Monico_LISN011_Final

Report
Ionosphere effects on GNSS positioning: data
collection, models and analyses
João Francisgo Galera Monico, Paulo De Oliveira Camargo, Haroldo Antonio
Marques, Heloisa Alves Da Silva
UNESP – FCT – Presidente Prudente, SP.
Bruno Bourgard
Septentrio NV, Leuven.
Luca Spogli
INGV, Rome.
Outline
• Infra-structure available for GNSS research and
applications in Brazil
• GNSS Services required in Brazil
• Brazilian Ionospheric Model
– Mod_ION
– Rinex_HO
• CIGALA Project
– Objectives
– Preliminary results
• Final Comments
Available Infra-structure
in South America/Brazil
SIRGAS GNSS data
• SIRGAS-CON GNSS
Network
Brazilian GNSS data (IBGE/INCRA)
• Brazilian
Continuous
GPS Network (RBMC).
Some
stations
are
operational since 1996
• ~100 stations
RBMC Real Time – RBMC_IP
• Data of about 30 Brazilian
GNSS
stations
are
distributed in real time,
using NTRIP protocol.
•
GNSS/GPS Active Network
at São Paulo State – Real time data
Meteorological and GNSS stations
• Meteorological stations are required to be collocated
with GNSS for GNSS/Met support
– 18 are available at São Paulo State (all stations were
calibrated)
GNSS demands in Brazil
Off shore applications
Air Navigation
Positioning in general
Precision agriculture
Rural Cadastre (50 cm or better – 1 sigma)
….
PA in Brazil is demanding 24 hours
RTK service
Concerning Air Navigation, Brazilian
authorities decided to invest in GBAS
instead of SBAS.
A system from Honeywell Aerospace is
under certification at Rio de Janeiro
Airport (Galeão). (Cosendey
presentation on Nov 09).
Challenges for such GNSS
applications
Ionospheric Scintillation!
São Paulo State Network RTK (VRS)
Preliminary results.
Local
Base/RTK
Initialization
TUPÃ
VRS (GNSS)
ARAC (GNSS) 84,13km
VRS_S (GPS)
ARAC_S (GPS) 84,13 km
1min 24 seg
8 min 4 seg
2 min 23 seg
12 min 19 seg
Sart
13:07:01 as
13:24:52 as
13:47:55 as
14:18:35 as
End
N. points
collected
13:18:17
13:43:41
14:08:33
14:44:30
205
205
205
205
Ionospheric Index (I95)
based on São Paulo
State GNSS Network
Developments on GNSS/Ionosphere at
FCT/UNESP
GNSS and Ionosphere
• A Ion-model based on GNSS has been under
development at FCT/UNESP since 1997;
– Mod_Ion (in-house iono model) generates Ionospheric
maps and coefficients for L1 users
• Ionospheric Index (Fp)
• Ionex files from Brazilian GNSS data
• Real time ionosphere maps of TEC/ROT and of the
correspondent delays on L1 (Aguiar – presentation
on Nov 9th).
Ionospheric Regional Model (MOD_Ion) (GPS & GLONASS)
•
•
Fi
TEC
(P
s
2r
 P )
s
1r
VTEC

s
cos( z  ) r
TEC
+ Fi
s
s
[(S p2 -S p1 ) + (R p2 -R p1 ) r ] + Fi
TEC
 P 21
= > i = G, Rk
Fi  f 2 i /(f 1i  f 2 i )
2
2
2
n4
VTEC  a1  a 2B 
s

m 4
{a j cos(i h )  a j  1 sin(i h )}  a n*2  3 h 
s
s
i 1
j  2i  1
s

{a jcos(i B )  a j 1 sin(i B )}
s
s
i 1
j  2i  10
n4
F(P  P ) r  S F (a 1  a 2 B 
s
2
s
1
s
r
s

{a j cos(i h )  a j 1 sin(i h )}  a n*2  3 h 
s
s
s
i 1
j  2i  1
m4

i 1
j  2i  10
{a j cos(i B )  a j 1 sin(i B )}  F(R P 2  R P1 ) r
s
s

F(S P2  S P1 )  F  P 21 .
s
s
Mod_Ion with inequality equation
• Problem: at some situations, even with calibrated
equipments, negative values of TEC are obtained.
• One solution: to apply inequality equation as follows:
n4
VTEC  a1  a 2B 
s

