Variations of Near-UV and Visual Solar
Spectral Irradiance as Measured by
VIRGO/SoHO and PREMOS/Picard
Werner Schmutz, Alexander Shapiro, and Christoph Wehrli
Talk based on Wehrli et al. 2013, A&A 556, L3:
”Correlation of spectral solar irradiance with solar activity
as measured by VIRGO”
SORCE Science Meeting Cocoa Beach, Florida
January 28-31, 2014
Introduction I
• Disagreeing Spectral Solar Irradiance
measurements – used as forcing –
produce different responses of the
terrestrial atmospheric chemical
composition
• Therefore, it is important to assess the
amplitudes (including the signs!) of the
variations as a function of wavelength.
from Ermolli et al. 2013
Introduction II
• With
VIRGO/SOHO
PREMOS/PICARD
has filter radiometer
observations we
can confirm the
observations
since
amplitudes
1996 in the
of the
visual
rotational
and
time scale
near
IR
of SIM and SOLSTICE
• measurements
It is possible to verify SORCE spectral
measurements in these pass bands
• No confirmation of longer time scales
• However,
the PREMOS/PICARD
sensitivity change of filter
possible with
radiometers
in timeuse
is not
really
 thus, no further
of PREMOS
data in
understood
…
this talk
from Ermolli et al. 2013
Overview
• Some words of caution about filter
radiometers;
• VIRGO/SoHO SPM data and
“as-little-as-possible” degradation
correction;
• Results for rotational time scales;
• Results for variation on an
activity-cycle time scale
28.01.2014
Werner Schmutz
4
Degradation rates on 4 S/C
1st 200 exposure days
Ratio to 1st Light
IPHIR on Phobos2 (1988) & SOVA on EURECA (1992)
VIRGO on SOHO (1996)
1
1
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0
335/500/862 nm
0
50
100
150
0.2
200
0
402/500/862 nm
0
SOVIM on SSI (2008)
1
4
6
Exposure time [d]
8
200
0.8
Ratio to 1st Light
28.01.2014
2
150
1
0.6
0.4cleanliness comes close to
Picard
SOHO
0.2
402/500/862 nm
0
100
PREMOS FR-A on Picard (2010)
PREMOS 535, 782 differs not much
from 0.8
VIRGO 500, 862, whereas 210
0.6
resembles
IPHIR 335.
0
50
0.4
0.2
10
0
210/535/782 nm
0
Werner Schmutz
50
100
150
Exposure time [d]
200
5
SOVA2/EURECA (1992)
Instruments returned after flight
Wehrli et al., 1995, Metrologia
Right: 335nm filter band pass degraded the same
way for both operational SPM-A and backup
SPM-B;
Left: Rapid solarization of UG1 blocking glass!
28.01.2014
Werner Schmutz
6
Sensitivity changes of
VIRGO filter radiometers
2
1.8
spectral irradiance [Wm -2nm-1]
1.6
1.4
Backup (B) instruments
1.2
1
0.8
67%
0.6
0.4
Operational (A) instruments
17%
0.2
4%
0
1996
28.01.2014
1998
2000
2002
2004
2006
Werner Schmutz
2008
2010
2012
2014
7
Sensitivity changes of VIRGO
backup filter radiometers
1.010
1.010
862 nm
rel. Variation
1.005
1.000
1.000
0.995
0.995
0.990
1996
2000
2004
2008
0.990
1996
2012
1.010
VIRGO
A
rel. Variation
1.000
0.995
28.01.2014
2004
2008
2012
rel. Variation (times 5)
400 nm
2000
2000
2004
2008
2012
1.010
1.005
0.990
1996
500 nm
1.005
TSI x 5 !!
1.005
1.000
0.995
0.990
1996
Werner Schmutz
2000
2004
2008
2012
8
Simple fit to sensitivity change
1.010
1.008
1.006
Sum of two exponential curves
y1 = 1.006*exp(-(t-1996)/1689)
• Measurements should be left as they
are, but …
1.004
rel. Variation
1.002
 … difficult to learn from the uncorrected
level 2 data
1.000
0.998
0.996
 Thus, correct,
BUT as little as possible !
