Transcript Solar Wind During the Maunder Minimum Leif Svalgaard Stanford University Predictive Science, San Diego, 4 Sept.

Solar Wind During the
Maunder Minimum
Leif Svalgaard
Stanford University
Predictive Science, San Diego, 4 Sept. 2012
1
Indicators of Solar Activity
• Sunspot Number (and Area,
Magnetic Flux)
• Solar Radiation (TSI, UV, …,
F10.7)
• Cosmic Ray Modulation
• Solar Wind
• Geomagnetic Variations
• Aurorae
• Ionospheric Parameters
• Climate?
• More…
Longest direct
observations
Rudolf Wolf
After Eddy, 1976
Solar Activity is Magnetic Activity
2
Unfortunately Two Data Series
Ken Schatten
Hoyt & Schatten, GRL 21, 1994
3
How Well was the Maunder Minimum Observed?
H&S
27
It is not credible that for many
years there were not a single
day without observations
Number of days per year with ‘observations’
4
Number of Observers per Year for GSN
More Realistic
Assessment
30
25
20
15
10
5
0
1600
1650
1700
1750
1800
1850
1900
1950
2000
H&S
1610
5% of 365 is ~20 days
1700
1825
Even after eliminating the spurious years
with ‘no missing data’ there are enough left
to establish that the Maunder Minimum had
very few visible sunspots and was not due to
general lack of observations
5
The Ratio Group/Zurich SSN has
Two Significant Discontinuities
At ~1946 (after Max Waldmeier took over) and at ~1885
6
Locarno
Sergio
Cortesi
Locarno is today the
reference station of
the official SIDC SSN
223
227
228
231
232
233
234
235
3
4
13
4
4
6
9
3
1
1
1
1
1
1
1
1
8
46
11
Effect of
Weighting of
Sunspots
223
227
228
231
232
233
234
235
3
4
13
4
4
6
9
3
1
1
6
1
2
4
4
1
8
46
20
SSN = 10*G+S
126
In the 1940s the observers in Zürich [and Locarno]
began to Weight spots. The net result is a ~20%
inflation of the official Zürich SSN since ~1945
100
26% inflated
Unweighted count red
7
Compared
with Sunspot
Area (obs)
1000
Rz
100
10
1
0.1
1
10
100
1000
10000
SA
0.1
Not linear relation,
but a nice power
law with slope
0.732. Use relation
for pre-1945 to
compute Rz from
Area, and note
that the observed
Rz after 1945 is
too high [by 21%]
8
Removing the discontinuity in ~1946,
by multiplying Rz before 1946 by 1.20, yields
Leaving one significant discrepancy ~1885
9
Wolf-Wolfer Groups
Number of Groups: Wolfer vs. Wolf
9
Wolfer
8
Yearly Means 1876-1893
7
6
Wolfer = 1.653±0.047 Wolf
5
R2 = 0.9868
Wolfer
4
3
2
80mm 64X
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
Number of Groups
12
10
Wolf*1.653
8
Wolfer
Wolf
6
4
Wolf
37mm 20X
2
0
1865
1870
1875
1880
1885
1890
1895
10
Making a Composite
Comparison Sunspot Groups and Greenwich Groups
10
Groups
9
8
7
6
Average
Quimby*
Wolfer
Winkler*
Wolf*
5
4
RGO*
3
Matched
on this
cycle
2
1
Year
0
1875
1880
1885
1890
1895
1900
1905
1910
1915
1920
Compare with group count from RGO [dashed line] and note its drift
11
Extending the Composite
Comparing observers back in time [that overlap first our composite and then
each other] one can extend the composite successively back to Schwabe:
Comparison Composite Groups and Scaled Zurich SSN
14
Composite
Zurich
12
10
8
6
4
2
0
1820
1830
1840
1850
1860
1870
1880
1890
1900
1910
1920
1930
There is now no systematic difference between the Zurich SSN
and a Group SSN constructed by not involving RGO.
1940
12
Why are these so different?
Observer
Wolfer, A., Zurich
2% diff.
Wolf, R., Zurich
Schmidt, Athens
Weber, Peckeloh
Spoerer, G., Anclam
Tacchini, Rome
Moncalieri
Leppig, Leibzig
Bernaerts, G. L., England
Dawson, W. M., Spiceland, Ind.
