claudio brunini , amalia meza , francisco azpilicueta …

A NEW IONOSPHERE MONITORING TECHNOLOGY BASED ON GPS

CLAUDIO BRUNINI1, AMALIA MEZA1, FRANCISCO AZPILICUETA1,MARIA ANDREA VAN ZELE3, MAURICIO GENDE3 and ALEJANDRO DIAZ2

1Universidad Nacional de La Plata, 2Instituto Tecnologico Buenos Aires, 3Consejo Nacional deInvestigaciones Cientıficas y Tecnicas, Argentina

Abstract. Although global positioning system (GPS) was originally planned as a satellite-basedradio-navigation system for military purposes, civilian users have significantly increased their accessto the system for both, commercial and scientific applications. Almost 400 permanent GPS trackingstations have been stablished around the globe with the main purpose of supporting scientific research.In addition, several GPS receivers on board of low Earth orbit satellites fitted with special antennas thatfocus on Earth’s horizon, are tracking the radio signals broadcasted by the high-orbiting GPS satellites,as they rise and set on Earth horizon. The data of these ground and space-born GPS receivers, readilyaccessible through Internet in a ‘virtual observatory’ managed by the International GPS Service,are extensively used for many researches and might possibly ignite a revolution in Earth remotesensing.

By measuring the changes in the time it takes for the GPS signals to arrive at the receiver as theytravel through Earth’s atmosphere, scientists can derive a surprising amount of information about theEarth’s ionosphere, a turbulent shroud of charged particles that, when stimulated by solar flares, candisrupt communications around the world. This contribution presents a methodology to obtain hightemporal resolution images of the ionospheric electron content that lead to two-dimensional verticaltotal electron content maps and three-dimensional electron density distribution. Some exemplifyingresults are shown at the end of the paper.

Keywords: GPS, ionosphere, global ionospheric maps

1. Introduction

1.1. IONOSPHERE

The ionosphere is that part of the upper atmosphere where the free electron densityis high enough to disturb the propagation of radio frequency electromagnetic waves(Hargreaves, 1992). Free electrons are mainly produced by the photoionization ofneutral atoms and molecules of the atmosphere evoked by the UV solar radiation,but many complex physical phenomena in the solar-terrestrial environment partic-ipate in the production and loosing of electrons and in determining their spatialdistribution and temporal variations.

The ionosphere can be divided into three broad geographical regions whosebehaviors are quite different. The equatorial or low latitude region, from about +30◦

to –30◦ of geomagnetic latitude, is characterized by the largest values of electroncontent and the presence of large gradients in the spatial distribution of the electron

Astrophysics and Space Science 290: 415–429, 2004.C© 2004 Kluwer Academic Publishers. Printed in the Netherlands.

416 C. BRUNINI ET AL.

density. In this region takes place the geomagnetic anomaly that produces two peaksof electron content at about 20◦ to north and south of the geomagnetic equator inthe sun-lighted hemisphere. The mid-latitude regions, from about ±30◦ to ±60◦

of geomagnetic latitude present the more regular variations, although ionosphericstorms can bring sudden changes up to about 20% or more of its total electroncontent. The polar or high latitude regions are dominated by the geomagnetic fieldand their changes are rather unpredictable.

The temporal variability is dominated by the Sun. Apart from the night-dayperiodicity, quite regular variations are associated with the solar cycle of about11 years and the seasons.

What is now called ionosphere—the name was introduced by Watson Wattaround 1930—has been studied for more than 100 years using different obser-vational techniques, today considered classical. A large contribution was madeby a global network of 100–200 vertical incidence ionozondes that started opera-tion during the International Geophysical Year 1957–1958. Incoherent backscatterradars were used after 1958 to explore the topside. In 1957 the spatial era beganenabling topside ionozondes on board satellites, observations of Farady rotation ontransionospheric signals emitted by geostationary satellites, Doppler method withrockets and satellites and probe techniques aboard spacecrafts.

