airport in the Bahamas (radiance 10) is no longer adjusted. It had previously suffered a “wrong way” adjustment for UHI correction (right) which seemed, well, just wrong, but now remaines unadjusted, with temperatures around 1900 similar to those today.
However, there do seem to be problems with location of stations. These have been highlighted and discussed by both Ron Broberg and Peter O’Neill, and the errors are serious enough to cause misclassification of quite a number of stations. Peter has even found examples of stations “out at sea”. The problem is finding all of them. As Peter says:
“The problem [accuracy of the latitude/longitude coordinates in the metadata] is, as they say, “even worse than we thought”. One of the consumers of GHCN metadata is of course Gistemp, and the implications of imprecise latitude/longitude for Gistemp are now considerably greater, following the change last month to use of satellite-observed night light radiance to classify stations as rural or urban throughout the world, rather than just in the contiguous United States as was the case previously. As about a fifth of all GHCN stations changed classification as a result, this is certainly not a minor change. But how can you judge night light radiance of stations which are not where you believe them to be?”
What about other changes? Well plenty of those too. There are many examples of stations dropping out of use and of those that suddenly wake up after years of apparent inactivity (at least as far as reporting temperatures for use in the GHCN datafile). Here are just a few examples those now reporting again:
- Russia: Bajkit, Vanavara, Selagoncy, Dudinka, Aleksandrovsk
- Mexico: Temosachic,Ch, Zamora,Mich.
- Turkey: Kirsehir, Bolu, Elazig, Gumushane
- Norway: Karasjok, Ona Ii, Fokstua Ii
From looking at how changes in the reporting stations can affect the temperature record (China, Turkey, more examples to come), this whole issue is a bit like mapping sand bars – trying to figure out certainties in a constantly changing landscape. But then, the whole climate issue is like that anyway.