The Ionosphere and its Effect on Satellite Navigation

1
The Ionosphere and its Effect on Satellite
Navigation

• Todd Walter
• Stanford University
• So. Cal. ION 11 September 2008
• http//waas.stanford.edu

2
What is the Ionosphere?

• The ionosphere refers to regions of the upper
  atmosphere where charged particles occur in
  sufficient densities to affect radio waves.

3
Ionosphere Influences

• Changes the Velocity of the Signals
• Group Delay
• Refraction of Radio Waves
• Amplitude and Phase Variations
• Errors Are a Function of
• Carrier Frequency
• Solar Activity
• Magnetic Latitude
• Seasonal and Time of Day/Night Variations
• Elevation and Azimuth of the Satellite

4
Ionospheric Effects
5
Sun and Magnetosphere
Coronal Mass Ejections
Aurora
Ionosphere
Sunspots
Image from AGU
6
Generation of the Ionosphere
7
Atmospheric Heights
8
Diurnal, Polar andEquatorial Variations

• Longitude Profile Viewed
• From Above North Pole

• Latitude Profile Viewed
• From Equatorial Plane

Note Distance of Shaded Region From Earth
Indicates Ionospheric Density (TEC) and Not
Height.
9
Ionospheric Delay
10
Seasonal Variations
Courtesy Pat Doherty Jack Klobuchar
11
11-Year Solar Cycles
12
Appleton Hartree Equation
13
Phase Advance
? X
0.163 m at L1
14
Code Delay
? X
0.163 m at L1
15
Ionosphere Observables
16
Thin-Shell Model
17
Obliquity Factor
18
Nominal Ionosphere - IPPs
19
Failure of Thin Shell Model
Quiet Day
Disturbed Day
20
Disturbed Ionosphere
21
Ionospheric Threat
22
20 November 2003 2015 2100 UT
Slide Courtesy Seebany Datta-Barua
23
Estimation of Ionospheric Gradients
T1
T2
IPP
S2
S1
S1
S2
S1
Slide Courtesy Jiyun Li
24
Nominal Day Spatial Gradients Between WAAS
Stations
Typical Solar Max Value Below 5 mm/km
Slide Courtesy Seebany Datta-Barua
25
Spatial Gradients Between WAAS Stations During
Anomaly

• Storm Values
• gt 40 mm/km
• up to 360 mm/km

Slide Courtesy Seebany Datta-Barua
26
Ionospheric Decorrelation(0th Order)
Typical Solar Max Value Below 2 mm/km
27
Equatorial Ionospheric Decorrelation
Typical Solar Max Value Below 11 mm/km
28
Disturbed Ionosphere Decorrelation
29
Solar Max Quiet Day
July 2nd, 2000
30
Temporal Gradients
Slide Courtesy Seebany Datta-Barua
31
Disturbance in Polar Region
4 m change
200 sec
32
GPS Modernization

• Does access to two civil frequencies solve all of
  our ionospheric problems?
• Increase user noise multipath
• 2.59 times larger than L1-only
• Scintillation
• Higher order effects

33
What is Scintillation?

• Local Disturbances in the Ionosphere that Alter
  Signal Amplitude and Phase of Received GPS Signal
• Refraction/Diffraction - Small Scale
  Irregularities
• Amplitude Fades
• Phase Variations
• Most Common in Equatorial and Polar Regions

34
Equatorial Scintillation
35
Global Distribution of Scintillation
36
Scintillation and Deep Signal Fading

• Signal to noise ratio (C/No) of PRN 11 (Mar. 18,
  2001)

Nominal
C/No
Scintillation (equatorial solar max)
25 dB
C/No
Slide Courtesy Jiwon Seo
100 sec
37
Scintillation and Navigation
GPS
WAAS
Scintillation Patches
Slide Courtesy Jiwon Seo
38
Scintillation and Navigation
GPS
WAAS
Scintillation Patches
Slide Courtesy Jiwon Seo
39
Severe Scintillation Data(sky view)
Slide Courtesy Jiwon Seo
40
Severe Scintillation Data

• 7 out of 8 satellites were affected by
  scintillation in our data set
• Worst 45 minutes based on S4 index during the 8
  days campaign at Ascension Island on Mar. 2001
  
• Collected using a NAVSYS DSR-100 receiver with a
  Rubidium frequency standard
• Dr. Theodore Beach, AFRL (S4 plots Raw IF
  sampled data)
• Processed in 50 Hz rate using a NordNav
  commercial software receiver
• 50 Hz C/No (signal to noise ratio) output was
  analyzed in this research

Slide Courtesy Jiwon Seo
41
Severe Scintillation Data (example)

• 50 Hz C/No outputs of all 8 satellites on sky
• (100 sec out of 45 min data as an example)
• Number of simultaneous loss of satellites is more
  important than number of fading channels

Slide Courtesy Jiwon Seo
100 sec
42
Simultaneous Loss of Satellites

• Chance of simultaneous loss is strongly dependent
  on reacquisition time of receiver

Slide Courtesy Jiwon Seo
20 sec Loss
18 sec
Max of 4 SV Loss
43
Simultaneous Loss of Satellites

• Chance of simultaneous loss is strongly dependent
  on reacquisition time of receiver

Slide Courtesy Jiwon Seo
2 sec Loss
18 sec
Max of 2 SV Loss
44
Number of Tracked Satellites

• Simulating 20 sec reacquisition time (WAAS MOPS
  limit)
• Using 45 minutes of severe scintillation data
• 4 or more 97.9 , 5 or more 92.3 , 6 or
  more 68.1

