Reposted from WUWT, this article by Talkshop contributor Lucy Skywalker is excellent background to the latest episodes in the hockey stick wars between Realclimate.org and ClimateAudit.org, where Steve McIntyre has just delivered two more blows to the ‘Team’ treering chronology-temperature reconstructions. These have been used to support the ‘Hockey Stick’ graph drawn by Michael Mann using statistical techniques whixh have since been heavily criticised by expert statisticians. Mann was also selective in which tree ring chronologies he used, favouring ones which supported his conclusion, and excluding ones which didn’t. This kind of selection bias is frowned on in all scientific disciplines; except paleo-dendro-climatology apparently, where they have been known to cheerfully admit to it.
However as we mentioned earlier on the subject of biological growth populations, this does not mean that one could not improve a chronology by reducing the number of series used if the purpose of removing samples is to enhance a desired signal. The ability to pick and choose which samples to use is an advantage unique to dendroclimatology.
Or more succinctly
you had to pick cherries if you want to make cherry pie
They really need to read Lucia’s post on cherry picking.
Let’s look closely and compare local thermometer records (GISS) with the Twelve Trees, upon whose treerings depend all the IPCC claims of “unprecedented recent temperature rise”.
For my earlier Yamal work, see here and here. For the original Hockey Stick story, see here and here.
Half the Hockey Stick graphs depend on bristlecone pine temperature proxies, whose worthlessness has already been exposed. They were kept because the other HS graphs, which depend on Briffa’s Yamal larch treering series, could not be disproved. We now find that Briffa calibrated centuries of temperature records on the strength of 12 trees and one rogue outlier in particular. Such a small sample is scandalous; the non-release of this information for 9 years is scandalous; the use of this undisclosed data as crucial evidence for several more official HS graphs is scandalous. And not properly comparing treering evidence with local thermometers is the mother of all scandals.
I checked out the NASA GISS page for all thermometer records in the vicinity of Yamal and the Polar Urals, in “raw”, “combined”, and “homogenized” varieties. Here are their locations (white). The Siberian larch treering samples in question come from Yamal and Taimyr. Salehard and Dudinka have populations of around 20,000; Pecora around 50,000; Surgut around 100,000; all the rest are officially “rural” sites. Some are long records, some are short.
Russia has two problems. First, many records stopped or became interrupted around 1990 after the ending of Soviet Russia; worst affected are the very telling Arctic Ocean records. Second, during Soviet Russia (and possibly now for all I know), winter urban records were “adjusted” downwards so that the towns could claim more heating allowances. Nevertheless, it will become clear that these issues in no way impede the evidence regarding treerings.
Click to enlarge these graphs. The first shows the 20 GISS stations closest to Yamal and the Polar Urals. The second shows treering width changes over time (only 10 of the 12 trees here). This was supposedly compared with local thermometer records, and used to calibrate earlier treering widths as temperature measurements to create a 1000-year temperature record. It was a pig to turn these graphs into a stack of transparent lines at the same scale as the GISS records for comparison, but finally, interesting material started to emerge.
I scaled all the GISS thermometer records to the same temperature scale, and ran them all from 1880 to 2020 at the same time scale (GISS graphs do not do this). I overlaid them as transparent lines along their approximate mean temperatures for comparison. Mean temperatures (visually judged) vary from around -2ºC (Pecora) to -13ºC (Selagoncy, Olenek, Hatanga, and Ostrov Uedine) and even -15ºC (“Gmo Im E.K. F”). The calibrations are degrees Centigrade anomaly, and decades.
Ha! Straightway we see clear patterns emerging. Let’s agree them:
Thermometer records: (1) time-wise, thermometers show temperatures rising from 1880 to 1940 or so; (2) temperatures fall a little from 1940 to 1970; (3) temperatures then rise a little but do not quite regain the heights of the 1940’s; (4) despite mean temperatures ranging from -2ºC to -15ºC (total means range 13ºC), and a range of temperature anomalies from each mean of only 9ºC from warmest year to coldest year, when mean temperatures are aligned, clear correlations emerge; (5) there are high variations between adjacent years. We shall investigate all this more closely in a minute.
Treering records: I’ve shown here the full records given for the 10 trees that runs from 1800 to 2000; but below, I use the same timescale as the thermometer records (1880-2020) for comparison. It is useful to see a few things here already: (6) treering sizes are increasing from 1830; (7) before that they show a decrease; (8) they do show correlation from 1880 on (this is NOT proof that the correlation is due to temperature).
Yamal area: (9) The 7 stations around Salehard seem to go in lock step with each other pretty well. (10) The five Yamal treering records (YAD) also correlate with each other, showing spikes around 1910, 1925, 1940, 1955, 1965, and 1980-1990. (11) But the treerings fall out with each other 1990-2000; and (12) these treering spikes do NOT correspond to the thermometer temperature spikes; but (13) there is a slight correlation with the longterm temperature; however, (14) crucially, there is no hockeystick blade in the thermometer record (15) nor is there one in the treering record if we remove the red YAD061 which is clearly an outlier – only a plateau’d elevation of the peaks throughout the 20th century starting before the real CO2/temp rise (and this is actually matched by pre-1800 values at times).
