Blowin in the Solar Wind Chillin in the Polar Cap

1
Blowin in the Solar Wind Chillin in the Polar
Cap

• Space Weather Prediction of the Polar Cap Index
• Lee Bargatze
• IGPP Seminar - March 15, 2005

2
What am I going to Say?

• Solar wind monitors can give at most about 1-hr
  advance warning of impending conditions about the
  Earth.
• Typically, a very simple is used to account for
  the delay between when a monitor measures in situ
  conditions and those conditions should be making
  an impact at the Earth.
• A new solar wind propagation model is available.
• Does it help predict activity? Yes but modestly.
• Demonstrate this using the PC Index.
• We cant start surfing till we learn to swim.
  Background!

3
Solar Source of Space Weather
From Tobias and Weiss 2005
From TRACE Website
4
Sunspot Number Magnetic Activity at Earth

• The sunspot number (yellow) and number of
  magnetically disturbed days (red) are highly
  correlated.

5
Solar Wind Starts Blowin near the Sun
All from Habbal and Woo 2005
6
One Month of Huffing and Puffing
From SOHO LASCO Website
7
Flare to Coronal Mass Ejection
From SOHO Website
8
The Solar Wind and the IMF

• The solar wind expands radially outward from the
  Sun
• The solar magnetic field is frozen into the solar
  wind and is stretched out yielding the IMF
• The feet of the field lines remain rooted in the
  rotating Sun. Rotation wraps the field into a
  spiral
• Planetary obstacles force the flow to divert
  around distorting the IMF

9
Coronal Mass Ejection can be Highly Geo
Effective
10
Why Space Weather Forecasting?

• Statement from the National Space Weather
  Program "Space weather" refers to conditions on
  the sun and in the solar wind, magnetosphere,
  ionosphere, and thermosphere that can influence
  the performance and reliability of space-borne
  and ground-based technological systems and can
  endanger human life or health. Adverse conditions
  in the space environment can cause disruption of
  satellite operations, communications, navigation,
  and electric power distribution grids, leading to
  a variety of socioeconomic losses.

Illustration from GEDAS Website
11
Effects of the large 1989 magnetic storm on the
US and Canadian power grid
12
Its Full of SatellitesTrust me
13
STS-39 Auroral Photograph
14
Beauty and Dynamics of the Aurora
From IMAGE FUV Imager
Oh Yes, the PC Index
15
Space Weather Gone Mainstream?
16
Can You say Lissajous Orbit
CAN EXPECT ONE HOUR WARNING

• Typical ACE Position
• X gse Distance 230 RE Upstream
• Y gse Excursions up to 40 RE
• Z gse Excursions up to 20 RE

Orbit Plots courtesy of ACE group at Caltech
17
ACE - an L1 Solar Wind Monitoring Satellite
18
What Did I Say? Part I Whats Next

• Remember only one hours notice. If your lucky!
  What can we do with one hour?
• The solar wind input time series must be
  corrected to account for variable propagation
  delay from L1 to the bow shock nose. Accomplish
  this using the ACE position vector, the solar
  wind velocity vector, and the normal vector
  defining phase front orientation.
• Dont forget the X/V propagation technique,
  simple as How long to drive to SD?.
• Can we do better? Weimer et al. 2002 2003
  gives some hope.
• The phase front normal directions can be
  determined using a non-standard minimum
  variance technique developed by Weimer et al.
  2003. He and his colleagues also presented a
  multi-spacecraft technique in 2002.
• Propagation delay correction using time-variable
  phase front normals can organize the solar wind
  magnetic field form well separated locations with
  a surprising high degree of accuracy.

19
Lets Force a Match Between Two IMF Time Series
From Weimer et al. 2002
20
Brute Forced Matching of Two IMF Time Series
From Weimer et al. 2002
21
Well Three Points in Space Always Define a Plane
The phase plane contains the IMF vector and the
position of ACE as well as the virtual positions
of IMP-8, Wind, and Geotail.
From Weimer et al. 2002
22
Surprising High Degree of Accuracy
From Weimer et al. 2002
23
Why MVA? What MVA?

• The advantage - only real-time data from one
  spacecraft is required!
• The measurements are radioed to the Earth and the
  data. A propagation is applied using the
  location where the in situ measurements were
  obtained, the solar wind velocity vector, and the
  orientations of phase fronts.
• A minimum variance technique
• developed by Weimer et al. 2003.
• This is in quotes because he got the
• desired results using a coding error!
• We and others found it.
• The error alters physical interpretation
• of the method but not the final results.

24
ACE - an L1 Solar Wind Monitoring Satellite
25
Minimum Variance AnalysisFor those Yearning for
Equations

• Calculate the mean magnetic field over a time
  interval with N samples
• Determine the variation of each vector sample
  from the mean field
• Calculate the mean cross product of deviations in
  the ith and jth components where i and j run
  from 1 to 3.
• Construct a (3 x 3) matrix from the elements Mij
• The diagonal components are the power in each
  component of the field

• Diagonalize the variance matrix obtaining the
  principal axis coordinate system for the
  variation in B during the interval
• Assume that the direction of minimum variance is
  the normal to a phase front in the solar wind

26
The Normal from Minimum Variance Analysis Follow
the Normals found with 4-Spacecraft
Weimer et al., 2002 2003
27
What the heck! Its New Years morning, 2
am.Im home alone these results wont match
28
We examined the eigenvector directions for
several forms of the IMF variance matrix
Thanks Professor Newman
29
Phase Plane Normal Determination