{a j cos(i h )  a j  1 sin(i h )}
s
s
i 1
j  2i  1
m4
 a n*2  3 h 
s

i 1
j  2i  10
{a j cos(i B )  a j  1 sin(i B )}  0
s
s
GNSS Ionospheric Products
• TEC Maps
IONEX Files
2nd and 3rd order Ionosphere corrections
• In-house software was developed (RINEX_HO)
• GPS Solutions, Online First: 21 April 2011, DOI:
10.1007/s10291-011-0220-1, "RINEX_HO:
second- and third-order ionospheric corrections
for RINEX observation files" by H. A. Marques, J.
F. G. Monico and M. Aquino
2nd and 3rd order Ionosphere corrections
• The earth’s magnetic field
– Dipolar Approximation
– International Geomagnetic Reference Field (IGRF) model
(IGRF11 model)
– Corrected Geomagnetic Model from PIM (Parameterized
Ionospheric Model)
• TEC
– From raw pseudoranges, from pseudoranges smoothed by
phase, or from Global Ionosphere Maps (GIM).
2nd order Ionosphere corrections
Bipolar – IGRF and Differences
CIGALA Project
“Concept for Ionospheric scintillation mitiGAtion for professional GNSS in
Latin America”
Goal: Understand the cause and implication of IS disturbances at low
latitudes, model their effects and develop mitigations through:
– Research of the underlying causes of IS and the development of state-of-the-art models
capable of predicting signal propagation and tracking perturbations
– Field measurement via the deployment in close collaboration with local academic and
industrial partners of multi-frequency multi-constellation Ionospheric Scintillation
Monitoring (ISM) network
– Design and implementation of novel IS mitigation techniques in state-of-the-art GNSS
receivers
– Field testing the mitigation techniques, leveraging the same partnership as during the
measurement campaign.
CIGALA partners
IS Monitoring Network in Brazil
• 8 ISM stations
• Latitudinal and longitudinal
distribution over Brazil
• Two stations at São José dos
Campos (crest of EIA) and
Pres. Prudente
• Data stored locally and sent to
repository at UNESP, Pres.
Prudente
• Data mirrored at INGV, Rome
CIGALA IS Monitoring Network in
Brazil
Continuous recording of :
• Amplitude scintillation index S4 : standard deviation of received power
normalized by its mean value
• Phase scintillation index σΦ : standard deviation of
de-trended carrier phase, with Phi60 its 60” version
• TEC (Total Electron Content)
• Lock time
• Code – Carrier Divergence
• Spectral parameters of phase Power Spectral Density:
– Spectral slope p
– Spectral strength T
• Raw high-rate I&Q correlation values (50Hz)
Septentrio PolaRxS ISM receiver is the
base of the CIGALA network
(c) CIGALA Consortium
PolaRxS: facts
 Track GPS, GLONASS, GALILEO, COMPASS, SBAS
 L1, L2, L5, E5a, E5b signals, including GPS L2C, GLONASS