Even when there are not defined boundaries, it is accepted that the ionosphereextends from about 50 to 2000 km above the Earth surface. Below 50 km the ionizingradiation has been completely absorbed by the higher layers of the atmosphere,while above 2000 km the atmospheric density is too low to produce any appreciableionization. Four broad regions called D, E, F and topside can be recognized in thevertical structure of the ionosphere (Figure 1). These regions may be further dividedinto several regularly occurring layers, such as F1 or F2.

Using large databases of classical observations covering different geographicalregions and different solar and geomagnetic conditions, several empirical iono-spheric models were established. Among them, the International Reference Iono-sphere (IRI) is probably the most widely used. IRI is continuously revised andupdated through an international cooperative effort sponsored by the Committeeon Space Research and the Union of Radio Sciences (Bilitza, 1990).

Thanks to many years of efforts the Climatology of the ionosphere, under-stood as the capability of predicting the mean condition of the electron densityand its quasi-periodic variations, is today well known. The variation of the so-lar activity and the plasma emission from solar corona produce, however, dra-matic changes in the spatial environment around the Earth. The study of thesechanges is a branch of knowledge called space weather (Radicella, 2000). Theuse of sophisticated high technology systems for telecommunication, navigationand space missions, has created the need to predict the meteorological condi-tions of the space around the Earth. Disruptions of the ionosphere caused by mas-sive solar flares can interfere with or even destroy communication systems, Earth

IONOSPHERE MONITORING BASED ON GPS 417

Figure 1. Vertical profile of the ionosphere (after Hargreaves, 1992).

satellites and power grids on Earth and can cause losses of millions of dollars indamages.

1.2. GPS AND EARTH SCIENCES

The orbiting component of global positioning system (GPS) consists of 24 satel-lites plus spares: four in each of six different orbital planes inclined 55◦ from theEarth’s equatorial plane. Orbiting about 20 000 km above the Earth’s surface, inalmost circular orbits, all satellites have periods of 11 h and 58 min. The satellitesare distributed within their planes so that from almost any place at least four areabove the horizon at any time (for details about GPS see Kleusberg and Teunissen,1996).

Today there are several thousands permanent GPS geodetic receivers aroundthe world, some in dense regional arrays and others widely scattered to help definethe International Terrestrial Reference Frame and to support global and regionalgeodetic researches. Figure 2 illustrates the current distribution of permanent globalGPS sites overseen by the International GPS Service (http://igscb.jpl.nasa.gov),composed by about 300 stations. In spite of some sizable coverage gaps (particularly


Figure 2. IGS network of permanent GPS tracking stations (after http://igscb.jpl.nasa.gov).

in Africa, parts of Asia and South America and vast ocean areas) that limit thecurrent global performance, the observations of the network provide quite goodglobal coverage.

In spite of its military origin, GPS applications have rapidly expanded into anunforeseen universe of scientific applications. Without exaggeration it could be saidthat GPS can now probe from the center most point of the Earth—the geocenter—tothe outermost boundary of the Earth system—the edge of the ionosphere (Table I).GPS is now well established as the workhorse technique for scientific geodesy,which encompasses determination of precise satellite orbits, measurement of de-tailed motion and deformation of the Earth’s tectonics plates, precise location of the

TABLE I

Contributions of GPS to earth sciences

Solid earth

Location and motion of the geocenter; structureof the deep interior; deformation of crustand lithosphere; post-rebound glacialdeformation; Earth rotation and polarmotion; shape of the Earth.

Oceans

Significant wave height; ocean geoid and globalcirculation; short term eddy scale circulationand surface winds; sea state.

Lower atmosphere

Climate change and weather modeling;profiles of atmospheric parameters;structure of boundary layers; winds;waves and turbulence; watervapor distribution.

Upper atmosphere

High resolution 2- and 3-D ionosphericimaging; interactions with solar radiation,solar wind and geomagnetic field; monitoringspace storms; energy transport mechanismsand calibration for altimetry.