Ricco, Palermo
Winkler, Jena
Merino, Madrid
Konkoly, Ogylla
Quimby, Philadelphia
Catania
Broger, M, Zurich
Woinoff, Moscow
Guillaume, Lyon
Mt Holyoke College
K-Factors
H&S RGO to Wolfer
1.094
1.117
1.135
0.978
1.094
1.059
1.227
1.111
1.027
1.01
0.896
1.148
0.997
1.604
1.44
1.248
1.21
1.39
1.251
1.603
1
1.6532
1.3129
1.5103
1.4163
1.1756
1.5113
1.2644
0.9115
1.1405
0.9541
1.3112
0.9883
1.5608
1.2844
1.1132
1.0163
1.123
1.042
1.2952
Begin
End
1876
1876
1876
1876
1876
1876
1876
1876
1876
1879
1880
1882
1883
1885
1889
1893
1897
1898
1902
1907
1928
1893
1883
1883
1893
1900
1893
1881
1878
1890
1892
1910
1896
1905
1921
1918
1928
1919
1925
1925
K-factors
1.8
This
analysis
1.6
1.4
1.2
1
H&S
0.8
0.8
1
1.2
1.4
1.6
1.8
2
No correlation
Number of Groups
12
10
Wolf*1.653
8
Wolfer
6
4
Wolf
2
0
1865
1870
1875
1880
1885
1890
13
1895
Why the
large
difference
between
Wolf and
Wolfer?
Because Wolf either
could not see groups of
Zurich classes A and B
[with his small telescope]
or deliberately omitted
them when using the
standard 80mm
telescope. The A and B
groups make up almost
half of all groups
14
Removing the discontinuity in ~1885 by
multiplying Rg by 1.47, yields
Only two adjustments remove most of the disagreement
and the evidence for a recent grand maximum (1945-1995)
15
The Effect on the Sunspot Curve
SIDC
No long-term trend the last 300 years
16
Removing the discrepancy between the Group
Number and the Wolf Number removes the
‘background’ rise in reconstructed TSI
I expect a strong reaction against ‘fixing’ the GSN from people that ‘explain’
climate change as a secular rise of TSI and other related solar variables
17
Typical Reconstruction
TSI ~ TSI0 +a·GSN + b·<GSN>11yr
Now
18
Kopp/LASP
Some More TSI
Reconstructions
Crucial question: is there a slowly
varying background? I think not.
19
The Auroral Record in Europe
45º
51º
55º
60º
Hungary
43º
55º
S. Sweden
Denmark
Effect of
Changing
Magnetic
Latitude
It is very difficult [impossible?] to calibrate accurately the auroral
record because of the unknown ‘civilization’ correction.
20
80-110 Year ‘Gleissberg Cycle’ in
Solar Activity Asymmetry?
Extreme Asymmetry during
the Maunder Minimum…
There are various dynamo
theoretical ‘explanations’ of NS asymmetry. E.g. Pipin,
1999. I can’t judge these…
Is this a ‘regular’ cycle or just
over-interpretation of noisy
data [like Waldmeier’s]?
‘Prediction’ from this: South will
lead in cycle 25 or 26 and
beyond. We shall see…
Zolotova et al., 2010
21
Asymmetric Solar Activity
22
18
Comparing Cycles 14 and 24
23
Polar Field Reversal SC24
WSO
24
How do we Know that the Poles
Reversed Regularly before 1957?
Wilcox & Scherrer, 1972
Svalgaard, 1977
“Thus, during last eight solar cycles
magnetic field reversals have taken
place each 11 year period”. S-M effect.
Vokhmyanin & Ponyavin, 2012
The predominant polarity = polar field polarity
(Rosenberg-Coleman effect) annually
modulated by the B-angle.
This effect combined with the RussellMcPherron effect [geomagnetic activity
enhanced by the Southward Component
of the HMF] predicts a 22-year cycle in
geomagnetic activity synchronized with
polar field reversals, as observed (now for
1840s-Present).
25
Cosmic Ray Modulation Depends
on the Sign of Solar Pole Polarity
The shape of the
modulation curve
[alternating ‘peaks’
and ‘flat tops’] shows
the polar field signs.
North pole
North pole
Miyahara, 2011
Svalgaard
& Wilcox,
1976
Ice cores contain a long
record of 10Be atoms
produced by cosmic
rays. The record can be
inverted to yield the
cosmic ray intensity.
The technique is not yet
good enough to show
peaks and flats, but
might with time be
refined to allow this.
26
The Cosmic Ray Record
17 pounds/yr
2 oz/year
Steinhilber et al. 2012
27
Cosmic Ray Proxy [Berggren et al.]
28
24-hour running means of the Horizontal Component of the low- & midlatitude geomagnetic field remove most of local time effects and leaves a
Global imprint of the Ring Current [Van Allen Belts]:
A quantitative measure of the effect can be formed as a series of the unsigned
differences between consecutive days: The InterDiurnal Variability, IDV-index
29
IDV is strongly correlated with HMF B,
but is blind to solar wind speed V
nT
10
9
8
7
6
5
4
3
2
1
0
IDV Independent of Solar Wind Speed
10
9
8
7
6
5
4
3
2
1
0
B obs
B calc from IDV
B obs median
B std.dev
100% =>
Coverage
1960 1965 1970
1975 1980 1985 1990
nT
600
V
20
500
400
15
IDV
300
?
