observation and numerical simulation of propagation of ulf ... · department of earth and planetary...
TRANSCRIPT
Observation and Numerical Simulation of Propagation of ULF Waves From the Ion Foreshock Into the
Magnetosphere
Kazue Takahashi(1), Lucile Turc(2), Emilia Kilpua(2), Naoko Takahashi(3), Andrew Dimmock(4), Primoz Kajdic(5), Minna Palmroth(2), Yann Pfau-Kempf(2), Jan Soucek(6),
Howard Singer(7), Tetsuo Motoba(1), and Craig Kletzing(8)
1. The Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, USA2. Department of Physics, University of Helsinki, Helsinki, Finland
3. Department of Earth and Planetary Science, Graduate School of Science, The University of Tokyo, Tokyo, Japan4. Swedish Institute of Space Physics (IRF), Uppsala, Sweden
5. Instituto de Geofísica Universidad Nacional Autónoma de México, Ciudad Universitaria, CDMX, México6. Institute of Atmospheric Physics, Academy of Sciences of the Czech Republic, Prague, Czech Republic
7. Space Weather Prediction Center, NOAA, Boulder, Colorado, USA8. Department of Physics and Astronomy, University of Iowa, Iowa City, Iowa, USA
Paper EGU2020-5562
EGU 2020 General Assembly, 3-8 May 2020
Abstract
2020-05-06 EGU
Observational studies have demonstrated that ULF waves excited in the ion foreshock are a main source of Pc3-4 ULF waves detected in the magnetosphere. However, quantitative understanding of the propagation of the waves is not easy, because the waves are generated through a kinetic process in the foreshock, pass through the turbulent magnetosheath, and propagate as fast mode waves and couple to shear Alfven waves within the magnetosphere. Recent advancement of hybrid numerical simulations of foreshock dynamics motivated us to analyze observational data from multiple sources and compare the results with simulation results. We have selected the time interval 1000-1200 UT on 20 July 2016, when the THEMIS, GOES, and Van Allen Probe spacecraft covered the solar wind, foreshock, magnetosheath, and magnetosphere. The EMMA magnetometers (L=1.6-6.5) were located near noon. We found that the spectrum of the magnetic field magnitude (Bt) in the foreshock exhibits a peak near 90 mHz, which agrees with the theoretical prediction assuming an ion beam instability in the foreshock. A similar Bt spectrum is found in the dayside outer magnetosphere but not in the magnetosheath or in the nightside magnetosphere. On the ground, a 90 mHzspectral peak was detected in the H component only at L=2-3. The numerical simulation using the VLASIATOR code shows that the foreshock is formed on the prenoon sector but that the effect of the upstream waves in the magnetosphere is most pronounced at noon. The Bt spectrum of the simulated waves in the outer magnetosphere exhibits a peak at 90 mHz, which is consistent with the observation..
2
Magnetic cloud on 20 July 2016
2020-05-06 EGU 30000 0300 0600 0900 1200 1500 1800 2100
UT
MAS (L=6.4)10 s samples
1000
500
0
40
20
0
-40-20
02040
-40-20
02040
-40-20
02040
100
50
0
906030
0
-500
0
10050
0-50
302010
0
201510
50
201510
50
V
(km
/s)
n p(c
m-3
) B
x(n
T) B
y(n
T) B
z(n
T) B
t(n
T) δ
B (n
T) θ
xB(d
eg)
Rbs
(RE)
Rm
p(R
E)SY
M-H
(nT
) A
L(n
T)
H
19-20 July 2016
GSM
200
nT
CME
GSM
GSM
2-h interval for data analysys
• Steady IMF– Magnitude ~15 nT– Cone angle < 30�
• Multiple spacecraft– THEMIS, GOES, and Van
Allen Probes
• Ground magnetometers– EMMA– N = 23– L = 1.6–6.5
20 July 20161000-1200 UT
403020100-10-20
Y SM
(RE)
30