0.994
y2 = 1 - 0.008*exp(-(t-1996)/ 1.3)
0.992
0.990
1996
28.01.2014
2000
2004
Werner Schmutz
2008
2012
9
Simple fit to sensitivity change
1.010
1.008
1.006
Sum of two exponential curves
y1 = 1.006*exp(-(t-1996)/1689)
Time scale
1689 yr
• Use data after
6 years in space,
i.e. 2002 - 2014
1.004
rel. Variation
1.002
1.000
0.998
• Linear degradation
correction only
0.996
0.994
y2 = 1 - 0.008*exp(-(t-1996)/ 1.3)
Time scale
1.3 yr
0.992
0.990
1996
28.01.2014
2000
2004
Werner Schmutz
2008
2012
10
rel. Variation
rel. Variation
Linearly detrended
SPM-B measurements
28.01.2014
1.002
1.002
1.000
1.000
0.998
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
0.996
2002 2004 2006 2008 2010 2012 2014
1.002
1.002
1.000
1.000
0.998
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
0.996
2002 2004 2006 2008 2010 2012 2014
Werner Schmutz
11
Correlation of SPM-B
measurements with TSI
28.01.2014
Werner Schmutz
12
Correlation between TSI and SSI (SIM)
Data 21/04/2004 - 22/08/2011
MONTHLY SNAPSHOTS DATA
DAILY DATA
0.6
0.8
Correlation
Correlation
0.6
0.4
0.2
400
-0.2
1000
200
600
800
Wavelength [nm]
0.0
-0.5
0.4
0.2
rotational
cycle
0.0
400
600
800
Wavelength [nm]
1000
0.6
Correlation
Correlation
activity
cycle
600
800
Wavelength [nm]
0.8
1.0
0.5
400
DAILY - SMOOTHED DATA
81-DAY SMOOTHED DATA
-1.0
200
28.01.2014
0.2
0.0
0.0
-0.2
200
0.4
-0.2
200
1000
Werner Schmutz
400
600
800
Wavelength [nm]
1000
13
Solar rotational time scale
Irradiance at 215 nm
28.01.2014
Werner Schmutz
14
Correlation of SPM-B
measurements with TSI
28.01.2014
Werner Schmutz
15
rel
0.9995
A&A 556, L3 (2013)
0.9990
1.001
862 nm
1
1.0005
2002
2004
SSI−500 variation
SSI−862 variation
1.0005
1.001
500 nm
2006
1
2008
2010
Fig. A.2. The same as Fig. 3 but for annual mean values.
0.9995
0.9995
402 nm
1.0005
SSI−402 variation
1.001
2012
1
0.9995
Table A.1. Slopes of the linear regression (shown in Fig. 4) between rel0.999
0.999
0.999
0.9995in the
1 annual
1.0005
1.001
0.999the0.9995
1
1.0005
0.999
0.9995
1
1.0005
ative variations
means
of
VIRGO
TSI
and1.001
SSI 0.999
monthly
TSI variation
TSI variation
TSI variation
snapshots.
1.001
Wehrli
et al.regressions,
2013
Fig. 4. VIRGO SSI (RGB) versus TSI variations based on annual means from 2002 to 2012. Solid lines indicate from
the robust
linear
the
dash dot and dashed lines are, respectively, 1σ and 2σ uncertainties.
CAUTION:
Channel
Slope
2-σ uncertainty
1-σ uncertainty
measurements
of the total
solar irradiance
used as a proxy for
Based on monthly R
snapshots, positive
correlations
of SSI
slope values
might
solar activity. Monthly backup measurements of the SPM-B in- with TSI were found at all three wavelengths. In the analysis
be862
affected
SSI-862
−0.71
0.13]
−0.34]
strument
are affected by
small, but [−1.55,
not negligible,
ageing ef-[−1.08,
of annual
averages, the0.22
correlations at
nm and atby
402the
nm besubtraction
of
the
fects ofSSI-500
the instrument that
mask the signature
solar activity. [1.37,
come negative.
