Ionospheric Tomography and Numerical Mapping of F2layer Critical Frequency Using GPS RO Data By L'C'

1
Ionospheric Tomography and Numerical Mapping of
F2-layer Critical Frequency Using GPS RO DataBy
L.-C. Tsai, W. H. Tsai, C.H. Liu

• Acknowledgements
• Bill Schreiner, Doug Hunt
• July 2, 2004

2
Geographic Distribution of Ionosondes and GPS/MET
Occultation Observations
3
The International Reference Ionosphere (IRI)
Structure (Rawer, 1981)
4
IRI-modeled foF2 Spherical Maps (under the
Inertial Coordinate)

• Spherical maps at the vernal equinox, summer
  solstice, autumnal equinox, and winter solstice
  in 1996

5
GPS/MET vs. IRI foF2 (the URSI map) Comparison

• The correlative scatter plot (the left panel)
  between the IRI-modeled foF2 and GPS/MET
  retrieved foF2 (49325 observations), and their
  difference distribution (the right panel)

6
The Position Projections at Ne Peaks (under the
Inertial Coordinate)

• 4776 observations recorded within the 13th prime
  time of GPS/MET

7
Retrieved foF2 Values vs. the Inertial Latitude
8
Numerical Fitting of foF2 by the Least Squares
Method

• 1. Gk(?,?), main latitudinal variation
• 2. Gk(?,?), mixed latitudinal and longitudinal
  variation
• - First order in longitude
• - Second order in longitude

9
Representation of foF2 Main Latitude Variation

• Daytime (the left panel) and nighttime (the right
  panel) foF2 fitting by least squares polynomials
  of degree 10 in latitude

10
The Optimum Separation of Noise

• Standard deviation of residuals ek vs. degree k
  for main latitude (in red) and mixed latitudinal
  and longitudinal (in blue and black) variation

11
The foF2 Numerical Map

• Main latitude trend mapping using the 13th prime
  data of GPS/MET

12
The foF2 Numerical Map

• Mixed latitude and longitude trend mapping (at
  the 1st order) using the 13th prime data of
  GPS/MET

13
The foF2 Numerical Map

• Mixed latitude and longitude trend mapping (at
  the 2nd order) using the 13th prime data of
  GPS/MET

14
The Comparison of the IRI-modelled and GPS/MET
foF2 numerical maps
15
Estimate of standard deviation error of foF2
representation from GPS/MET
16
The geometry of the GPS-LEO occultation problem
for ionosphere observations
P1 is an occulting LEO point, and P2 is an
auxiliary LEO point with the same radial distance
of P1.
17
An example of link sideview of close-up/grouped
occultation observations for tomography
reconstruction

• The cell resolutions
• Latitude 0.5º
• Altitude 10 km
• (Longitude 30º)

18
Basic equations for ionospheric tomography
inversion
19
Three contiguous tomographic images (Tsai et.
al, JASTP,2002)

• Reconstructed from the actual GPS/MET data
  recorded from on Feb. 23, 1997.

20
Question How to evaluate the determined
tomography?

• To find optimal reconstruction strategy using the
  MART algorithm in space-based ionospheric
  tomography
• 2D tomography from a reference GPS/MET experiment
  through the IRI model
• 3D tomography from a reference COSMIC/ROCSAT-3
  experiment through the IRI model
• Compare the determined 2D/3D tomography to the
  true IRI-modeling values

21
A reference GPS/MET experiment

• Two contiguous true ionospheric images obtained
  from the IRI model

22
A reference GPS/MET experiment

• The log images for the occultation observations
  within each of the two revolutions of the
  Micro-Lab1 satellite
• The reference TECs are simulated by the line
  integrations of the IRI modelled electron
  densities along the calibrated paths

23
A reference GPS/MET experiment

• Two electron density images determined,
  separately, by the MART algorithm using the
  occultation data within each of the two
  revolutions

24
A reference GPS/MET experiment (combining RO data
within two resolutions)

• The log image for combined occultation
  observations within the two revolutions of the
  Micro-Lab1 satellite

25
A reference GPS/MET experiment (combining RO data
within two resolutions)

• The electron density image by the MART algorithm
  using all occultation data within the two
  revolutions

26
A reference GPS/MET experiment

• RMS electron density error images for
  reconstructed tomographys using occultation
  observations within one (left), and two (right)
  LEO orbital revolutions

27
A reference COSMIC experiment STK (Satellite
Tool Kit) simulated COSMIC orbits2003.7.2
0000 0159
28
A reference COSMIC experiment

• LEO-LEO link projections (199 occultations) onto
  the earth surface
• Different color presents links to different LEO
  satellite

z
x
y
29
An example of 3D tomography (2003.7.1.
18001940 UT)

• h 200 600 km , ?h 20 km
• The cell resolution 1º latitude, 5º longitude,
  10 km altitude

30
Validation of reconstructed 3D images (I)

• The left panels show the reconstructed images,
  and the right panels show the true images.

h200 km
h200 km
h300 km
h300 km
31
Validation of reconstructed 3D images (II)

• The left panels show the reconstructed images,
  and the right panels show the true images.

h400 km
h400 km
h500 km
h500 km
32
Summary

• The foF2 values retrieved from RO data can be
  used to represent ionospheric characteristics
  with an error of lt1Mz standard deviation.
• With a cutoff t2 in the Students distribution,
  the total term number of the mapping polynomials
  is 90 which is one order less than the original
  IRI numerical mapping.
• The average error of foF2 representation around
  the mid of years is better than these around the
  beginning (or ending) of years.
• The tomographic reconstructions of using combined
  reference GPS/MET data within one and two passes
  of the LEO orbit show convincingly improved
  agreement to the original IRI-modelled images.
• The COSMIC mission has 6 space-based GPS
  receivers and could provide twice of occultation
  observations to determine 3D tomographic electron
  densities.