100
4 or more tracked SVs
5 or more
Time Percentage
Slide Courtesy Jiwon Seo
6 or more
65
20 sec
2 sec
Reacquisition Time
45
Number of Tracked Satellites

• Simulating 2 sec reacquisition time
• 4 or more 100 , 5 or more 100 , 6 or
  more 98.3
• WAAS MOPS limit (20 sec) should be reduced

100
4 or more tracked SVs
5 or more
Time Percentage
Slide Courtesy Jiwon Seo
6 or more
65
20 sec
2 sec
Reacquisition Time
46
Higher Order Model of n

• To approximate n with higher accuracy, expand the
  Appleton-Hartree Equation in powers of signal
  frequency.
• Brunner Gu (1991), Tucker Fannin (1968),
  Bassiri Hajj (1984)
• Group index of refraction

Slide Courtesy Seebany Datta-Barua
47
1st Order (X)
120 m
Quiet
Active
Slide Courtesy Seebany Datta-Barua
48
Higher Order Phase Errors
20 cm
Quiet
Active
Slide Courtesy Seebany Datta-Barua
49
Higher Order Range Errors
20 cm
Quiet
Active
Slide Courtesy Seebany Datta-Barua
50
Summary

• The ionosphere has a major impact on GPS L1-only
  operations
• Accuracy, Integrity, Availability, Continuity
• The arrival of a second civil frequency will
  address the largest issue
• Significant effects remain, particularly in
  equatorial regions
• Accuracy Continuity still affected

51
Small-scale Irregularity
52
Artificial Undersampled Scenario
53
WAAS Measurements
54
Artificial WAAS Undersampling Scenario
55
Real Undersampled Condition
56
WAAS Measurements
57
Single-Freq. Klobuchar Iono. Delay Correction
Algorithm

• a and b parameters are pre-stored functions of
  date not based on current observations
• Application of model reduces overall
  single-frequency user ionosphere delay errors by
  approximately 50

58
Scintillation Location
59
Higher Order Terms
60
Group Velocity
61
Ionosphere Spatial Anomaly Discovered in 4/6/00
Storm
Observed gradient 6 m vertical delay at 7 km IPP
separation (86 cm/km) 19 km effective
separation, since IPP moved in opposite direction
from anomaly
62
Aurora Borealis
63
Physical Mechanisms
E-Field Drift

• Solar Flux
• Electric Fields
• Magnetic Fields
• Diffusion
• Winds

64
Ionospheric Chemistry

• Above the Mesopause, Solar UV Disassociates
  Molecular Oxygen into Atomic Oxygen
• Solar UV and Soft X-Rays Ionize the Neutral
  Constituents (Primarily Atomic Oxygen and
  Hydrogen)
• Electrons Diffuse to Greater Heights Following
  Magnetic Lines of Force
• Loss Primarily Through Molecular Recombination

O O2 O2 O g O2 e- O O g
65
Chapman Function
66
Observation Equations
Small L1/L2 differences (residual
inter-frequency biases)
L1 and L2 errors are not identical
67
Ranging Impacts and Errors
68
Useful Combinations

• Ionosphere-Free Combinations
• Ionospheric Estimates
• L1

g
-
PR
(
L
)
PR
(
L
)
1
2
g
-
1
code-minus-carrier (divergence)
-
PR
(
L
)
PR
(
L
)
2
1
g
-
1
Noisy Unbiased Estimates
Biased Quiet Estimates
69
Dual Frequency Smoothing
70
WAAS Supertruth Data

• Raw Data Collected From Each WRS
• 3 independent receivers per WRS
• Postprocessed to Create Supertruth
• Carrier tracks leveled to reduce multipath
• Interfrequency biases estimated and removed for
  satellites and receivers
• Comparisons made between co-located receivers
  (voting to remove artifacts)
• Multipath and Bias Residuals up to 50 cm
• Without Voting, Receiver Artifacts Cloud Results
  and Make It Impossible to See Tails of the
  Distribution

71
Solar Events
72
Magnetosphere

• The region of space to which the earths
  magnetic field is confined by the solar wind.

Slide Courtesy Seebany Datta-Barua
Courtesy of Windows to the Universe
http//www.windows.ucar.edu/
73
Correlation Estimation Process
74
Disturbed Ionosphere
75
Southward Progress of Anomaly Wave Front over
1.5 Hours
76
Active Ionosphere
Slide Courtesy Seebany Datta-Barua
77
Decorrelation Between IPPs During Ionospheric
Disturbance
Slide Courtesy Seebany Datta-Barua
78
Map of 2 Minutes Aggregate WAAS Supertruth Data
UTC 06/04/2000 213212 - 213402
20
44
18
43
Boston
16
42
14
41
New York City
12
40
Vertical Ionospheric Delay in m
Washington, D.C.
10
39
ipp direction of motion
8
38
6
37
Storm front direction of motion
4
36
Slide Courtesy Seebany Datta-Barua
2
35
0
34
-82
-80
-78
-76
-74
-72
-70
Only ipps to svn 40 at 213212 and 213402 shown
79
Ionospheric Decorrelation Function (0th Order)
80
Preliminary Decorrelation Findings

• Nominal Ionosphere is Relatively Smooth
• Nearby IPPs Well Correlated
• Confidence About a Single Measurement Can Be
  Described As
• s2 sm2 (0.5 m d0.5 m/1000km)2
• There Appears to Be a Deterministic Component
• Next Try Removing a Planar Fit

81
Ionospheric Decorrelation About a Planar Fit (1st
Order)
82
Ionospheric Decorrelation Function (1st Order)
83
Equatorial Sigma Estimate1st Order