Solar Rotational Modulations of SSI and
Correlations with the Variability of TSI
Jae N. Lee1,2 , Robert F. Cahalan2,3, and Dong L. Wu2
1.
2.
3.
Joint Center for Earth Systems Tech., University of Maryland, Baltimore County, MD
NASA Goddard Space Flight Center, Greenbelt, MD
Applied Physics Lab/Johns Hopkins University, Laurel, MD
1.
2.
TSI (TIM, ACRIM III, VIRGO)
SSI (SIM V19, SOLSTICE, TIMED/SEE, SATIRE-S)
were continuous and almost completely clean,
d of 220 days (7 months from 21 May 2007
ber 2007). Seven maxima and seven minima
daily noise values yield a time series where
ant frequency is 1/27 day (Figure 3).
of the fact that these changes in the APD
ng daytime, approximately from sunrise till
ected that these were related to the Sun and
tion. To validate this we compared our data
olar indices data. However, after we gathered
dices (Solar Radiation and Climate Experictral irradiance, 2007, available online http://
u/sorce/index.htm; NOAA National Weather
Weather Prediction Center, 2007, available at
http://www.swpc.noaa.gov/) (sunspot number, 10.7 cm radio
flux, Lyman alpha flux) we found out that none of these solar
indices, or any other measured quantity, revealed a continuous 27-day periodic fluctuation throughout this period.
[13] Figure 4 represents our daily measured VLF APD
(2 – 5 kHz) together with three different solar indices, which
do not always correlate well with each other. However,
when we compare our measured index when all three
indices agree, we notice that there is a strong negative
correlation between our VLF index and the solar ones. This
implies that during enhanced solar activity, i.e., with higher
values for the solar indices, we detect less atmospheric
noise, and vice versa when the solar indices are at minima.
However, unlike all the solar indices presented, our ELF-
Sun-­‐Climate connec/on in solar rota/onal /me scale Reuveni and Price (2009) Takahashi et al. (2014, SORCE mee/ng presenta/on) Cloud(varia5ons(
(OLR)
Africa
Lightening Lightning,in,MC
Pacific
America
Atlan/c
2003.8
variation 2007: TSI vs Lyman-,
4. Daily VLF
noise levels at rotational
2 – 5 kHz
versus different solar indices. When
0.2
0.2 there is a good
TSI
ment between all three solar indices (red, blue, and green), they all show negative
correlation with
0.15
Lyman-,
LF index (black).
0.1
0.05
0
0.1
4 of 7
0
-0.1
-0.15
-0.2
-0.2
0.8
TSI
0.6
Lyman-,
0.6
0.4
0.4
0.2
0.2
0
-0.05
-0.1
0.8
2004.7
rotational variation 2003/2004: TSI vs Lyman-,
0
-0.2
-0.2
-0.4
-0.4
-0.6
-0.6
-0.8
-0.8
-­‐  VLF lightening signal correlates with solar ac/vity only when the solar rota/onal signals are large. -­‐  There are 27-­‐day like signals in lightening ac/vity, but not always correlated with solar rota/onal signals. Sun Climate Symposium
Lee, Cahalan, and Wu
2
Savannah, GA, 11/11/15