Earth’s center of mass, variations in the Earth’s rotation and pole motion, etc. Appli-cations of GPS to Earth sciences include high-resolution two- and three-dimensionalionospheric imaging and atmospheric limb sounding to recover precise profiles oftroposphere parameters.

1.3. IONOSPHERIC MODELING USING GPS OBSERVATIONS

Ionospheric models could be roughly classified as either theoretical or empirical.Although the former are generally more complicated and can describe qualita-tively the main characteristics of the ionosphere, they lack precision. The em-pirical models are fitted using average values obtained from large databases thatgather information collected from Earth, rockets and artificial satellites, at dif-ferent times during the day, at different epochs during the year and during vary-ing levels of solar and geomagnetic activity (Llewellyn and Bent, 1973; Bilitza,1990).

GPS-based ionospheric models are a particular type of empirical models in thesense that are able to describe ‘the weather of the day’, while classical empiricalmodels are usually designed to provide monthly mean values. From the point ofview of ionospheric studies, GPS represents the latest generation of satellites thatmay be used to study the ionosphere. Researches on the matter can be traced back tothe 1980s (Kleusberg, 1986; Feess and Stephens, 1987; Lanyi and Roth, 1988; Wildet al., 1989) and were multiplied in the following years (Coster et al., 1992; Hajjet al., 1994; Wilson et al., 1995; Schaer et al., 1996; Jakouski et al., 1996; Daviesand Hartmann, 1997). A detailed description about applications of GPS to iono-spheric studies and a very comprehensive list of references is provided by Manucciet al. (1999). Many researches were focused on GPS observations analysis to studyirregularities and perturbations in the ionosphere, including those caused by largegeomagnetic storms (Van Dierendonck et al., 1993; Coco et al., 1995; Coker et al.,1995; Ho et al., 1996, 1998; Aarons et al., 1997; Pi et al., 1997; Afraimovich et al.,2000).

On May 1998 IGS created the Ionosphere Working Group (Feltens and Schaer,1998), and soon after five different centers started computing and making avail-able several GPS-derived ionospheric products, mainly two-dimensional world-wide grids of vertical total electron content. To make feasible interchanges andcomparisons, the so-called IONEX (Ionosphere Map Exchange) standard formatwas established (Schaer et al., 1998). The five different centers that currently de-liver VTEC maps to IGS are Jet Propulsion Laboratory (Manucci et al., 1998),European Space Agency (Feltens and Schaer, 1998), Center for Orbit Determina-tion in Europe (Schaer, 1999), Universidad Politecnica de Cataluna (Hernandez-Pajares et al., 1999) and Energy Mines and Resources of Canada. These usedifferent algorithms to generate grids of VTEC with time resolution of at least2 h.


2. GPS-Based Ionospheric Models Developed by La Plata Ionospheric Group

2.1. EXTRACTING IONOSPHERIC INFORMATION

FROM GPS OBSERVATIONS

Each GPS satellite broadcasts two carries at frequencies of approximately 1.5 GHz(L1) and 1.2 GHz (L2), both of them modulated by binary pseudorandom codes(Kleusberg and Teunissen, 1996). A GPS receiver is able to correlate the incomingsatellite signals with replicas of them generated by the receiver. The correlation lagis related with the propagation time of the satellite signal, and then, with the geomet-rical satellite-receiver range. Double-frequency geodetic receivers provide differenttypes of simultaneous range measurements for every in-view satellite, some relatedwith the phase of the modulations and other associated with the phase of the car-riers. Several biases affect the measurements, the most important of which are thetropospheric and ionospheric range delays, receiver and satellite clocks errors andcarrier phase ambiguities. From the point of view of the current discussion, the mainfeature of that biases is whether they depend or not on the frequency of the carrier.When simultaneous carrier phase observations in both frequencies are subtracted,the satellite-receiver geometrical range and all frequency independent biases areremoved and the so-called geometry–free linear combination, �4, is obtained:

�4 = �1 − �2 = �I + τR + τ S, (1)

where �1 and �2 are the carrier phase observations at both frequencies (correctedby the carrier phase ambiguities); �I is the difference between the ionospheric rangedelays for the carriers L1 and L2; and τ R and τ S are inter-frequency electronic rangedelays produced in the hardware of the receiver and the satellite, respectively.