20
10
0
-10
-20
-30
XSM(RE)
20151050-5-10
15
10
5
0
-5
-10
-15
Bow shock
Magneto-pause
TH-E
TH-DTH-A
VA-A
VA-B
G-13G-14
G-15
EMMA
Dawn
Dusk
XSM(RE)
Y SM
(RE)
(a)
(c)
(b)
Density radial profile
2020-05-06 EGU
(a)
(b)1000-1100 UT
fT1
(mHz
) ~13 h MLT
EMMA
101
102
ρeq
(am
u.cm
-3)
101
102
103
104
105
1110987654321L
20
15
10
SC p
oten
tial
(V)
THEMIS-A 0800-1400 UT
20 July 2016
1000 UT 1200
(c)
Plume
4
• THEMIS-A spacecraft potential– Plume in the outer
magnetosphere
• Magnetoseismic analysis of EMMA data– Smooth plasmaspheric mass
density profile at L < 6
Dayside spacecraft observations
2020-05-06 EGU 5
• Foreshock/sheath – Broadband– No strong peak at
theoretical frequency !"#
• Magnetosphere– Broadband – Cross-phase indicates
instances of tailward propagation
20 July 201620
10
0
Bt
(nT)
OMNI
1510
50
R mp
(RE)
THEMIS-A200
100
0
B t (nT)
1000 1030 1100 1130 1200UT
15010050
0
fuw
(mHz
)
906030
0
θxB
(deg
)
36027018090
0
φyz
B(d
eg)
THEMIS-D
n
i(c
m-3
)
100
50
0
B t (nT)
906030
0
10-1
101
103
θxB
(deg
)
THEMIS-E
n
i(c
m-3
)
10-1
101
103
100
50
0
B t (nT)
906030
0
θxB
(deg
)
MagnetosheathForeshock
MagnetospherePlume
10.59 1.3
10.22
10.89 2.2
10.32
11.17 3.0
10.41
11.41 3.8
10.50
11.63 4.5
10.59
LMLATMLT
PS
D(n
T2/H
z)
Freq
uenc
y
(mHz
)
10 110 210 310 4
PS
D(n
T2/H
z)
10 -110 010 110 2
0
50
100
150
Freq
uenc
y
(mHz
)
Bt
20 July 2016THEMIS-D
11.34 -4.7
11.71
11.07 -3.8
11.80
10.77 -2.9
11.89
10.44 -2.0
12.00
10.09 -1.1
12.12
LMLATMLT
10 110 210 310 4
PS
D(n
T2/H
z)
0
50
100
150
Freq
uenc
y
(mHz
)
Bt
PSD
((mV\
m)2 /
Hz)
THEMIS-A
10 -410 -310 -210 -1
0
50
100
150 Eφ
1000
9.14 -8.7
12.301030
8.54 -8.3
12.451100
7.89 -7.9
12.621130
7.18 -7.6
12.841200
6.41 -7.6
13.10UT
LMLATMLT
-180
0
180
Cros
s-ph
ase
(
deg)
0
50
100
150 Eφ−Bt
Foreshock M’sheath
0
50
100
150THEMIS-E
Bt
PlumeInstumental
(b)(a)
!"# mHz= 7.6+, nT cos2345
Foreshock waves just outside the magnetosheath
2020-05-06 EGU 6
20 July 2016
-50
0
50
Bz
(nT)
(a)
100
50
0
Bt
(nT)
1100 1102 1104 1106 1108 1110UT
(b)
(c)PS
D(n
T2/H
z)
Frequency (mHz)
Bz
Bt
20 50 200 500
80 mHz
10010
GSE
THEMIS-E
100
101
102
103
• Large amplitude• Transverse
component (Bz) Broad peak at 80 mHz(~fuw)
• Compressional component (Bt)– Strong power at
30-40 mHz
Comparison of waves between space and ground
2020-05-06 EGU
1000 9.14 -8.7
12.30
1020 8.75 -8.4
12.39
1040 8.33 -8.1
12.50
1100 7.89 -7.9
12.62
1120 7.42 -7.7
12.76
1140 6.93 -7.6
12.92
1200 6.41 -7.6
13.10
UTLMLATMLT
(a)
(b)
(c)
MLTSOD 12.80 13.13 13.45 13.78 14.11 14.44 14.77MLTPPN 11.98 12.31 12.63 12.96 13.29 13.62 13.95
2July 20 2016
10 -210 -110 010 110 2
0
50
100
150 THA Bt
10 -210 -110 010 110 2
0
50
100
150 SOD H (L = 5.4)
10 -210 -110 010 110 2
0
50
100
150 PPN H (L = 2.2)
0.0
0.5
1.0
Cohe
renc
e
0
50
100
150 THA -SOD
0.0
0.5
1.0
Cohe
renc
e
0
50
100
150 THA-PPN (e)
(d)
Freq
uenc
y
(mHz
)
PS
D(n
T2/H
z)
PSD
(nT2
/Hz)
PS
D(n
T2/H
z)
Freq
uenc
y
(mHz
)Fr
eque
ncy
(m
Hz)
Freq
uenc
y
(mHz
)Fr
eque
ncy
(m
Hz)
Plume
7
• Power spectra – not similar
between space and ground
• Coherence– Low at all
frequencies
EMMAgram
2020-05-06 EGU 8
THEMIS-A and EMMA Magnetometers 20 July 2016
T_boxcar = 21 s
1.05
0.00
-1.05 1.05
0.00
-1.05
L
1 2 3 4 5 6 7
L
1 2 3 4 5 6 7
1000 1005 1010 10151233 1238 1243 1247
UTMLT
H (nT)
D (nT)
(b)
(c)
0.90
0.00
-0.90 0.90
0.00
-0.90
L
1
2
3
4
L
1
2
3
4
1005 1006 1007 1008 1009 10101152 1153 1154 1155 1156 1157
UTMLT
D (nT)
H (nT)
(e)
(f)
-505
Bt
(nT)
-505
Bt
(nT)
(a)
(d)
PPNZAG
PPNZAG
