Transcript A Norwegian Ionosphere Model based on GPS Data

A Norwegian
Ionosphere Model
Based on GPS Data
Anna B.O. Jensen
Nordic Institute of Navigation
Oslo, June 2008
Outline
 Navigation in the arctic
 Ionospheric activity at high latitudes
 SATREFTM Ionosphere Model
 Verification towards IGS GIM
 Comparison with EGNOS ionosphere
model
 Summary
Navigation in the arctic (1)
 Norway is located at high
latitudes, from 60° to 80° N,
mainly in the arctic region
 GPS is used extensively as a navigational mean
in the arctic
 Unfortunately, GNSS performance is limited in
the arctic compared to mid latitudes
Navigation in the arctic (2)
 Much offshore activity
in the Norwegian Sea
 Need for reliable
navigation
 In the future:
 Increasing activity and
more traffic due to global
warming and more oil
and gas exploration
 Increasing need
for navigation
Ionospheric activity at high latitudes
 At high latitudes characteristics of the
ionospheric activity are different than at
mid latitudes:
Higher ionospheric variability
Increased amount of scintillation
 This does affect navigation users e.g. by:
Reduced accuracy
Poor signal tracking (loss of lock)
Ionospheric activity at high latitudes
Feb. 28, 2008
70°N network
60°N network
Trimble gpsnet
software
TM
SATREF
Ionosphere Model (1)
 Several ionosphere models exist, but
they are generally poor for high latitudes
 In 2007 the NMA therefore started
development of a regional Norwegian
ionosphere model based on the
SATREFTM network of GNSS stations
TM
SATREF
Ionosphere Model (2)
 The model is based on:
GPS data from selected
SATREFTM stations
Estimation of ionospheric
delays in the stations
Spatial interpolation to
obtain nationwide grid
model
Test area
Test data
Day of year
Date
K-index
DOY 015
Jan. 15, 2008
2–4
Iono.
activity
Low
DOY 033
Feb. 02, 2008
2–6
Medium
DOY 059
Feb. 28, 2008
2–6
Medium
DOY 324
Nov. 20, 2007
1-5
Medium
Verification towards GIM (1)
 Verification with respect to the Global
Ionosphere Model (GIM) of the IGS
 IONEX files retrieved from the IGS web
site, and L1 ionosphere delays extracted
for comparison with SATREFTM
Ionosphere Model
 20 grid points used for verification
Verification towards GIM (2)
 Differences, SATREFTM minus IGS GIM
Date
Jan. 15, 2008
Mean
[ meter ]
-0.02
Std. dev.
[ meter ]
0.12
Feb. 02, 2008
-0.03
0.12
Feb. 28, 2008
-0.04
0.18
Nov. 20, 2007
-0.05
0.19
 30 second sampling, 20 grid points
Verification towards GIM (3)
 Summing up:
Mean of differences of 2 - 5 cm is
basically negligible
• Indicates no offset between the two models
Standard deviation of 12 – 19 cm
• Occur mainly because no filtering is applied
to the SATREFTM model
• Lower standard deviation on the day with low
ionospheric activity
Comparison with the EGNOS iono. model
 Verification of the SATREFTM model
towards the IGS GIM showed acceptable
results
 Therefore, the SATREFTM model is now
used for a preliminary evaluation of the
performance of the EGNOS ionosphere
model in the arctic
Comparison with EGNOS iono. model
 Differences, SATREFTM minus EGNOS
Date
Std. dev.
[ meter ]
0.15
Samples
Jan. 15, 2008
Mean
[ meter ]
-0.12
Feb. 02, 2008
-0.10
0.20
6230
Feb. 28, 2008
-0.15
0.20
6435
Nov. 20, 2007
-0.12
0.23
6189
 16 grid points
6868
Selected grid point – Feb. 28, 2008
 EGNOS: blue, GIM: green, SATREFTM: red
 EGNOS model is biased
Test area
Selected grid points – Feb. 28, 2008
 EGNOS bias for upper grid point
Selected grid points – Feb. 2, 2008
 Another day - again EGNOS bias for same point
16 grid points, Jan. 15, 2008
Future work
 Modify model to run in real time
Lots of programming
 Further investigations to decide on:
Coverage area
Grid spacing
Number of SATREFTM stations to include
Temporal update interval
Information to users – web application
Summary
 Development of the SATREFTM Ionosphere
Model has been initiated
 Verification of the SATREFTM Ionosphere Model
towards the IGS GIM show very good results
 Comparison with EGNOS model show
deviations for some grid points
 Improvement expected with new EGNOS version this
summer
Acknowledgments
 Thanks to the Norwegian Space Centre
for providing support for the work
 Thanks to Ola Øvstedal, Norwegian
University of Life Sciences in Ås, for
valuable discussions during the
development phase