is more
1.65
[1.02, of
2.27]
1.92] A negative
0.79correlation for the IR channel
2
We note that there are several processes (e.g. shift of the peak likely real (R2 = 0.22) than for thelinear
UV channel
(R
=
0.01),
trend
SSI-402
−0.57
[−1.45,
0.32]
[−0.96,
−0.17]
0.01
wavelengths of the filters due to the diffusion of the filter lay- but both are not significant at the 2-σ level. However, the robust
ers, development of spectral leaks, radiation damage of the sili- (R2 = 0.79) positive correlation of monthly snapshots at 500 nm
con detector, etc.) that could influence the ageing and, in princi- remains statistically significant at the 5-σ level after smoothple, would make it a non-monotonic function of time. Presently ing out the solar rotation period. This result based on VIRGO
there
no reliable way to to
quantitatively
characterize these
pro- degradation
observations clearly
contradicts previous reports (Harder et al.
andis comparable
the uncertainty
of the
correction.
cesses or even to confirm their existence in the VIRGO/SPM 2009) of anti-correlated spectral irradiance at the peak of the soTherefore,
arepotential
the most
sensitive
toenergy
small
deviadata.
Therefore, thethese
only waypoints
to take these
effects into
lar spectral
distribution
with the solar cycle.
account
introduce
a sophisticatedcorrection
empirical model
of theaccordingly
tionsisinto the
degradation
and
the least re16
Werner
Schmutz
28.01.2014
2
„Max-Min“ Irradiance Variation
Fig. 1 from Haigh et al. (2010)
28.01.2014
Werner Schmutz
17
Wavelength band variations
Example:
variability
does not
depend on
the
wavelength
8%
of TSI
23%
39%
of TSI of TSI
27%
of TSI
<4%
of TSI
neglected
28.01.2014
Werner Schmutz
from Ermolli et al. (2013)
18
Channel
SSI-862
SSI-500
SSI-402
period
= 104)
elated
Slope
−0.71
1.65
−0.57
2-σ uncertainty
[−1.55, 0.13]
[1.02, 2.27]
[−1.45, 0.32]
R2
0.22
0.79
0.01
1-σ uncertainty
[−1.08, −0.34]
[1.37, 1.92]
[−0.96, −0.17]
400-450
nm: VIRGO B
and comparable to the uncertainty of the degradation
correction.
nm: VIRGO G
Therefore, these points are the most sensitive to 450-700
small deviations in the degradation correction and accordingly the least reliable. When the data are restricted to the central cluster (within
a square region of 1 ± 0.0005), the slope is reduced to +0.561
[0.436, 0.686], but remains strictly positive.
The trend of annual-mean data shown in Fig. A.2 is apparent
to the eye: the green curve (500 nm) resembles the black (TSI)
while the red (862 nm) and blue (402 nm) lines almost correVIRGO B+Glevel 1.0. We
spond to images of TSI mirrored at normalization
note that the temporal position of the extrema of SSI variations NRLSSI
may be affected by the choice of the selected period. However
the sign of the correlation remains unaffected (see Sect. 3.1).
VIRGO R
67%
28.01.2014
42%
-16%
Werner Schmutz
7%
19
snapshots.
Channel
SSI-862
SSI-500
SSI-402
riod
104)
ated
Slope
−0.71
1.65
−0.57
2-σ uncertainty
[−1.55, 0.13]
[1.02, 2.27]
[−1.45, 0.32]
1-σ uncertainty
[−1.08, −0.34]
[1.37, 1.92]
[−0.96, −0.17]
R2
0.22
0.79
0.01
and comparable to the uncertainty of the degradation correction.
Therefore, these points are the most sensitive to small deviations in the degradation correction and accordingly the least reliable. When the data are restricted to the central cluster (within
a square region of 1 ± 0.0005), the slope is reduced to +0.561
[0.436, 0.686], but remains strictly positive.
The trend of annual-mean data shown in Fig. A.2 is apparent
to the eye: the green curve (500 nm) resembles the black (TSI)
VIRGO
while the red (862 nm) and blue (402 nm)
linesB+G
almost correspond to images of TSI mirrored at normalization level 1.0. We
note that the temporal position of the extrema of SSI variations
may be affected by the choice of the selected period. However
the sign of the correlation remains unaffected (see Sect.