• H component– Poleward phase
propagation – Transient
pulsations at low L
• D component– Weaker poleward
propagation– No transient
pulsations
Cross-phase analysis
2020-05-06 EGU 9
• Power spectra
– Peaks at ~30 and
~90 mHz
– The higher
frequency matches
fuw
• Cross spectra
– T1, T2, and T3
waves detected
(a)
(b)
(c)
PPN (L = 2.21)
PSD
(nT2
/Hz)
PSD
(nT2
/Hz)
Cro
ss ph
ase
(deg
)
EMMA 20 July 2016 H component Midpoint L = 2.16
100011.9811.79
103012.4712.28
110012.9612.77
113013.4513.26
120013.9513.76
UTMLTPPNMLTZAG
10 -210 -110 010 110 2
10 -210 -110 010 110 2
-90
0
90
0
50
100
150
0
50
100
150
0
50
100
150
ZAG (L = 2.11)
PPN - ZAG
Freq
uenc
y
(mHz
)
PPN (L = 2.21)
T3
T1
T2
Comparison with previous studies
2020-05-06 EGU
5 September 2002 1830-1850 UT (Clausen et al. 2009)
20 July 2016 1000-1020 UT (This study)
Cluster-3(L = 4.5)
Geotail(Foreshock)
(a)
(b)
(i)
(h)
(g)
(f)
(e)
(d)
(c)
(j)
PINA (4.1)
10-4
10-2
100
102
104
10 100Frequency
(mHz)
GILL(6.2)
10-4
10-2
100
102
104
FCHU (7.4)
HD
10-4
10-2
100
102
104
PSD
(nT2 /H
z)
PPN (2.2)
10-4
10-2
100
102
104
10 100Frequency
(mHz)
MEK (3.9)
10-4
10-2
100
102
104
KIL (6.3)
10-4
10-2
100
102
104
10-4
10-2
100
102
104
10-4
10-2
100
102
104
Bt
10-4
10-2
100
102
104
THEMIS-D(Foreshock)
10-4
10-2
100
102
104
THEMIS-A(L = 8.9)
Bt
Trace
Trace
T1T2 T3
41 mHz88 mHz
Mag
netic
field
trac
e PS
D
(
nT2 /H
z)
Frequency (mHz)
THEMIS-E20 July 2016 1100-1120 UT
ISEE-1 (Le and Russell,1992) 28 Dec 19861: 1030-1050 UT2: 1400-14203: 1950-2010
12
3
1(f×4, p×3)
100
101
102
103
104
100 101 102 103
10
• Comparison with ISEE study– Higher frequency– Higher power– Broader spectral peak
• Comparison with Cluster study– No single outstanding spectral
peak
Local time dependence of wave power
2020-05-06 EGU 11
Bt component
Van Allen Probe B
20 July 20161000-1200 UT
Freq
uenc
y
(mHz
)
PSD(nT2/Hz)
Van Allen Probe A
0.01
0.1
1
RMS
ampli
tude
(nT
)
10:002016-07-20
10:30 11:00 11:30 12:00UT
THEMIS-A
70-110 mHz
GOES-13
GOES-14
GOES-15
10 -210 -110 010 1
0
50
100
150 THEMIS-A
10 -210 -110 010 1
0
50
100
150 GOES-13
10 -210 -110 010 1
0
50
100
150 GOES-14
10 -210 -110 010 1
0
50
100
150 GOES-15
10 -210 -110 010 1
10 -210 -110 010 1
0
50
100
150
0
50
100
150
20151050-5-10
15
10
5
0
-5
-10
-15
Bow shock
Magneto-pause
TH-A
VA-A
VA-B
G-13G-14
G-15
Dawn
Dusk
YSM(RE)
X SM
(RE)
• Nightside waves– Much weaker than
dayside waves– No indication of
upstream wave propagation
Vlasiator simulation
2020-05-06 EGU 12
• Magnetosphere is included
– Magnetopause at 5 Re
– Realistic Alfven velocity
Proton density
Alfven velocity
Simulated waves
2020-05-06 EGU
1.0
0.5
0.0
cohe
renc
e
1.0
0.5
0.0cohe
renc
e
10 100Frequency (mHz)
10-3
10-2
10-1
100
101
102
VS-1vsVS-2
VS-1vsVS-3
(e)
(d)
(c)
(b)
(f)
VS-3. (3.2, 3.8) VS-2. (4.9, -0.9)
VS-1. (4.7, -1.7)(a)
600550500450400350Time(s)
Bz
2 nT
Bz P
SD(n
T2/H
z)20 50 200 500
VS-1
VS-2
VS-3
VS-1VS-2
VS-3
13
• Foreshock– Short coherence
length– Multiple wave
packets
• Magnetosphere– Wave power peaks
at noon– Low coherence
between noon and dusk
– Rapid power decrease toward dawn/dusk
Summary
2020-05-06 EGU
• A magnetic cloud in the solar wind produces an ion foreshock populated with ULF waves with higher-than-usual frequencies and complex spectra.
• In the magnetosphere and on the ground, there are only weak signatures of oscillations directly driven by the foreshock waves.
• Numerical simulation indicates short spatial scales and a multiple frequencies of the foreshock waves.
14