TSI and SSI   SSI is a distribu/on of TSI, SSI varia/on is wavelength dependent, and also /me scale dependent. Solar spectrum originates at different levels of the solar atmosphere.   With current satellite observa/ons, more than 100 solar rota/onal cycles of 27-­‐day varia/on are covered (2003 – Aug. 2013).   Dis/nct solar rota/onal modula/ons of SSI at each wavelength can be iden/fied in terms of amplitude and phase.   The phase of the rota/onal mode obtained from SSI is compared with that of TSI. Lee, Cahalan, and Wu
3
Sun Climate Symposium
Savannah, GA, 11/11/15
Sta/s/cal analysis : EEMD (Ensemble Empirical Mode Decomposi/on) [Huang et al., 1998; Ruzmaikin et al., 2007; Barnhart and Eichinger, 2011] Low frequency varia/on High frequency varia/on Fourier Transform EEMD Hilbert-­‐Huang Transform with /me variable amplitudes and phases Lee, Cahalan, and Wu
4
Sun Climate Symposium
Savannah, GA, 11/11/15
The 27-­‐Day Rota/onal Varia/ons in TSI Observa/ons: from SORCE/TIM, ACRIMSAT/ACRIM III, and SOHO/VIRGO [Lee et al., 2015] Total Solar Irradiance (TSI) at 1 AU (W/m 2)
1363
1362
SORCE Total Solar Irradiance - daily average
maximum
declining SC23
1361
missing
minimum SC23
1360
1359
1358
1357
rising SC24
04
05
06
07
08
09
10
11
12
13
14
15
Year
-­‐  The rota/onal signals are well captured with three independent space-­‐borne TSI measurements. -­‐  They agree well with each other, with large amplitudes in declining and rising phases, but with small amplitudes in minimum phase. -­‐  How about SSI? Lee, Cahalan, and Wu
5
Sun Climate Symposium
Savannah, GA, 11/11/15
Rota/onal Varia/on : T: TSI
SI vsvUV
s L(121nm)
yman-­‐α (121 nm) rotational variation
10
5
10
2003
5
0
0
−5
−5
−10
−10
TIM/TSI 5
2004
SOLSTICE TIMED/SEE 5
0
0
−5
−5
5
2005
•  Standardized amplitudes are es/mated by removing the mean and by dividing with σ. 5
0
0
−5
−5
5
5
ATSI (t) =
0
0
−5
−5
5
2007
0
0
Lee, Cahalan, and Wu
−5
ITSI (t) − ITSI
σ (ITSI )
,
AUV (t) =
IUV (t) − IUV
σ (IUV )
•  The modes from two observa/ons of Lyman-­‐α line (SORCE/SOLSTICE and TIMED/SEE) agree well in amplitudes and phases. •  The modes of Lyman-­‐α line are not always in phase with TIM/TSI modes. •  TSI varia/ons are affected by solar proton event (SPE), but UV varia/ons are not. 2006
5
A few examples : How rota/onal varia/ons of SSI are related with that of TSI? −5
6
Rota/onal Varia/on : TSI vs Lyman-­‐α (121 nm) rotational variation : TSI vs UV
5
5
2008
0
0
−5
TIM/TSI SOLSTICE TIMED/SEE 5
−5
5
2009
0
0
−5
−5
5
5
2010
0
0
−5
−5
5
-­‐ Rota/onal varia/ons of TSI and Lyman-­‐α are ohen out-­‐
phase during high solar ac/vity period, but are in-­‐
phase during solar minimum (2008-­‐2010). 5
2011
0
0
−5
−5
5
5
2012
0
0
−5
5
−5
5
2013
0
0
7
−5
Lee, Cahalan, and Wu
−5
Sun Climate Symposium
Savannah, GA, 11/11/15
Rota/onal Varia/on : TSI vs MUV (240 nm) rotational variation : TSI vs UV (240nm)
10
5
10
2003
5
0
0
−5
−5
−10
−10
5
TIM/TSI 2004
SOLSTICE SIM 5
0
0
−5
−5
5
2005
Similar but TSI with MUV (240nm) - They are ohen out-­‐phase during high solar ac/vity period. -­‐ Modes from SIM and SOLSTICE match well. 5
0
0
−5
−5
5
5
2006
0
0
−5
−5
5
5
2007
0
0
−5
−5