10
200
5
100
B
0
1960
18
16
0
1970
1980
1990
2000
2010
IDV vs. Solar Wind Speed V (1963-2010)
IDV
1995 2000 2005 2010
14
HMF B as a Function of IDV09
10
B nT
12
1963-2010
10
8
8
6
6
4
2
4
y = 1.4771x0.6444
2
R = 0.8898
y = 0.4077x + 2.3957
2
R = 0.8637
4
10
0
0
2
6
8
12
14
2
R = 0.0918
2
V km/s
IDV
16
0
350
400
450
500
550
30
Space
Climate
10
Climatological Solar Wind Cycle (Base #13-#23)
Density
np
9
8
7
IMF
B
6
5
4
Since we can reconstruct
B, V, and n for 11 solar
cycles we can determine
an ‘average’ profile of the
solar wind through the
solar cycle
Speed
V/100
3
2
SSN
Rz
1
0
0
Years 11
22
33
44
10
9
66
10
200
Average of Solar Cycles 13-23 (1890-2003)
Average of Solar Cycles 13-23 (1890-2003)
9
180
Proton Density n
8
55
8
160
7
Flow Pressure
7
140
6
IMF Magnitude B
4
Solar wind Speed V o
3
2
6
120
Obs.1965-2003
5
Obs.1965-2003
100
5
80
4
60
3
40
Rz
1
2
Rz
1
20
Phase in Solar Cycle (years)
0
0
2
4
6
Phase in Solar Cycle (years)
0
8
10
12
0
2
4
6
0
8
10
12
31
Solar Activity 1835-2011
Sunspot Number
Monthly Average Ap Index
60
Ap Geomagnetic Index (mainly solar wind speed)
50
40
30
20
10
B nT
0
1840
Heliospheric Magnetic Field Strength B (at Earth) Inferred from IDV and Observed
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
10
B (IDV)
8
6
13
23
4
0
1830
B (obs)
Heliospheric Magnetic Field at Earth
2
1840
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
Year
32
Variation of ‘Open Flux’
Since we can also estimate solar wind speed from geomagnetic indices
[Svalgaard & Cliver, JGR 2007] we can calculate the radial magnetic flux
from the total B using the Parker Spiral formula:
Radial Component of Heliospheric Magnetic Field at Earth
6
Br nT
5
Ceiling
4
R2 = 0.0019
3
2
Floor
1
Year
0
1830
1840
1850
1860
1870
1880
1890
1900
1910
1920
1930
1940
1950
1960
1970
1980
1990
2000
2010
There seems to be both a Floor and a Ceiling and most importantly no longterm trend since the 1830s.
33
Floor and Ceiling of Solar Wind
Alfvénic Mach number
Solar Cycle Variation of Solar Wind Alfvenic Mach Number at Earth
500
Alfven Mach number
450
400
16
14
Ceiling
12
350
10
300
250
200
8
Floor
150
6
0.5
MA = V Np /(20 B)
Sunspot Number
4
100
2
50
0
1750
250
20
21
22
23
24
0
1800
1850
1900
1950
2000
2050
2100
2150
Bartels Rotation
2200
2250
2300
2350
2400
2450
8
Observations
seem to suggest that the magnitude of the solar cycle
variation
SSN
Plasma Beta
7
is invariant, i.e. does not depend on the size of the cycle. In particular, that the
200
value
at solar minimum is the same, ~12.25, in every cycle.
6 34
5
OMNI Explanation of MA
Consider first the multi-species nature of the solar wind plasma: protons, alphas, electrons.
We use subscripts p, a and e for these. N is density, T temperature, V flow speed, m mass Let
Na = f*Np Ne = Np + 2*Na = Np*(1+2f) Mass density = mp*Np + ma*Na + me*Ne = mp*Np +
4*mp*f*Np = mp*Np * (1+4f) Thermal pressure = k * (Np*Tp + Na*Ta + Ne*Te) = k * (Np*Tp +
f*Np*Ta + (1+2f)*Np*Te) = k*Np*Tp * [1 + (f*Ta/Tp) + (1+2f)*Te/Tp] Flow pressure =
Np*mp*Vp**2 + Na*ma*Va**2 + Ne*me*Ve**2 = Np*mp*Vp**2 + f*Np*4*mp*Va**2 =
Np*mp*Vp**2 * [l + 4f*(Va/Vp)**2] Rewrite: Mass density = C*mp*Np Thermal pressure =
D*Np*k*Tp Flow pressure = E*Np*mp*Vp**2 Where C = 1+ 4f D = 1 + (f*Ta/Tp) + (1+2f)*Te/Tp
E = 1 + 4f*(Va/Vp)**2 Now, some issues. 1. f is typically in the range 0.04-0.05, although
there are significant differences for different flow types. 2. Ta/Tp is typically in the range 4-6. 3.