3.1). R
VIRGO
SORCE
28.01.2014
Werner Schmutz
20
Channel
SSI-862
SSI-500
SSI-402
eriod
104)
lated
Slope
−0.71
1.65
−0.57
2-σ uncertainty
[−1.55, 0.13]
[1.02, 2.27]
[−1.45, 0.32]
1-σ uncertainty
[−1.08, −0.34]
[1.37, 1.92]
[−0.96, −0.17]
R2
0.22
0.79
0.01
and comparable to the uncertainty of the degradation correction.
Therefore, these points are the most sensitive to small deviations in the degradation correction and accordingly the least reliable. When the data are restricted to the central cluster (within
a square region of 1 ± 0.0005), the slope is reduced to +0.561
[0.436, 0.686], but remains strictly positive.
The trend of annual-mean data shown in Fig. A.2 is apparent
to the eye: the green curve (500 nm) resembles the black (TSI)
while the red (862 nm) and blue (402VIRGO
nm) lines
almost correB+G
spond to images of TSI mirrored at normalization level 1.0. We
note that the temporal position of the extrema of SSI variations
may be affected by the choice of the selected period. However
the sign of the correlation remains unaffected (see Sect. 3.1).
VIRGO R
28.01.2014
Werner Schmutz
NRLSSI
21
Channel
SSI-862
SSI-500
SSI-402
Slope
−0.71
1.65
−0.57
2-σ uncertainty
[−1.55, 0.13]
[1.02, 2.27]
[−1.45, 0.32]
1-σ uncertainty
[−1.08, −0.34]
[1.37, 1.92]
[−0.96, −0.17]
R2
0.22
0.79
0.01
and comparable to the uncertainty of the degradation correction.
Therefore, these points are the most sensitive to small deviations in the degradation correction and accordingly the least reliable. When the data are restricted to the central cluster (within
a square region of 1 ± 0.0005), the slope is reduced to +0.561
[0.436, 0.686], but remains strictly positive.
The trend of annual-mean data shown in Fig. A.2 is apparent
to the eye: the green curve (500 nm) resembles the black (TSI)
while the red (862 nm) and blue (402VIRGO
nm) lines
almost correB+G
spond to images of TSI mirrored at normalization level 1.0. We
note that the temporal position of the extrema of SSI variations
may be affected by the choice of the selected period. However
the sign of the correlation remains unaffected (see Sect. 3.1).
VIRGO R
• VIRGO/SOHO and SIM/SORCE
measurements are in contradiction to
each other in the visible part of the solar
spectrum
period
= 104)
elated
• But, taken into account the uncertainties
of the VIRGO-TSI correlations, this does
SORCE
not imply that the large UV-amplitudes
of
the SORCE measurements are wrong
28.01.2014
Werner Schmutz
22
Stellar observations
Most of stellar data are in Strömgern (b+y)/2
b
28.01.2014
Werner Schmutz
y
23
Stellar brightness variations
versus activity
positive correlation
(faculae-dominated stars)
The Sun is very close to the
threshold of the transition
The idea is not new. See for example Preminger
between positive and
et al. (2011):
negative correlation.
“Therefore, we should not be surprised to find
Thus, the small residual could
Sun-like stars for which activity and
be either, positive or negative.
photospheric radiative flux are anticorrelated, as they are for the Sun. How,
negative
correlation
then, can we explain
that some
Sun-like stars
(spot-dominated
stars)
exhibit activity directly
correlated with
photospheric brightness? It could be due to
the angle from which we view these stars”
Fig. from Lockwood et al. (2007)
28.01.2014
Werner Schmutz
24
Activity models for stars
Result:
the
Sun is faculae-dominated
Stellar measurements from Lockwood et al. (2007)
independently on the viewing angle
Shapiro et al. (2014) submitted:
SATIRE model extrapolated to stars with
differentislevels
of magnetic
The conclusion
backed
up by the activity
entire parameter
space of magnetic activities
28.01.2014
Werner Schmutz
25
Conclusions
 The “simple” question if the Spectral Solar
Irradiance is positively 😋 or negatively 😕
correlated with Total Solar Irradiance is
answered by VIRG/SOHO as:
 500 nm is clearly positive !
