Transcript Slide 1

Reconciling Group and Wolf Sunspot
Numbers Using Backbones
Leif Svalgaard
Stanford University
5th Space Climate Symposium, Oulu, 2013
The ratio between Group SSN and Wolf [Zürich,
International] SSN has a marked discontinuity ~1882:
Reflecting the well-known secular increase of the Group SSN
1
Why a Backbone? And What is it?
Building a long time series from observations made over
time by several observers can be done in two ways:
• Daisy-chaining: successively joining
observers to the ‘end’ of the series,
based on overlap with the series as it
extends so far [accumulates errors]
• Back-boning: find a primary observer
for a certain [long] interval and
normalize all other observers
individually to the primary based on
overlap with only the primary [no
accumulation of errors]
When several backbones have been constructed we can
join [daisy-chain] the backbones. Each backbone can be
improved individually without impacting other backbones
Chinese Whispers
Carbon Backbone 2
The Wolfer Backbone
Alfred Wolfer observed 1876-1928 with the ‘standard’ 80 mm telescope
Years of overlap
1928
1876
Rudolf Wolf from 1860 on
mainly used smaller 37
mm telescope(s) so those
observations are used for
the Wolfer Backbone
80 mm X64
37 mm X20
3
Normalization Procedure
Number of Groups
Number of Groups: Wolfer vs. Wolf
12
9
Wolfer
8
Yearly Means 1876-1893
10
Wolf*1.653
7
8
Wolfer = 1.653±0.047 Wolf
6
2
R = 0.9868
5
Wolfer
6
4
4
3
Wolf
2
2
F = 1202
1
Wolf
0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
0
1860
1865
1870
1875
1880
1885
1890
1895
For each Backbone we regress each observers group counts for each year against
those of the primary observer, and plot the result [left panel]. Experience shows that
the regression line almost always very nearly goes through the origin, so we force it
to do that and calculate the slope and various statistics, such as 1-σ uncertainty.
The slope gives us what factor to multiply the observer’s count by to match the
primary’s. The right panel shows a result for the Wolfer Backbone: blue is Wolf’s
count [with his small telescope], pink is Wolfer’s count [with the larger telescope],
and the orange curve is the blue curve multiplied by the slope. It is clear that the
4
harmonization works well [at least for Wolf vs. Wolfer].
Regress More Observers Against Wolfer…
5
The Wolfer Group Backbone
Wolfer Group Backbone
14
12
10
14
Wolfer Backbone Groups
Weighted by F-value
Number of Observers
Standard Deviation
12
10
8
8
6
6
4
4
2
2
0
1840
0
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
The Wolfer Backbone straddles the interval around 1882 with good coverage
(~9 observers) and with reasonable coverage 1869-1925 (~6 observers). Note
that we do not use the Greenwich [RGO] data for the Wolfer Backbone.
6
Hoyt & Schatten used the
Group Count from RGO
[Royal Greenwich
Observatory] as their
Normalization Backbone.
Why don’t we?
Because there are strong
indications that the RGO
data is drifting before ~1900
José Vaquero found a similar
result which he reported at the
2nd Workshop in Brussels.
Sarychev & Roshchina report in Solar Sys.
Res. 2009, 43: “There is evidence that the
Greenwich values obtained before 1880
and the Hoyt–Schatten series of Rg before
1908 are incorrect”.
Could this be caused by
Wolfer’s count drifting? His kfactor for RZ was, in fact,
declining slightly the first
several years as assistant
(seeing fewer spots early on –
wrong direction). The group
count is less sensitive than the
Spot count and there are also
the other observers…
7
The Schwabe Backbone
Schwabe received a 50 mm telescope from Fraunhofer in 1826 Jan 22. This
telescope was used for the vast majority of full-disk drawings made 1826–1867.
Years of overlap
?
Schwabe’s House
For this backbone
we compare with
Wolf’s observations
with the large
80mm standard
telescope
8
Regressions for Schwabe Backbone
9
The Schwabe Group Backbone
Schwabe Group Backbone
8
7
10
Weighted by F-value
Number of Observers
Schwabe Backbone Groups
9
8
6
Standard Deviation
7
5
6
4
5
4
3
3
2
2
1
0
1790
1
1795
1800
1810
1815
1820
1825
1830
1835
1840
1845
1850
1855
1860
1865
1870
1875
1880
0
1885
Comparing Schwabe Backbone with Hoyt & Schatten Group Number
9
8
1805
10
9
Groups
8
7
7
6
Schwabe Backbone
5
6
5
H&S Groups
4
4
3
3
2
2
1
1
0
1790
1800
1810
1820
1830
1840
1850
1860
1870
1880
0
1890
10
Joining the Two Backbones
Comparing Schwabe with Wolfer backbones over 1860-1883 we find a normalizing factor of 1.55
Comparing Overlapping Backbones
Reducing Schwabe Backbone to Wolfer Backbone
12
12
10
10
1860-1883
Wolfer
1.55
8
6
6
4
4
2
0
1860
Wolfer = 1.55±0.05 Schwabe
8
Schwabe
R2 = 0.9771
2
Wolfer
Schwabe
0
1865
1870
1875
1880
1885
1890
1895
1900
0
1
And can thus join the two backbones covering ~1825-1946
2
3
4
5
6
7
11
8
Comparison Backbone with GSN and WSN
Sunspot Group Numbers 1790-1945
14
12
Backbone
Groups = GSN/13.149
Groups = WSN/12.174
10
8
6
4
2
0
1790
1800
1810
1820
1830
1840
1850
1860
1870
1880
1890
1900
1910
14
The discrepancy
around 1860 might
be resolved using
the newly digitized
Schwabe data
supplied by Arlt et
al. (This meeting).
12
10
8
6
4
2
0
1790
1800
1810
1820
1830
1840
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
Scaling all
curves to
match for
1912-1946
shows that the
combined
backbone
matches the
scaled Wolf
1920
1930
1940
12
Number
Conclusions
• Using the ‘Backbone’ technique it is possible to
reconstruct a Group Sunspot Number 18251945 that does not exhibit any systematic
difference from the standard Wolf [Zürich, Intnl.]
Sunspot Number
• This removes the strong secular variation found
in the Hoyt & Schatten GSN
• And also removes the notion of a Modern Grand
Maximum
13