Stephen McIntyre Profile picture

Aug 7, 2019, 17 tweets

Here are two series which were included in PAGES 2017 - both from Pakistan (from about same latitude as bristlecone pines.) One went up in late 20th, one went down. Guess which one was selected into PAGES 2019. Cherry-picking by climatists has gone well beyond being a joke.

2/ here's another example from PAGES2K Asian network: two juniper (JUSP) sites at different elevations in Bagrot Valley, Pakistan. One had upspike at very end, the other had decline. No points for guessing which is in PAGES2019.

3/ another contradiction: in the SW US at near-identical latitude, lower boundary trees said to be precipitation proxies, while upper boundary (higher altitude) more temperature related. Even at upper boundary, very dry in US and validity as temperature proxy questionable.

4/ from 229 Asia tree ring sites in PAGES 2019 (already screened), PAGES 2019 selected 47 sites according to criteria which accentuated 20th century increase. More sites from each of Mongolia and Pakistan than either Russia or China

5/ of 8 Russian tree ring series in PAGES 2019, 7 (!?!) come from exactly the same site: Ust Koksa Hill, Altai (ring width, early width, late width, max density, min density, late density, early density). Site at tricorner of Russia, Kazakh and Mongolia. Other site EAST of Japan.

6/ the Briffa troika of Russian tree ring sites (Yamal, Taimyr and the bodged version of Polar Urals) are used in PAGES2019 Arctic network, rather than Asia network.

7/ here's another puzzling and potentially serious PAGES 2019 problem -one which also impacts PAGES 2017 and PAGES 2013. I spot checked tree ring series with highest 2005 values - two over 4.4 SD units - paki033 and indo005. Note that a couple of authors over-represented.

most extreme (over 5 SD units in 2005) was paki033, a PIGE series from Pakistan (35 30N; 74 45E). It is used in PAGES 2013, PAGES 2017 and PAGES 2019. PAGES version has dramatic upspike, since 1991 and especially since 2002.

measurement data is at NOAA . Examining the underlying data, there are some issues. Usually, tree ring specialists require 10+ cores for valid chronology (exceptionally 5+ cores). paki033 doesn’t have 5 cores until 1776, only 2 cores as late as 1762. www1.ncdc.noaa.gov/pub/data/paleo…

but there's something MUCH worse. paki033 chronology NOT archived, only measurements. I calculated chronology using standard technique. Totally different result: no huge late upspike. Instead long decline with modest late recovery. ??!?? @onturenio, you should look at this

@onturenio I had similar problem when I tried to replicate chronology for INDO005: instead of the dramatic upspike of PAGES2K series, I obtained only a minor recovery from long decline.

@onturenio The Asian tree ring series in PAGES2017 were identical to the Asia tree ring data in PAGES 2013. To my knowledge, the PAGES 2013 chronologies had not been previously published in a technical publication, instead appearing first in PAGES 2013.

@onturenio I've triple checked the calculations. While the versions differ dramatically, they also clearly come from the same data (due to high frequency similarity). Where did PAGES2K chronologies for paki033 and indo005 come from? How were they calculated?

@onturenio this doesn't impact ALL the series. Chronology for nearby paki027 site is easily replicated. Not sure what's going on with the two series shown above. They emerged from spot check of high closers, not cherry pick from examining numerous series.

@onturenio here's another PAGES tree ring chronology PAKI021 - where PAGES version goes up, while chronology calculated from underlying measurement data at NOAA goes down.
Kalash Valley Bumburet - PIGE. Used in P13, P17 and P19.

@onturenio one more Pakistan tree ring series (PAKI036) in which PAGES version is more HS than chronology calculated from underlying data according to standard method. Less dramatic difference than others, but still very distinct.

@onturenio the NOAA archive for this measurement data gives PAGES2K (2013) as academic citation -NOT a specialist publication of the Pakistan tree ring data.

Share this Scrolly Tale with your friends.

A Scrolly Tale is a new way to read Twitter threads with a more visually immersive experience.
Discover more beautiful Scrolly Tales like this.

Keep scrolling