In the range of frequencies of GPS signals the ionosphere behaves as a dispersivemedium and the Appleton–Hartree theory provides the refraction index (de Munckand Spolastra, 1992). Based on this result, the ionospheric range delay, I f , at fre-quency f , can be computed. It results directly proportional to the electron densityintegrated along the slant path of the signal from the satellite to the receiver andinversely proportional to the square of the frequency of the carrier:

I f = −40.3

f 2

∫slant

N ds, (2)

where N is the three-dimensional electron density distribution. From Eqs. (1) and(2):

�4 = k∫

slantN ds + τR + τ S, (3)

where k = −40.3( 1f 21

− 1f 22

).


Subtracting observations made simultaneously in both frequencies cancels allfrequency–independent biases and highlights the ionospheric information sought.Unfortunately, this information is biased by the electronic inter-frequency rangedelays produced in the hardware of receivers and satellites.

2.2. TWO-DIMENSIONS GLOBAL IONOSPHERIC MODELING

To model the ionosphere in two dimensions, the so-called thin layer approximationis adopted (Manucci et al., 1999; Schaer, 1999). This model represents the iono-sphere through a thin spherical shell with equivalent total electron content, locatedapproximately 400 km above the Earth’s surface, close to the peak of the electrondistribution (Figure 3). Signals coming from a satellite S cross the thin layer atthe so-called piercing point P, with a zenith distance z′, and reach the receiver E,with a zenith distance z. The so-called solar-fixed coordinate system, X, Y, Z—ageocentric system co-rotating with the sun—is adopted. In such a system the sunstays practically quiet, then temporal variations of the electron content are slowand can be averaged for some short periods of time. The coordinates to describethe bi-dimensional distribution of the total electron content over the shell are thesolar-fixed longitude λ, and the geomagnetic latitude φ.

The slant total electron content along the ray path of the signal,∫

slant N ds, isrelated with the vertical total electron content along the vertical through the piercingpoint,

∫vertical N dv, using the simple approximation

∫vertical N dv ∼= 1

cos z′∫

slant Nds.The bi-dimensional distribution of the vertical total electron content is representedby a spherical harmonics expansion, whose coefficients are kept constant for some

Figure 3. Basic geometry for the thin layer model in the solar-fixed co-ordinate system.


period of time, typically 2 h (Brunini, 1998). Then, Eq. (3) leads to the final ex-pression of the equation of observation:

�4 = k1

cos z′

L∑l=0

M∑m=l

{alm cos

(2π

mλ

24

)

+ blm sin

(2π

mλ

24

)}Plm(sin φ) + τR + τ S. (4)

Equation (4) contains as unknowns the coefficients of the expansion and theinter-frequency electronic range delays for every receiver and every satellite. Theseunknowns are fitted by least squares using the GPS observations belonging to theglobal network. Typically we compute daily batch solutions in which a constantvalue for every receiver and every satellite inter-frequency electronic range delayis estimated, and a different set of constant coefficients are adjusted for each 2 hinterval.

All the information necessary to generate this type of models (satellite ephemerisand GPS observations) can be downloaded via anonymous ftp. It is thereforepossible to fit a large time series of models and plot the corresponding globalVTEC maps to perform a large scale temporal analysis of the VTEC variabilityand to study its correlation with geomagnetic and solar activity (Brunini et al.,2002).

