AbstractFree electrons existing in the terrestrial ionosphere severely affect the transmission ofradio waves coming from satellites orbiting around the Earth. That effect, unwanted inmost cases, is the main source of information for ionospheric research. In this contributionwe present a method for combining GPS and TOPEX/Jason observation in a consistentmodel approximation, which exploits the complementary benefits of both satellitetechniques. Basically, TOPEX/Jason observations are used to improve the vertical totalelectron content interpolation (vTEC) in open ocean regions and to improve the calibrationof the GPS satellite and receiver instrumental biases (DCB), while GPS observations areused to achieve high temporal resolution of vTEC variability.

IntroductionThe ionospheric information provided by dual-frequency signals traversing the ionosphereis related to the total electron content TEC, defined as the electron density Ne integralalong the signal ray path:

GPS provides slant TECs, sTEC, between the broadcasting satellites and the groundreceivers while TOPEX/Jason provides vertical TECs, vTEC, between the satellite and the seasurface. Slant sTEC and vTEC are both related through a mapping function F. The unit formeasuring TEC is the TECU (TEC Unit) and 1TECU=1016 electrons/m2

La Plata Ionospheric Model LPIMGPS

LPIMGPS, Brunini created in 1998, approaches the hole ionosphere by a spherical shell ofinfinitesimal thickness located 450km above the Earth´s surface. Mapping function used is:

where z’ is the satellite zenith distance at the piercing point .The two-dimensional global vTEC distribution on the shell is described by a sphericalharmonic expansion up to 15 degree and order:

The coordinates used for the expansion are the modip latitude , introduced by Azpilicuetain 2005, and local time (LT). Input data come from GPS measurements.

Our purpose: an upgraded version of LPIM, LPIMGPS+TP

This work focuses on the updating of LPIM in order to add the new experimental datasource that provides satellite altimetry over ocean regions, just where GPS observationsare very scarce. The new version, LPIMGPS+TP , will be able to assimilate long GPS andTOPEX data series, and compute high temporal/spatial resolution maps of the global TEC.

GPS observable and dataGPS ionospheric observable is geometry free linearcombination carrier-to-code leveled , which isrelated with sTEC and the hardware delay (DCB)of the receiver and the satellite bi and bk through:

RINEX data (ftp//garner.ucsd.edu/pub/) for 300Continuos tracking stations were processed.The stations were selected from International GNSS Service (IGS) global network(http://igscb.jpl.nasa.gov) so that there are not more than one in each 5°x5° square. Asshown in figure 1, they are highly concentrated over the continental region of thenorthern hemisphere.

TOPEX/Jason observable and dataTOPEX measurement over the Earth’s ocean regions provides a direct measurement ofvTEC through the following equation:

where RK is the range measured by the altimeter at a frequency fK=13.575 GHz and RC isthe range measured at a frequency fC=5.3 GHz.

TOPEX satellite orbits at an altitude of 1336 km collectinginformation over oceanic earth surface between ±66° oflatitude. It pass over the same point of the ocean surfaceeach 9.9 solar days two hours (local time) early, performing127 revolutions with a ground track equatorial spacing of316 km (figure 2). Therefore, around 96 days are neededfor 24 hours local time coverage. Topex data weredownloaded from NASA-JPL Physical Oceanography

Distributed Active Archive Center (ftp://podaac.jpl.nasa.gov/pub/sea_surface_height/topex_poseidon/topex_gdr/).

ResultsGPS and TOPEX series of observations were processed for a period of 100 days. The periodgoes from 1st January 2002 to 10th April 2002 and corresponds to a high solar radiationactivity.The coefficients alm and blm of the expansion series have been estimated by least squaresadjustments every 2 hours UT interval and a daily calibration offset for each receiver andsatellite, bi and bk, has been obtained. This procedure was first applied using only GPSmeasurements and then repeated using both GPS and TOPEX ones.

a) Global ionospheric maps (GIMs) comparisons

GIMs produced by LPIMGPS and LPIMGPS+TP observations were assessed. Figure 3 shows the vTEC global distribution for the period [14:00-16:00] hs of Universal Time (UT) for day 008-2002 obtained with LPIMGPS+TP. In order to study the influence of TOPEX data in a global scale, we have computed, witheach model, a mean GIM for each period of 2 hours using 100 days. The differencebetween both models for the period [14-16] hs of TU is mapped in figure 4. Thedifference is minimal in the modip northern hemisphere continental region and it reachesa mean value of 2.88 TECU in the modip southern hemisphere over the ocean.

b) Comparisons between LPIM and TOPEX measurements.For every TOPEX determination, vTECTP, two vTEC values were computed, one usingLPIMGPS, vTECLPIM_GPS, and the other using LPIMGPS+TP, vTECLPIM_GPS+TP . Both series areplotted in figure 5 , together with TOPEX measurements,against TU.