• Postulate that the solar wind consists of
  multiple adjacent flux tubes roughly aligned with
  the Parker spiral
• At any instant the direction of a flux tube is
  determined by the mean field
• As the monitor passes from one flux tube to
  another due to solar wind flow the field rotates
  slightly
• In a short interval this rotation is about an
  axis orthogonal to the mean field
• The axis of rotation can be found by minimum
  variance analysis of the magnetic field
  orthogonal to the mean field

Bruno et al. 2004
30
If You Can Visualize This

• GSE Coordinate Axes
• Red Points Black Line
• The IMF time history for an 28-min, 105-point
  time series.
• Red Line
• Corresponding IMF Mean Vector
• Blue Ellipsoid/Vector
• Variance ellipsoid and minimum variance
  direction from standard MVA
• Green Ellipsoid/Vector
• Variance ellipsoid and minimum variance
  direction for MVA without IMF mean subtraction

31
IMF and Phase Plane Normal Distributions
32
The Phase Plane Normals Are a Bit Surprising
Jan 1999
IMF
PPN

• The IMF distribution peaks near the ecliptic and
  along the Parker spiral direction.
• The PPNs must lie somewhere along the great
  circle that is oriented at 90 degrees with
  respect to the IMF mean vector.
• The PPNs are not confined to the ecliptic plane.
  For this month, the toward and away sector PPN
  distributions are significantly different.

Z
Y
Toward Sector
Away Sector
33
The Ugly Reality of Phase Plane Normals Note
that the denominator goes to zero when the phase
plane normal is perpendicular to the solar wind
velocity vector. This deficiency in the equation
has been account for by forcing the phase plane
normal to lie within the so-called limiting angle
taken equal to 45 in this study.
34
ACE - an L1 Solar Wind Monitoring Satellite
35
Propagation Delay Time Difference Distributions

• Black Curves - Convection propagation delay minus
  PPN propagation delay.
• Red Curves - Corotation propagation delay minus
  PPN propagation delay.
• Dashed Curves - Difference using PPN propagation
  delay with no limiting angle.
• Half width at half max is about 8 minutes when
  using a 45 degree limiting angle.
• This means that impulse response features should
  be better resolved, if the PPN model is accurate.

36
A Space Weather Forecasting Breakthrough?

• Plus or minus 8 minutes is not much!
• Guess you could turn off or remove a critical
  subsystem (unplug your computer).
• Probably more of research interest
• Lets consider the PC index

37
Magnetospheric Current Systems orWhy Doesnt It
Look Like a Dipole?
38
When the IMF is Southward!

• The magnetosphere responds initially via
  convection.
• Both ends of the field lines are moving. Just
  follow the numbers.
• Its hard to see so I will help!

39
PC - The Polar Cap Index
?H and ?D are the deviations of the components
measured in geographic coordinates relative to
quiet-time levels. ? is the geographical
longitude of the station. DE is the average
declination angle (DE 117º at Vostok, 285º at
Thule). ? is the angle between the noon-midnight
line and the direction of antisunward
convection. UT is universal time.
where Em VBTsin4?/2 is a solar wind
input parameter called the merging electric field
? Hproj ?Em ?
Normalization yields the PC index
PC ( ?Hproj ?)/ ?
Troshichev and Andrezen 1985 Vennerstrøm and
Friis-Christensen 1985
40
Propagation Model Effect on Solar
Wind Input-Magnetic Index Correlation
The Correlation coefficient as a function of time
lag for the VBs and PC index input-output series
from July 1999. The red curve is for VBs
calculated after applying the PPN propagation
method. The green, yellow and blue curves are
for VBs found after propagating using the
convection, half-way, and corotation methods,
respectively.
41
VBs-PC Index Response FilterSVD using 5, 10,
15, 20, 25 and all terms
Nov 1999
42
VBs-PC Index Response Filters - 1999
Impulse Response
Time Lag (hr)
43
What Did I Tell You?

• A little about the solar wind and the
  magnetospheres response.
• The new solar wind propagation model usually
  yield shifts of less than 10 minutes.
• Maybe not of too much space weather significance
  but may be good for resolving the time lagged
  response of the PC index. Astro does mind!
• May help tell when a solar wind is not very
  useful.

44

• Discussion Future Directions
• Using the PPN propagation delay technique to
  date leads only to a modest increase in the
  correlation between VBs and PC index time series
  despite the fact that the PPN technique can be
  used to greatly increase inter-spacecraft
  correlations of solar wind conditions as observed
  at ACE, WIND, IMP-8, GEOTAIL etc.
• We may have to face the possibility that solar
  wind conditions as measured at L1 may not impact
  the magnetosphere (perhaps 10 of the time?).
  This scenario may occur when the IMF is radial
  and PPN is perpendicular with the Earth-Sun line.
• The study will be expanded first to include all
  available solar wind input data from ACE and then
  to utilize other the interplanetary monitors. We
  also plan to properly handle shock propagation.
  New solar wind input variables will be tested.
• So much left to do with the PC index. Issues to
  address include

• Meeting the challenge - i.e. propose a unified
  definition
• Assess index saturation
• Determine the relative contributions of Hall
  and field-aligned currents to PC
• Assess the By-effect