2.3. THREE-DIMENSIONS GLOBAL IONOSPHERIC MODELING

The model in two dimensions represents the ionosphere as a simple spherical layerof depreciable width. It only accounts for the integrated effect of the free electronsfrom the upper to the lower part of the ionosphere. While ground-based ionosphericmaps represent a big advance for ionosphere weather, they are only bi-dimensionaland give almost no information on the vertical electron distribution. This limita-tion can be removed by the introduction of horizontal cuts through the ionosphereaffordable by space born GPS receivers.

At the same time that ground-based GPS geodesy was being developed in the1980s, investigators were turning their attention to the use of flight GPS receivers onlow orbiting Earth satellites for a variety of applications in scientific Earth remotesensing. Combined data from a large number of ground-based and some space GPSreceivers will enable high resolution two- and three-dimensional snapshot imagingof the global ionosphere.

Computerized tomography is a technique, whereby, an image of an object is con-structed from a set of projections, or integrated densities, taken along many linesthrough the object. This technique is used in medicine to produce two-dimensionalcross-sectional X-ray images and more recently is used in three-dimensional acous-tic and seismic tomography to analyze the oceans and solid Earth.


Three dimensional tomography of the ionosphere has been performed by theionospheric group of Universidad Nacional de La Plata using ground base GPSreceivers and NASA GPS Met observations (Meza, 1999). Sub-daily imaging ofthe global VTEC and electron density profiles were obtained.

An expansion in spherical harmonics dependent on the solar-fixed longitude λ

and the geographic latitude φ, analogous to that used for two-dimensional model,is kept in this case. The vertical profile of the electron density is modeled througha Chapman type function f (h) dependent on the height h over the Earth surface(Hargreaves, 1992) as follows:

N (λ, φ, h) = f (h)L∑

l=0

M∑m=l

{alm cos

(2π

mλ

24

)

+ blm sin

(2π

mλ

24

)}Plm(sin φ) and

f (h) = exp

(1 −

(h − hm

H

)− exp

(−h − hm

H

)), (5)

where hm is the height of maximum electron density and H is the scale height of theChapman profile. Introducing Eq. (5) into Eq. (3), the expression of the equationof observation for three-dimensional modeling is obtained:

�4 = kL∑

l=0

M∑m=l

alm

∫path

f (h) cos

(2π

mλ

24

)Plm(sin φ) ds

+ blm

∫path

f (h) sin

(2π

mλ

24

)ds + τR + τ S. (6)

The integrals of last equation depend on the path of the signals from the GPSsatellites to either the ground-based or the space-born receivers, and can be evaluatedbecause the coordinates of satellites and receivers are known. The coefficients ofthe expansion are estimated, together with the electronic range delays, using theleast square method.

2.4. REGI ONAL VTEC M ODELS

The term regional within this context must be understood as an area covered by anetwork of permanent stations where the baseline lengths connecting any pairs ofstations are no larger that 500 km.

The starting point for this type of development is Eq. (3), previously presentedwith a detailed consideration of its problems derived from the appearance of theelectronic receiver and satellite range delays and transformation from slant to


vertical total electron content. The assumptions specified in Section 2.2 about thethin layer model are still valid (for details, see Brunini et al., 2001).

As now the objective is just to model a region of the ionosphere, VTEC mustdepend on three coordinates: two associated with the space distribution of totalelectron content and the third one accounting for the time changes suffered by theionosphere along the day. So VTEC should be represented by a function of time tand of geographic longitude λg and latitude φ.

The strategy chosen is to fix the central station of the network, as the statingpoint for a spatial series expansion in the form of a polynomial of degree n, usuallywith coefficients depending on t following a trigonometric-term polynomial:

VTEC (λg, φ, t) = a0(t) +n+1∑i=1

ai (t) (λg − λ0)i−1(φ − φ0)n−i−1 and

(7)

ai (t) =m∑

j=1

[αi j cos

(2π j

24t

)+ βi j sen

(2π j

24t

)]i = 0, n,

where λ0 and φ0 are the geographic longitude and latitude of the central station.Usual values for n are 2 or 4 and for m is 12. These parameters determine the

spatial and temporal resolution capacity of the model.Once the expression for VTEC and the ‘obliquity factor’ 1

− cos z′ are replaced inEq. (3) the complete set of unknowns is composed of the (n + 1) × 2 × m coefficientsof the expansion plus one additional electronic range delay τ R for each receiver andτ S for each satellite involved in the observations. Then, a least squares adjustmentis performed using a data package corresponding to one day of observations fromthe complete network of stations, estimating in this way values for the unknowns.