Differences were binned into bands of 1 hour width and the mean values of differencesand its standard deviations were plotted against local time in figure 6. The procedure wasrepeated but in this case the results were binned into modip latitude bands 3°width(fig. 7).

We have plotted in blue, at the top, differences between old and new version of LPIM. InLT a constant bias is observed, but as a function of modip a greater deviation in the modipsouthern hemisphere is observed. At the botton, both versions are compared with TOPEXobservations. Reduction of biases observed when new version is used (in green)encourages us to continue exploring proposed combination technique.

c) Diferential code biases (DCB)DCB of receivers and satellites were treated as additional unknowns and were assessedwith harmonic coefficients, assuming that they remain constant throughout each day. AsDCB daily variation over the 100 days analyzed was not signifactive, we are seeking for asingle solution for each instrumental biases in the whole period. It involves solving asystem of approximately 60 million equations at a time. We are working on this.

ConclusionsWe have achieved our purpose combining GPS and TOPEX/Jason TEC measurements in aconsistent model. The power of GPS relies upon its capability for providing a good temporalresolution for describing the variability of the vTEC distribution worldwide; its majorweaknesses are the need of involving some crude approximations for calibrating thesatellite and receiver biases and to interpolate the vTEC in open ocean regions rather farfrom the observing sites. In spite of not being totally free of instrumental biases, TOPEXand Jason provide an almost direct and accurate measurement of the vTEC withoutinvolving model approximations, and their observations are just located over ocean regionswhere GPS observations are rather scarce; the major limitation is the quite long interval ofapproximately three months, needed for covering all geographic longitudes and local timeswith TOPEX or Jason observations.

Over the ocean, LPIMGPS (plotted in red)extrapolates the structure of the EquatorialAnomaly, due to the lack of GPS data overthe sea. The influence of TOPEX data is seenin the profile of green line obtained withLPIMGPS+TP.Differences between values calculated andobserved of vTECs have been obtained:

ΔvTEC1=vTECLPIM_GPS+TP –vTECLPIM_GPS

ΔvTEC2= vTECTP– vTECLPIM_GPS+TP

ΔvTEC3= vTECTP– vTECLPIM_GP

Global vTEC maps and instrumental DCB estimation based on combined GPS and TOPEX/Jason measurementsIsabel Bibbó1,2, Claudio Brunini1,3, Francisco Azpilicueta1,3, Diego Janches4

1Facultad de Ciencias Astronómicas y Geofísicas, Universidad Nacional de La Plata, Argentina, 2Facultad de Ciencias Naturales y Museo, UNLP, 3CONICET, Argentina , 4 NorthWest Research Associates, CoRA Divission, USA.

M

0 0

( , ) cos 2 sin 2 (sin )24 24

lL

lm ml lm

l m

mh mhvTEC h a b P

4 k

i LL sTEC b b

Fig.1: IGS continuous tracking stations usedin this work.

Fig.2: Ground trucks performed by Topexsatellite during a complete cycle.

1

cos '

sTECF

vTEC z

Fig.6: Differences between values calculated and observed of vTECs as a function of LT

Fig.7: Differences between values calculated and observed of vTECs as a function of modip

Fig.3: vTEC global distribution for [14-16] hs of TU for day 008-2002 obtained withLPIMGPS+TP

Fig.4: <vTECGPS+TP>-<vTECGPS> global distribution for [14-16] hs of UT for [001-100] days of 2002.

Fig.5: vTEC measured by TOPEX and modeled according LPIM against TU

0

S

TEC Neds

ΔvT

EC

[TE

CU

]

4L

0.18( )

2.2 / C KvTEC R R

mm TECU

TPvTECvTEC

LPIM_GPSvTEC

LPIM_GPS+TP

2

3

ΔvTECΔvTEC

1ΔvTEC

ΔvT

EC

[TE

CU

]

ΔvT

EC

[TE

CU

]

2

3

ΔvTECΔvTEC

1ΔvTEC

ΔvT

EC

[TE

CU

]