Excuse me for wondering if treerings beat to a different drum than temperature – it is certainly curious that there appears to be something causing correlations in the treerings. Wind? Sunspots? The moon? But let’s check by zooming in a little closer…
Salehard close-up: (16) all the nearby thermometer records mirror Salehard closely, although stations are up to 500 miles apart, the range of mean temperatures is -2ºC to -9ºC, and the range of annual temperatures at each station is up to about 9ºC – altogether a remarkable consistency. Click to see animated version of these records. (17) The close fit of Mys Kamennij (pale sea-blue) is particularly significant, since it is maritime and rural, and the same distance as Salehard from the treering site (some 120 miles), but in the opposite direction; (18) Ostrov Waigatz (Vaigach Island) shows the same pattern but with greater extremes; (19) in comparison with all this, the treering records show virtually no correlation at all – yet since treering differences between summer and winter exist at all, one would expect to see some correlation with warmer and colder years. (20) Perhaps if a far larger sample were used, a correlation might be detected, but clearly (21) we have trees here that are far too individual – especially YAD061.
Polar Urals: Here are seven station records around the Polar Urals site, compared with the five Taimyr (POR) treering records. (22) Mean temperatures are lower here – further North but also more continental, so perhaps the summers are as warm as Yamal, with similar near-treeline environment. (23) more noise in the temperature record, but the overall pattern is still the same; (24) 1943, 1967, 1983 are warm in common with the Salehard records, and 1940 is cold; other years are harder to compare. (25) The early fragmentary records for Dudinka and Turuhansk still fit together and overlay the Salehard records well, showing clear temperature rise between 1880 and 1940. (26) The treering records are fairly coherent, more so than the Yamal ones, and (27) they fit the Yamal records’ spikes in 1910, 1925, 1940, 1955, 1965, and 1980 on, but (28) again, this does not fit the temperature record.
The best of both record series: Really rural thermometer records from the maritime Arctic: (29) show the strongest pattern yet which (30) fits the other two sets of thermometer records but (31) does not fit the treering records even though (32) the treerings show coherent patterns within themselves, despite the two sites being some 800 miles apart.
Briffa’s full chronology: The Yamal chronology Briffa used (black) is compared with Polar Urals (green) and shows recent temperatures exceeding the Medieval Warm Period but (33) this is highly questionable, as is the recent final uptick. No MWP supports the alarmist “Unprecedented!” yet Polar Urals generally have been shown to fit local thermometer records better than Yamal for the period of overlap.
More GISS Arctic graphs: There are many serious problems with GISS but we can only take the evidence here. (34) GISS 64ºN+ shows a misleading trend line – temperatures rise to 1940, fall to 1970, rise to 2000 but not higher than 1940, then level off after 2000; (35) I don’t know what stations went into this composite – the final uptick alerts my suspicions to some UHI or other station problems; (36) Tamino takes the biscuit for cherrypicked trends in the GISS 80ºN+ North Polar winter record (sea green) – it clearly opposes the general worldwide fall in temperatures 1940-1970. However, it’s interesting to see such extremes.
GISS’ homogeneity adjustments: Thankfully, only a few of these Russian records are “adjusted”. But the alterations are telling. Surgut spikes upwards over Salehard from about 1960 on – but (36) the adjustment (probably UHI) is perversely done by truncating and moving earlier records upwards, instead of adjusting later records downwards. And (37) why were Salehard’s and Ostrov Uedine’s earlier “raw” records omitted in the adjusted records? I think every correction here will tend to amplify global warming trends.
GISS world temperatures, 2008: This map was shown in Tingley & Huybers’ latest Hockey Stick presentation at PAGES conference. GISS’ own station records around Yamal and Polar Urals appear to show (38) this map is misleading, since according to GISS’ own records, above, averages local to Yamal / Polar Urals after 2000 are at the most 1.5ºC anomaly (above local mean).
CRU Arctic temperatures, seasonal anomalies: (graph by romanm) Since this is from uncheckable individual station records, (39) the figures could be contaminated by various “correction” factors, (41) UHI is especially likely in the winter. But note that (42) the difference in character between months, and between summers and winters, is striking – summers have hardly changed – and (43) still no definitive Hockey Stick as per illustrations and per Briffa’s Yamal treering record, nothing beyond the range of natural patterns clearly evidenced here. Even the known slight overall increase during the twentieth century takes place mainly earlier in the century.
Conclusions: There is no sign whasoever of a Hockey Stick shape with serious uptick in the twentieth century, in the thermometer records. Yet these records are clearly very consistent with each other, no matter how long the record or how cold, high, or maritime the locality, with a distance span of over a thousand miles. Neither does the Hockey Stick consistently show in the treerings except in the case of a single tree. Even with thermometer records that are incomplete and suffering other problems, the “robust” conclusion is –
“Warmist” treering proxy temperature evidence is falsified directly by local thermometer records.