Do We Know that the 1990s were the Warmest Decade of the Millennium

1
Do We Know that the 1990s were the Warmest Decade
of the Millennium?

• Stephen McIntyre
• climateaudit.org
• Toronto Ontario
• 2009 International Conference on Climate Change
• New York
• March 9, 2009

2
Questions

• How do they know that 1998 was the warmest year
  of the millennium?
• They dont. Minor variations in data
  versions (e.g. bristlecones) yield opposite
  results.
• Does it matter?
• For big policy, probably not. But its
  an interesting question and statistics should
  be done right.

3
IPCC 2001 The Rise of the Stick
The Hockey Stick graph, attributed to Mann et
al (Nature 1998, GRL 1999 (MBH), occurred six
times in IPCC TAR. Inset John Houghton at IPCC
Press Conference, 2001
4
The Soundbites The Warmest Decade and Year
 The 20th century was the warmest in the Northern
Hemisphere in the past 1000 years. The 1990s was
the warmest decade on record and 1998 was the
warmest year - in Canada and internationally. -
David Anderson, April 5, 2002  The 20th century
was the warmest in the Northern Hemisphere for
the past 1000 years and the 1990s the warmest
decade on record... The science of climate change
has been subjected to international scrutiny,
open to all qualified experts, peer review,
atmospheric modeling and process studies.
Liberal Party of Canada Caucus, Aug. 22, 2002
5
As Dr Thompsons Thermometer Shows,

• the vaunted Medieval Warm Period (the third
  little red blip from the left below) was tiny in
  comparison to the enormous increases in
  temperature in the last half-century - the red
  peaks at the far right of the graph. These
  global-warming skeptics - a group diminishing
  almost as rapidly as the mountain glaciers -
  launched a fierce attack against another
  measurement of the 1000 year correlation between
  CO2 and temperature known as the hockey stick,
  a graphic image representing the research of
  climate scientist Michael Mann and his
  colleagues.
• Gore actually shows Manns reconstruction
  spliced with the Jones instrumental record.

6
IPCC 2007 Spaghetti Graphs
It is likely that this 50-year period was the
warmest Northern Hemisphere period in the last
1.3 kyr,
Excerpt from IPCC 2007 Box 6.4 Figure 1..
7
Multivariate Calibration
McIntyre and McKitrick, 2009 Paleoclimate
reconstructions are an application of
multivariate calibration, which provides a
theoretical basis for confidence interval
calculation (e.g., Osborne 1991 Brown and
Sundberg, 1987). Inconsistency among proxies
sharply inflates confidence intervals. Applying
the inconsistency test of Brown and Sundberg to
Mann et al. A.D. 1000 proxy data shows that
finite confidence intervals cannot be defined
before 1800. MBH2009 The method of uncertainty
estimation (use of calibration/validation
residuals) is conventional Luterbacher et al
2004 Wahl and Ammann 2007
8
Small and uninteresting perturbations of
data/technique
de Leeuw (1988) We usually do not want a small
and uninteresting perturbation of our data to
have a large effect on the results of our
technique. Classical statistics has always
studied stability by using standard errors or
confidence intervals. Gifi thinks this is much
too narrow and other forms of stability are
important as well
9
MM2005 on slight variations of methods and data

Redrawn from MM2005b
10
MBH98 AD1400 Proxies
11
There is no dispute over properly specified
calculations

Top Left MBH, 2003 top right Bürger et al
2006 bottom drawn from calculations using Wahl
and Ammann 2007 algoithm. Middle- varies PC
method (their Scenario 5) right - varies
bristlecones (Their Scenario 6).
12
NAS 2006

• For periods prior to the 16th century, the Mann
  et al. (1999) reconstruction that uses this
  particular principal component analysis technique
  is strongly dependent on data from the Great
  Basin region in the western United States
  bristlecones.
• some reconstructions are not robust with respect
  to the removal of proxy records from individual
  regions (see, e.g., Wahl and Ammann in press).

13
IPCC 2007

• The McIntyre and McKitrick 2005a,b criticism
  relating to the extraction of the dominant modes
  of variability present in a network of western
  North American tree ring chrono-logies, using
  Principal Components Analysis may have some
  theoretical foundation, but Wahl and Amman (2006)
  also show that the impact on the amplitude of the
  final reconstruction is very small (0.05C).