Lee, Cahalan, and Wu
8
Sun Climate Symposium
Savannah, GA, 11/11/15
Rota/onal Vrotational
aria/on : TvsSI vs MUV (240 nm) variation : TSI
UV (240nm)
5
5
2008
0
0
−5
TIM/TSI SOLSTICE −5
SIM 5
5
2009
0
0
−5
−5
5
5
2010
0
0
−5
−5
5
-­‐ TSI and MUV are in phase during minimum phase, but ohen out-­‐phase during solar ac/vity rising period. 5
2011
0
0
−5
−5
5
5
2012
0
0
−5
−5
5
5
2013
0
0
−5
−5
Lee, Cahalan, and Wu
9
Sun Climate Symposium
Savannah, GA, 11/11/15
Rota/onal Vrotational
aria/on : TSI vs H-­‐α (656 nm) variation : TSI vs Hα 656.3nm
10
10
2003
0
0
−10
TIM/TSI 10
SIM −10
SIM unscreened 10
2004
0
0
−10
−10
10
10
2005
0
0
−10
- TSI and VIS are in-­‐phase during whole analysis period (2003 – 2011). -­‐ Only 5 outliers from SIM are excluded for the EEMD analysis to remove the outstanding cycles appeared in purple doied curve. -­‐ Outliers from SIM are chosen from high frequency mode comparison with TSI. −10
10
10
0
0
−10
−10
10
W/m2/nm 2006
10
2007
0
−10
Lee, Cahalan, and Wu
Slope = 0.0011
0
−10
10
Sun Climate Symposium
Savannah, GA, 11/11/15
Rota/onal Varia/on : TSI vs. H-­‐α (656 nm) 2008
rotational variation : TSI vs UV
5
5
0
0
−5
2009
TIM/TSI SIM SIM unscreened −5
5
5
0
0
−5
−5
2010
5
5
0
0
−5
TSI and SIM H-­‐α (656 nm) −5
2011
5
5
0
0
−5
−5
11
Lee, Cahalan, and Wu
Sun Climate Symposium
Savannah, GA, 11/11/15
Phase of the solar rota/onal modes : 2005 -­‐  Rota/onal varia/ons are well defined in UV and 600-­‐ 900 nm, but not at photospheric wavelength (250 – 600 nm) and IR above 900 nm. mode of TSI: 2005 mode of SSI: 2005 350
350
300
300
250
2005: day of year
days : 2005 days : 2005 rotational mode : 2005
200
150
100
50
-5
0
5
3
2
250
1
200
0
150
-1
100
-2
50
120 140 160 180 200 220 240
wavelength (nm)
SOLSTICE Lee, Cahalan, and Wu
12
400
600
800
1000
1200
1400
1600
SIM Sun Climate Symposium
Savannah, GA, 11/11/15
-3
Linear correla/ons of the 27-­‐day mode between TSI and SSI Correlation of the 27-day mode between SSI and TSI
1
SIM
SATIRE-S
SOLSTICE A
0.8
r
0.6
0.4
0.2
0
-0.2
100
150
200
SOL-A
250
wavelength (nm)
300
350
SOL-B
SIM
SATIRE-S
0.8
0.6
SIM
SATIRE
0.4
0.2
SIM
0
-0.2
400
Lee, Cahalan, and Wu
400
Correlation of the 27-day mode between SSI and TSI (400-2400 nm)
1
r
-­‐  Correla/ons begin to grow at 250 nm, correla/ons from SOLSTICE, SIM and SATIRE track each other. -­‐  Correla/ons from SIM is >0.8 in 600 – 900 nm. -­‐  Correla/ons from SATIRE are nearly 1 above 400 nm, but shih at opaque minimum at 1650 nm. SOLSTICE B
SATIRE
600
800
1000
13
1200
1400
wavelength (nm)
1600
1800
2000
2200
Sun Climate Symposium
Savannah, GA, 11/11/15
2400
Summary
  Solar rotational modulations of spectral irradiance are wavelength dependent.
  Rotational variations of VIS and NIR (~600-900 nm) are in-phase with that
of TSI, but variations of FUV and MUV are not always in-phase.
  Close agreement of the rotational variations from SIM and SOLSTICE is
encouraging. We can address rotational variations from SORCE SSI
measurements with high confidence.