The importance of regional models is connected to the need of the geodesycommunity to correct the effect caused by the ionosphere on the propagation ofan electromagnetic wave, being the typical case users of single frequency GPSreceiver. But apart from this, the availability of a good ionospheric model could behelpful when determining hi-precision networks even when using double frequencyGPS receiver and can importantly reduce the observation time needed for precisedetermination.

3. Examples of Two- and Three-Dimensional Ionospheric Imaging

A detailed description of the results that are shown here as examples were extractedfrom Meza et al. (2000, 2002a, 2002b, 2002c) and Brunini et al. (2002). Figures 4aand 4b show examples of global quasi real time snapshot of the ionospheric verticaltotal electron content produced at La Plata.

Figure 4a shows a typical distribution of the VTEC in a quiet geomagneticconditions. Typical prominent and well-known features can be identified: the


Figure 4. (a) Vertical total electron content imaging of the global ionosphere obtained at La Plata fora period of two hours, on December 28, 1997. Coordinates are solar-fixed longitude (–12h to +12h)and geomagnetic latitude (−90◦ to +90◦); values are in TEC Units (1 TECU = 1016 electrons/m2).(b) Series of ionospheric snapshots every 4 h during the May 15, 1997 geomagnetic storm.

maximum delayed with respect to the local noon and shifted to the southern hemi-sphere, corresponding with the date close to the southern summer solstice; theequatorial anomalies with its typical structure of two peaks of VTEC and the in-between valley following the curvature of the geomagnetic equator in the sun lighted


hemisphere; etc. Figure 4b shows 12 selected 2-h VTEC maps, from 15th to16th May, covering the development of the May 15, 1997 geomagnetic storm, thestrongest of that year. The impact of the storm over the VTEC distribution is quiteclear in contrast with the quiet behavior shown in the previous figure. In particular,a large depletion of the VTEC takes place during the recovery phase of the storm.A deeper study shows a good correlation between the outstanding features of theglobal electron content during the storm and some well established geomagneticindices, confirming the ability of GPS to identify large scale storm features.

Figure 5 shows different vertical profiles of electron density for zero solar-fixedlongitude (the direction towards the sun) and different latitudes. Figures 5a–5cshow profiles adjusted with La Plata model for February 3rd, 4th and 5th, 1996,while Figure 5d shows the corresponding profile computed with the IRI model.

Figure 5. Vertical profiles of electron density for zero solar-fixed longitude and different latitudes(shown in the box below each figure). (a), (b) and (c), show results obtained with La Plata model and(d) with IRI model.


Keeping in mind that GPS gives sub-daily resolution while IRI provides monthlymean values, discrepancies between the two models are rather small.

4. Conclusions

Even when the current geographical coverage of the IGS tracking network limitsthe spatial and time resolution of the ionospheric global models, GPS represents anunprecedented opportunity for ionospheric scientists, since its observations providealmost global coverage with simultaneity and time continuity, at low cost for theusers and are readily accessible. In spite of the fact that the rather simple empiricalmodels used to fit the observations must be improved, we believe that GPS iono-spheric maps are of great help for a better understanding of the complex ionosphericenvironment and the global response of the ionosphere to geomagnetic storms.