28.01.2014
Werner Schmutz
26
Green 500 nm
1.002
1.002
500 nm
1.010
rel. Variation
Level 2 uncorrected
1.000
1.000
0.998
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
0.996
2002 2004 2006 2008 2010 2012 2014
1.002
1.002
1.000
1.000
1.005
0.995
0.990
1996
rel. Variation
1.000
2000
0.998
2004
2008
2012
Year
0.996
2002 2004 2006 2008 2010 2012 2014
TSI
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
1.010
28.01.2014
Werner Schmutz
27
Conclusions
 The “simple” question if the Spectral Solar
Irradiance is positively 😋 or negatively 😕
correlated with Total Solar Irradiance is
answered by VIRG/SOHO as:
 500 nm is clearly positive;
 862 nm is straight, i.e. a very small amplitude
(SIM/SORCE “ok”);
 400 nm is difficult, possibly negative but
maybe, also an additional instrumental
effect.
28.01.2014
Werner Schmutz
28
rel. Variation
rel. Variation
Linearly detrended
SPM-B measurements
28.01.2014
1.002
1.002
1.000
1.000
0.998
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
0.996
2002 2004 2006 2008 2010 2012 2014
1.002
1.002
1.000
1.000
0.998
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
0.996
2002 2004 2006 2008 2010 2012 2014
Werner Schmutz
29
Conclusions
 VIRG/SOHO:
 500 nm is clearly positive;
 862 nm is straight, i.e. very small amplitude;
 400 nm is difficult, possibly negative but
maybe also an additional instrumental
effect.
 Thus, published SIM/SORCE long-term trends
are NOT confirmed by VIRGO/SOHO and most
likely incorrect – in the visible
 But:
Large UV amplitudes are not excluded !
28.01.2014
Werner Schmutz
30
Thank you
for your attention
PICARD
SOHO
0.1
What determines the spectral profile of the solar irradiance variability?
F - C [W m-2nm-1]
0.0
200
0.15
0.10
300
400
500
600
700
Brightness contrasts of faculae and quiet Sun (disk averaged)
CN violet
system
LTE
CH G band
NLTE
0.05
0.00
200
300
400
500
Wavelength [nm]
600
700
NLTE effect in H-
SORCE data vs. Models
Rotational variability
Activity trends
from Ermolli et al. (2013)
also Ball et al. (2011), Lean & Deland (2012), Unruh et al. (2012)
Can the modelers change the 11-year variability without spoiling the rotational one?
In principle, YES
noise to the
rotational
variability
possible contribution to
the activity trends
Does the level of the quiet Sun activity change with time?
Hanle effect diagnostics allows one
to go beyond the spatial resolution
of the available magnetogramms
L. Kleint et al.: Solar turbulent magnetic fields: Non-LTE modeling of the Hanle effect in th
differential effect (very robust)
line ratios depend on the magnetic
field
from Kleint et al. 2011
turbulent magnetic field, G
BUT no clear trend in 2011-2013
data (Rameli et al. 2014)
there is a trend (3σ)
from Kleint et al. 2011
REMARK: The exact value of the magnetic field depends on the number of strong
assumptions since a tricky NLTE radiative transfer calculations are involved.
Nevertheless, the conclusion about the trend is robust.
presently there is NO evidence for a link between the irradiance and turbulent
magnetic field
VIRGO and PREMOS
rel. Variation
VIRGO:865 nm
VIRGO:500 nm
1.002
1.002
1.000
1.000
0.998
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
0.996
2002 2004 2006 2008 2010 2012 2014
rel. Variation
PREMOS:607 nm
PREMOS:607 nm (/annual)
1.002
1.002
1.000
1.000
0.998
0.998
0.996
2002 2004 2006 2008 2010 2012 2014
0.996
2002 2004 2006 2008 2010 2012 2014
PREMOS 607 nm
normalized to mean (y / <y>)
1.002
1.001
1
0.999
0.998
2010
2011
2012
2013
2014
2013
2014
normalized to linear (y / A*t + B)
rel. Variation
1.002
1.001
1
0.999
0.998
2010
2011
2012
normalized to lin-sin (y / A*t + B + C*sin(2pi*t/365 + D))
1.002
1.001
1
0.999
0.998
2010
2011
2012
2013
2014