What about Te? Feldman et al, JGR, 80, 4181, 1975 says that Te is almost always in the
range 1-2*10**5 deg K. Te rises and falls with Tp, but with a much smaller range of variability.
Kawano et al (JGR, 105, 7583, 2000) cites Newbury et al (JGR, 103, 9553, 1998)
recommending Te = 1.4E5 based on 1978-82 ISEE 3 data. So we'll use Te = 1.4E5 deg K for
our analysis. 4. What about (Va/Vp)**2? We should probably let this be unity always. If we let
f=0.05, Ta=4*Tp, Va=Vp, and Te=1.4*10**5, we'd have C = 1.2 D = 1.2 + 1.54E5/Tp E = 1.2
Characteristic speeds: Sound speed = Vs = (gamma * thermal pressure / mass density)**0.5
= gamma**0.5 * [D*Np*k*Tp /C*mp*Np]**0.5 = gamma**0.5 * (D/C)**0.5 *(k*Tp/mp)**0.5 With
the above assumptions for f, Ta, Va, and Te, and with gamma = 5/3, we'd get Vs (km/s) = 0.12
* [Tp (deg K) + 1.28*10**5]**0.5 Alfven speed = VA = B/(4pi*mass_density)**0.5 =
B/(4pi*C*mp*Np)**0.5 With the above assumptions, we'd get VA (km/s) = 20 * B (nT)/Np**0.5
and MA = V/Va = (V * Np**0.5) / 20 * B
35
For MA
= 7.5
at all
Maxima
Question:
Where would
the MHD
calculations
fall in this
diagram?
36
For MA
= 12.25
at all
Minima
MA =(V Np0.5)/(20 B)
The marks
the B = 4 nT
contour of the
‘Floor’ in HMF
37
‘Burning Prairie’ => Magnetism
Foukal & Eddy, Solar Phys. 2007, 245, 247-249
38
Relationship between Solar Radio Flux and Sunspot Number
240
Growing
Deficiency
of Sunspots
F10.7 sfu
220
SIDC 1996-2012
200
F = 1.0731 R + 59.2
r2 = 0.9497
180
160
Waldmeier 1947-1970
140
F = 0.9325 R + 55.0
r2 = 0.9938
120
100
80
Canonical Relationship SSN and F10.7
250
R SSN
SSN
60
0
20
40
60
80
100
120
140
160
180
200
1952-1990
200
150
Observed Sunspot Number Divided by Synthetic SSN (1951-1990)
1.4
600
100
y = -0.0000114x 3 + 0.0038145x 2 + 0.5439367x + 63.6304010
R2 = 0.9931340
50
1.2
R>10
500
F10.7
0
0
1.0
20
40
60
80
100
120
140
160
180
400
0.8
0.6
?
200
0.4
0.2
0.0
300
R
100
R
19
20
21
22
23
24
0
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
39
200
Deficit of Small Spots
Large
Small
Lefevre & Clette, SIDC
40
The Livingston & Penn Data
Temp.
From 2001 to 2012 Livingston and Penn have measured field strength and brightness at the
darkest position in umbrae of 1843 spots using the Zeeman splitting of the Fe 1564.8 nm line.
Most observations are made in the morning [7h MST] when seeing is best. Livingston
measures the absolute [true] field strength averaged over his [small: 2.5″x2.5″] spectrograph
aperture, and not the Line-of-Sight [LOS] field.
41
Umbral
Intensity
[Temperature]
and Magnetic
Field
Intensity
1
cycle 24
0.8
0.6
B
Gauss
cycle 23
0.4
1000
1500
2000
2500
3000
3500
4000
42
Evolution of
Distribution
of Magnetic
Field
Strengths
Distribution of Sunspot Magnetic Field Strengths
2005-2008
2009-2011
1000
1250
1500
1750
Sunspots form by assembly
of smaller patches of
magnetic flux. As more and
more magnetic patches fall
below 1500 G, fewer and
fewer spots will form
1998-2004
2000
2250
2500
2750
3000
3250
3500
3750
Gauss
43
Calibration
Change
MWO Plage
Strength Index
We see
fewer
sunspots
for given
MPSI
Cycle variation
and Trend
?
44
Working Hypothesis
• The Maunder Minimum was not a deficit of
magnetic flux, but
• A lessening of the efficiency of the process
that compacts magnetic fields into visible
spots
• This may now be happening again
• If so, there is new solar physics to be
learned
45