14
Wahl and Ammann (in press 2005 in press
2006 Sep. 2007)

• when the full information in the proxy data is
  represented by the PC series i.e. enough to get
  the bristlecones in, the impact of PC
  calculation methods on climate reconstruction in
  the MBH method is extremely small a slight
  modification to the original Mann et al.
  reconstruction is justifiable for the first half
  of the 15th century (0.050.10?), which leaves
  entirely unaltered the primary conclusion of Mann
  et al.
• Bristlecones add necessary verification
  skill indicating that the records carry
  important climate information at the level of
  eigenvector patterns in global surface
  temperatures. These results are valid
  notwithstanding issues concerning these proxies
  empirical relationship to local/regional surface
  temperatures after 1850

15
Wegman 2006

• Wahl and Ammann argue that if one adds enough
  principal components back into the proxy, one
  obtains the hockey stick shape again. This is
  precisely the point of contention
• A cardinal rule of statistical inference is that
  the method of analysis must be decided before
  looking at the data. The rules and strategy of
  analysis cannot be changed in order to obtain the
  desired result. Such a strategy carries no
  statistical integrity and cannot be used as a
  basis for drawing sound inferential conclusions.

16
A Dozen Independent Studies
Mann (Senate Testimony, 2003) More than a dozen
independent research groups have now
reconstructed the average temperature of the
northern hemisphere in past centuries.. Wigley
(AP Interview, 2003) about a dozen independent
studies suggest the 20th century was warmer than
normal, Wigley points out. Mann (Letter to
House Committe, 2005) Recent work since the TAR
has provided further support for this conclusion
the Stick, which is now common to more than a
dozen independent studies published in the
peer-reviewed scientific literature. Wikipedia
2009 More than a dozen independent research
groups have now reconstructed the average
temperature of the northern hemisphere in past
centuries.
17
Not Entirely Independent

• IPCC 2007 As with the original TAR series,
  these new records are not entirely independent
  reconstructions inasmuch as there are some
  predictors (most often tree ring data and
  particularly in the early centuries) that are
  common between them, but in general, they
  represent some expansion in the length and
  geographical coverage of the previously available
  data (Figures 6.10 and 6.11
• NAS 2006 Because the data are so limited,
  different large-scale reconstructions are
  sometimes based on the same datasets, and thus
  cannot be considered as completely independent.

18
Key proxies are repeated in so-called
independent spaghetti graph studies
Above Adapted from Wegman 2006 Figure 5.8.
Bristlecones/ foxtails are used in all but three
spaghetti graph studies. Polar Urals and
Tornetrask versions re used in every study.
19
Ecologists warm California MWP
Miller et al 2006 (Quat Res) Using contemporary
distributions of the species, we modeled
paleoclimate during the MWP to be significantly
warmer (3.2 deg C annual minimum temperature)
and slightly drier (-24 mm annual precipitation)
than present.
Left- Bristlecone treeline (Lamarche 1973)
declined since MWP right Subfossil medieval
trees above present treeline at Whitewing Mt,
Sierra Nevadas, California.
20
Heavy Equipment

• IPCC 2007 the possibility of investigating
  these issues further is restricted by the lack
  of recent tree ring data at most of the sites
  from which tree ring data discussed in this
  chapter were acquired
• Michael Mann paleoclimatologists are attempting
  to update many important proxy records to the
  present, this is a costly, and labor-intensive
  activity, often requiring expensive field
  campaigns that involve traveling with heavy
  equipment to difficult-to-reach locations (such
  as high-elevation or remote polar sites). For
  historical reasons, many of the important records
  were obtained in the 1970s and 1980s and have yet
  to be updated.

21
Ababneh 2006 did not replicate Graybills
bristlecone chronology
Left red MBH PC1 (used in IPCC 2007 Box 6.4)
green Graybill Sheep Mt chronology. Right -
Ababneh (2006, 2007) chronology.. All series
scaled to 1400-1980. 950-1100 and 1856-1980
highlighted.
22
Almagre CO Bristlecones are not setting records
Decline in recent portion of the chronology. The
most noticeable feature is the low growth in
the1840-50s associated with drought in other
Colorado tree ring chronologies.
Left- MBH PC1 (bristleocnes) right updated
Almagere chronology. In dimensionless chronology
units, basis 1.
23
Controversial Starbucks Hypothesis Proven!
Left 7 am at Starbucks, Colorado Springs right
bristlecones at 3600 m
24
Almagre program identified exact trees during
update and measurements online within 3 months
Left Pete Holzmann sampling Tree 31 in 2007
right- Donald Graybill sampling same tree in
1987.
25
Strip Bark simple statistical models dont
apply
Graybill and Idso 1993 Another tree selection
factor that is crucial to our findings involves
tree form... Strip bark trees were the primary
focus of investigation wherever possible,
(Brunstein USGS 2006http//www.climateaudit.org/?p
2239
26
Key proxies are used over and over.
Above From Wegman 2006 Figure 5.8 showing proxy
overlap IPCC 2007 Box 6.4 Fig 1.
27
Polar Urals Briffa (Jones) 1995
Above Polar Urals version of Briffa et al 1995,
used in Jones et al 1998 reconstruction. All
series scaled to scale of Briffa et al 1995.
950-1100 and 1856-1980 highlighted.
28
Evidence for warm MWP in Siberia
Left - Polar Urals treeline declined since
Medieval Warm Period (from Shiyatov 1995) right
subfossil medieval trees in foreground above
modern treeline in background (Esper)