  Rotational variations from SATIRE-S matches with observed SSI modes
below 280 nm, and exactly matches with TSI mode above 400nm. Lee, Cahalan, and Wu
14
Sun Climate Symposium
Savannah, GA, 11/11/15
Acknowledgment
  SORCE Team
  TIMED and SATIRE Teams
  Tom Woods, Jerry Harder, and Marty Snow
Lee, Cahalan, and Wu
15
Sun Climate Symposium
Savannah, GA, 11/11/15
Back ups
Lee, Cahalan, and Wu
16
Sun Climate Symposium
Savannah, GA, 11/11/15
Low frequency and High frequency varia/ons of TSI and Lyman-­‐α Low frequency varia/on (a) low frequency variation : TSI vs Lyman-,
1361.4
8
TIM/TSI
SOLSTICE/Lyman-,
1361.2
•
7.5
1361
6.5
W/m2 W/m
2
7
1360.8
1360.6
6
1360.4
1360
•
5.5
1360.2
01/04
01/05
01/06
01/07
01/08
01/09
01/10
01/11
01/12
#10
01/13
-3
5
•
HIgh frequency varia/on (b) high frequency variation : TSI vs Lyman-,
4
1.5
SOLSTICE/Lyman-,
TIM / TSI
3
1
2
0
-1
-0.5
-2
-1
-3
01/04
01/05
01/06
01/07
01/08
01/09
01/10
01/11
01/12
#10 -3
-1.5
01/13
W/m2 W/m2
0
-4
•
0.5
1
The phase of low and high frequency varia/ons of TSI and Lyman-­‐α line are compared. Amplitudes of high frequency varia/on Low frequency varia/ons are in-­‐
phase with each other. High frequency varia/ons? Distribu/on of low and high frequency modes 350
TSI HF mode : Histogram and fitted Normal curve
•  TSI HF mode is not normally distributed. •  SIM SSI difference between high and low TSI regime has a dip at 1600nm. 300
frequency
250
200
150
100
50
0
-4
-2
W/m2
-1
0
1
TSI HF mode : Histogram in high and low regime
160
40
140
35
120
30
100
80
25
20
60
15
40
10
20
5
0
-4
-3
Lee, Cahalan, and Wu
-2
W/m2
-1
0
Lyman , HF mode : Histogram
45
frequency
frequency
180
-3
1
0
-3
-2
-1
18
0
W/m2/nm
1
2
3
#10 -3
Sun Climate Symposium
Savannah, GA, 11/11/15
Eric Richard’s TSI vs SSI
I (λ ) = I 0 exp(− Ox ) ≈ I 0 (1 − Ox )
1. Relation between irradiance I and Opacity
à here O is opacity and x is distance times density. Note O depends on the
wave length lambda 2. Assume there are two main source of absorption such as O1: H- and O2:H2
à Then Lee, Cahalan, and Wu
I (λ ) = I 0 (1 − O1 x1 − O2 x2 )
19
Sun Climate Symposium
Savannah, GA, 11/11/15
3.  Let us see if the following fig makes sense
à the y axis is defined as d (λ ) =
I − I ref
I ref
=
=
Lee, Cahalan, and Wu
I 0 (1 − O1 x1 − O2 x2 ) − I 0 (1 − O1 x1 − O2 x2 )ref
I 0 (1 − O1 x1 − O2 x2 )ref
O1 (x1,ref − x1 ) + O2 (x2,ref − x2 )
(1 − O1 x1 − O2 x2 )ref
20
Sun Climate Symposium
Savannah, GA, 11/11/15
4. We can compare two wave length 1000 and 1600. d (1000) =
d (1600) =
O1 (1000)(x1,ref − x1 ) + O2 (1000)(x2,ref − x2 )
(1 − O1 (1000) x1 − O2 (1000) x2 )ref
O1 (1600)(x1,ref − x1 ) + O2 (1600)(x2,ref − x2 )
(1 − O1 (1600) x1 − O2 (1600) x2 )ref
5. Note in general lager fluctuation for lambda 1000 than 16000 because 1
1
>
(1 − O1 (1000) x1 − O2 (1000) x2 )ref (1 − O1 (1600) x1 − O2 (1600) x2 )ref
6. Also, depending on relative abundance (density ) of H- and H2, that is x1 and x2,
the sign of d(1000) and d(1600) can be different. That is while one grows , the other
could decrease. Lee, Cahalan, and Wu
21
Sun Climate Symposium
Savannah, GA, 11/11/15
SSI composite: TSI(high) – TSI(low)
Lee, Cahalan, and Wu
22
Sun Climate Symposium
Savannah, GA, 11/11/15
Lee, Cahalan, and Wu
23
Sun Climate Symposium
Savannah, GA, 11/11/15
SSI composite: TSI(asc) – TSI(des)
Lee, Cahalan, and Wu
24