There is plenty of effort in processing GPS data to form VTEC maps, but webelieve that less effort is currently spent on their validation and interpretation. DailyVTEC maps with 15-min resolution from Jet Propulsion Laboratory are availablesince 1995 and global maps with 2-h resolution, submitted by five IGS IonosphericAnalysis Centers, are available in IONEX format, for at least last three years. Ina further work we will try to combine them with other ionospheric data to studyglobal ionospheric climatology and global ionospheric response to geomagneticstorms.

Several dual-frequency GPS receivers on board low Earth orbits satellites likethe German Challenging Minisatellite Payload (Champ) and the Argentine Satelitede Aplicaciones Cientificas-C (SAC-C), fitted with special antennas that focus onEarth’s horizon, are tracking the radio signals broadcast by each of the 28 high-orbiting GPS satellites as they rise and set on Earth horizon. A single GPS receiverin low orbit can acquire more than 500 soundings a day, spread uniformly acrossthe globe. This large amount of data, free available for the scientific community,will enhance the knowledge of the three-dimensional structure of the ionosphere.

The biggest advantage of the ionospheric monitoring technology based on GPSmay well be its low cost. GPS receivers, comparable in size and complexity to anotebook computer, can be built for a fraction of the cost of traditional space-bornesensors and placed unobtrusively on many low-orbiting spacecraft. Since mostEarth satellites already carry such devices for timing and navigation, employingthose instruments for science purposes might possibly ignite a revolution in Earthremote sensing.

Acknowledgements

The authors would like to thank the organizer for the invitation to participate onthe Eleventh United Nations/European Space Agency Workshop on Basic Space


Science, hosted by the Comision Nacional de Actividades Espaciales (CONAE),Universidad Nacional de Cordoba, and Universidad Nacional de la Plata, on be-half of the Government of Argentina, September 9–13, 2002, Falda del Carmen,Cordoba, Argentina, in which this contribution was presented.

References

Aarons, J., Mendillo, M. and Yantosca, R.: 1997, Radio Sci. 32(4), 1535–1550.Afraimovich, E.L., Altyntsev, A.T., Kosogorov, E.A., Larina, N.S. and Leonovich, L.A.: 2000, J.

Atmos. Sol. Terr. Phys. 63(17), 1841–1849.Bilitza, D. (ed.): 1990, International Reference Ionosphere, NSSDC 90-22, Greenbelt, MD.Brunini, C.: 1998, PhD Thesis, Universidad Nacional de La Plata, Argentine, p. 191.Brunini, C., Meza, A. and Dıaz, A.: 2001, Regional Vertical Total Electron content Using GPS

Observations, Proceedings of the IAG 2001 Scientific Assembly, CD edition, September 2–7,Budapest, Hungary.

Brunini, C., Van Zele, A., Meza, A. and Gende, M.: 2002, J. Geophys. Res., Space Phys. 108 (A2)10.1029.2002 IA009346, 2003.

Coco, D.S., Gaussiran, T.L. and Coker, C.: 1995, Radio Sci. 30(6), 1869–1874.Coker, C., Hunsucker, R. and Lott, G.: 1995, Geophys. Res. Lett. 22(23), 3259–3262.Coster, A.J., Gaposchkin, E.M. and Thorton, L.E.: 1992, Navigation 39, 191–204.Davies, K. and Hartmann, G.K.: 1997, Radio Sci 32(4), 1695–1703.de Munck, J.C. and Spolastra, T.A.Th. (eds.): 1992, Proceedings of the Symposium on Refraction of

Transatmospheric Signals in Geodesy, The Netherlands Geodetic Commission, Publications onGeodesy, Delft, The Netherlands, No. 36, New Series.

Feess, W.A. and Stephens, S.G.: 1987, IEEE Trans. Aerosp. Electron. Syst. 23(3), 332–338.Feltens, J. and Schaer, S.: 1998, in: J. M. Dow, J. Kouba and T. Springer (eds.), IGS Products for the

Ionosphere, Proceedings of the IGS Analysis Center Workshop, 225–232, Darmstadt, February9–11.