L2C and Galileo E5 AltBOC
Very low phase noise OCXO
100Hz signal intensity and phase output for all signals
Computation of S4, sf , TEC, spectral parameters,... for all
satellites and signals
Interoperable ISMR file format
Multiple Interfaces: 4 RS232, USB, Ethernet
Rugged IP65 housing
Temperature range: -40C to 60C
Powering: 9-30V ; 6W
PolaRxS Phi60 Noise Floor <0.03rad
24-h Spirent simulation, Perfect GPS signal, L1
Receiver optimize for Maximum Tracking
availability during Strong Scintillation
Normal Receiver
Optimized ISM receiver
60
10
8
Loss-of-lock probability [%]
Loss-of-lock probability [%]
50
40
30
20
10
0
20
6
4
2
0
20
1
15
10
1
15
0.8
0.8
10
0.6
5
PLL bandwidth [Hz]
0.6
5
0.4
S4 level
PLL bandwidth [Hz]
Data bearing signals
Simulated with CSM on Spirent
0.4
S4 level
Receiver optimize for Maximum Tracking
availability during Strong Scintillation
Normal Receiver
Optimized ISM receiver
60
10
8
Loss-of-lock probability [%]
Loss-of-lock probability [%]
50
40
30
20
10
0
20
6
4
2
0
20
1
15
0.8
10
1
15
0.8
10
0.6
5
PLL bandwidth [Hz]
0.6
5
0.4
S4 level
PLL bandwidth [Hz]
0.4
S4 level
Pilot Signal (L2C)
Simulated with CSM on Spirent
Comparison with currently deployed
GSV equipment
• Scintillation free midlatitude location
(Nottingham)
• GPS L1CA
• 24h recording
• S4: correlation
coefficient = 0.9
• Phi60:
– PxS: 0.0292
– GSV: 0.0547
PRN19
Field Validation (C/N)
L1
• CIGALA receivers PRU1 and PRU2 at Presidente Prudente
• February to April 2011
L2
Field Validation (CCSTDDEV)
L1
• CIGALA receivers PRU1 and PRU2 at Presidente Prudente
• February to April 2011
L2
Using GLONASS for IS monitoring
• GPS and GLONASS orbits are complementary to increase spatial and temporal
observability of the ionosphere
• GLONASS provides open signals on both L1 and L2 in all SV
Moderate Scintillation Occurrence (S4)
observed using GPS vs. GLONASS
GPS
EIA
GLONASS
EIA
• INGV GBSC software is used to draw maps of rate of occurrence of S4>0.25 as a function of
lat/long or lat/time
• Maps plotted for L1 observations between Feb and April 2011
• Increased probability of scintillation clearly observable in EIA post-sunset
• Very good match between GPS and GLONASS observation => data can be merged
Moderate Scintillation Occurrence (Phi60)
observed using GPS vs. GLONASS
GPS
GLONASS
• INGV GBSC software is used to draw maps of rate of occurrence of Phi60>0.25 as a function
of lat/long or lat/time
• Maps plotted for L1 observations between Feb and April 2011
• EIA observable for GPS
• No match GPS and GLONASS observations
Understanding lack of Phi60 observability
when using GLONASS signal
• Short term stability of the
GLONASS satellite clock lower
than GPS
• Small scale phase scintillation
cannot be measured from single
frequency observation
• Solution:
Using differenced L1/L2
measurement to cancel the
satellite clock effect
Strong Scintillation Event on
Sept 25, 2011
S4 During Scintillation
L2C
L1CA
PRU2, Sep-25, 2011
PRU2, Sep-25, 2011
1.4
1.4
PRN2
PRN15
PRN15
1.2
PRN26
S4 from L2C (elevation mask of 20 deg)
S4 from L1CA (elevation mask of 20 deg)
1.2
1
0.8
0.6
0.4
0.2
0.8
0.6
0.4
0.2
0
0
0
•
•
1
0.5
1
UTC time [hours]
1.5
2
0
0.5
1
UTC time [hours]
S4 reported continuously during scintillation
S4 in L2 reported thanks to PRN15 (L2C) pass
1.5
2
SigmaPhi during Scintillation
L2C
L1CA
PRU2, Sep-25, 2011
PRU2, Sep-25, 2011
1.5
1.5
PRN15
PRN2
PRN15
Phi60-L2C [rad] (elevation mask of 20 deg)
Phi60-L1CA [rad] (elevation mask of 20 deg)
PRN26
1
0.5
0.5
0
0
0
0.5
1
UTC time [hours]
•
1
1.5
2
0
0.5
1
UTC time [hours]
sphi reported continuously on ISM optimized receiver
1.5
2
Tracking robustness (Cycle Slips)


Phase tracking continuous during the whole event despites the very high S4
level
3 cycles slips seen on L1CA (PRN15)
No cycles slips on L2C!
PRU2, PRN15, Sep-25, 2011
carrier phase
0.5
nav bit error
0.4
0.3
Detrended L1 carrier phase [cycles]

0.2
0.1
0
-0.1
-0.2
-0.3
-0.4
-0.5
1500
2000
2500
UTC time [hours]
3000
Effect on Real Time Precise Point
Positioning
PRU2, Sep-25, 2011
1.4
PRN2
PRN15
1.2
S4 (elevation mask of 20 deg)
PRN26
1
0.8
0.6
0.4
0.2
0
0
0.5
1
1.5
2
2.5
3
UTC time [hours]
434
433.8
PPP Height [m]
433.6
433.4
433.2
433
432.8
432.6
0
0.5
1
1.5
UTC time [hours]
2
2.5
3
PPP service continuous during
the whole event
Up to 40cm error during event
(service specification is 12cm
95%)
Final comments
• Brazil is a very challenge place for GNSS
applications, mainly due to the Ionosphere behavior
in the equatorial region;
• Several applications are already suffering the
effects of such problem (IS) and will increase in the
next two years;
• In the PA and aviation there is a need for more
developments and tests;
• CIGALA network will continue collecting data
after the final of the project (March 2012) and may
provide data for scientific purpose.
More information?
http://gege.fct.unesp.br

similar documents