29
Polar Urals Update
Above Polar Urals version incorporated into
Esper et al 2002. All series scaled to scale of
Briffa et al 1995. 950-1100 and 1856-1980
highlighted.
30
Yamal Substitution (Briffa 2000)
Above - Yamal chronology from Briffa 2000, used
in all but one subsequent study. All series
scaled to scale of Briffa et al 1995. 950-1100
and 1856-1980 highlighted.
31
Tornetrask Briffa 2000 vs Grudd 2006
Left Version used in IPCC 2007. Right -
Grudd 2006. All series scaled to scale of Briffa
et al 1995. 950-1100 and 1856-1980 highlighted.
32
Are ice cores any help?

• wide variety of ice core patterns since MWP.
• dO18 at Mt Logan, Yukon went down in 20th
  (attributed to regional change)
• Elevated dO18 at Law Dome, Antarctica ignored by
  NAS panel

Ice core dO18 from N to S.
33
Different Spaghetti Recipes

• NRC panel used 4 reconstructions as comfort for
  comparing modern-medieval levels, but did not
  verify that cited reconstructions met proxy
  quality standards. Medieval-modern relationship
  reverses with trivial and justifiable variations
  in proxy selection.

34
Glacier Retreat Alberta
Retreat of Saskatchewan Glacier (in Alberta) in
1999 exposed in situ stumps dated 2800-2900 BP
(14C) presumably not exposed in MWP In August
1999, a severe rainstorm resulted in stream
avulsion along the NE flank of the Saskatchewan
Glacier snout When first examined in early
September, erosion through a 3- to 5- m sequence
of glacial outwash and overlying till had exposed
17 sheared stumps rooted within a well-preserved
paleosol . By September 2000, the meltwater
channel had shifter southward and had eroded
through 5 m of fluted moraine deposits exposing 2
additional rooted stumps and flushing an
additional 40 detrital boles onto the outwash
surface. The stumps were rooted in a deeply
weathered pedogenic surface and prior to their
exposure were buried by 3-5 m of sediment. With
one exception, these dates collectively reflect
the consequences of a single Neoglacial advance
of the Saskatchewan Glacier into an established
valley bottom forest.
35
Quelccaya Glacier, Peru
Distichia muscoides, a component of Andean peat
deposits, exposed in recent Quelccaya glacier
retreat. Present limit is 400-500 m lower than
the discovery. Peat samples have a wide variety
of 14C dates - many about 4500 BP.
36
Green Alps
Left- modern view of a pass in the Alps (with
glacier lines of 1922 and 1856) right -
Schlüchter and Jörin 2004 reconstruction of view
Roman times of 2000 years ago
37
Does it matter?
38
Why did IPCC feature the least important figure?
It is one of those odd ironies that the figure
that everyone keeps talking about is actually the
least important from the point of view of
understanding climate responses to forcing. Gavin
Schmidt, 2005
39
Is it the Least Important Figure?
Left IPCC 2001 Synthesis for Policy makers
Figure SPM-10b with MBH reconstruction Right
IPCC 1990 version of Lamb (about 1 deg range)
together with IPCC 2007 SPM range of scenarios
yellow shows constant 2000 GHG.
40
Worse than We Thought
There is (yet another) odd inversion about the
idea that if we were to switch from, say, MBH
(less variance) to Moberg (more) that would
somehow imply a reduction in expected future
warming. That is completely wrong. If the past
temperatures varied more, it implies a higher
sensitivity to forcing, and therefore a higher
future William Connolley, 2005
41

• Visit www.climateaudit.org for
• ongoing discussion.
• Presentation will be posted at www.climateaudit.or
  g/pdf/newyork09 with bibliography