Hajj, G.A., Ibanez-Meier, R., Kursinski, E.R. and Romans, L.J.: 1994, Int. J. Img. Syst. Technol. 5(2),174–178.

Hargreaves, J.K.: 1992, The Solar-Terrestrial Environment, Cambridge Atmospheric And Space Sci-ence Series, Cambridge University Press, p. 420.

Hernandez-Pajares, M., Juan, J.M. and Sanz, J.: 1999, J. Atmos. Solar Terr. Phys. 61, 1237–1247.

Ho, C.M., Manucci, A.J., Lindqwister, U.J., Pi, X. and Tsurutani, B.T.: 1996, Geophys. Res. Lett. 23,3219–3222.

Ho, C.M., Mannucci, A.J., Sparks, L., Pi, X., Lindqwister, U.J., Wilson, B.D., Iijima, B.A. and Reyes,M.J.: 1998, J. Geophys. Res. 103(A11), 26409–26420.

Jakouski, N., Sardon, E., Egler, E., Jungstand, A. and Klahn, D.: 1996, in: G. Beutler, G.W. Hein,W.G. Melbourne and G. Seebr (eds.), About the use of GPS measurements for ionospheric studies,GPS Trends in Precise Terrestrial Airborne and Spaceborne Applications, IAG Symposium 115,Springer Verlag, New York.

Kleusberg, A.: 1986, Manuscr. Geod. 11, 256–261.Kleusberg, A. and Teunissen P. (eds.): 1996, GPS for Geodesy, Lecture Notes in Earth Sciences,

Vol. 60, Springer, p. 407.Lanyi, G.E. and Roth, T.: 1988, Radio Sci. 23(4), 483–492.Llewellyn, S.K. and Bent, R.: 1973, Documentation and description of Bent ionospheric model,

Report AFCRL-TR-73-0657, AD772733, Geophysics Laboratory, Hanscom Air Force Base.


Manucci, A.J., Wilson, B.D., Yuan, D.N., Ho, C.M., Lindqwister, U.J. and Runge, T.F.: 1998, RadioSci. 33(3), 565–582.

Manucci, A.J., Iijima, B.A., Lindqwister, U.J., Pi, X., Sparks, L. and Wilson, B.D.: 1999, GPS andIonosphere, Revised Submission to URSI Reviews of Radio Science, Jet Propulsion Laboratory,Pasadena, CA.

Meza, A.: 1999, Three Dimensional Ionospheric Models from Earth and Space Based Gps Observa-tions, PhD Thesis, Universidad Nacional de La Plata, Argentine, p. 183.

Meza, A., Brunini, C. and Kleusberg, A.: 2000, Geofısica Internacional 39(1), 21–27.Meza, A., Diaz, A., Brunini, C. and Van Zele, A.: 2002a, Radio Sci. 37(3), 9—1,

10.1029/2001RS002482.Meza, A., Brunini, C. and Kleusberg, A.: 200b, Adv. Space Res. 30(2), 307–312.Meza, A., Brunini, C., Bosch, W. and Van Zele, A.: 2002c, Adv. Space Res. 30(2), 401–406.Pi, X., Manucci, A.J., Lindqwister, U.J. and Ho, C.M.: 1997, Geophys. Res. Lett. 24(18), 2283–2286.Radicella, S.M.: 2000, Space weather, Fısica de la tierra, Vol. 12, 189–204.Schaer, S., Beutler, G., Rothacher, M. and Soringer, T.: 1996, Daily Global Ionosphere Maps Based

on GPS Carrier Phase Data Routinely Produced by the CODE Analysis Center, Proceedings ofthe IGS Analysis Center Workshop, Silver Spring, Maryland, March 19–21.

Schaer, S., Gurtner, W. and Feltens, J.: 1998, in: J.M. Dow, J. Kouba and T. Springer (eds.), IONEX:The IONosphere Map EXchange format Version 1, Proceedings of the IGS Analysis CenterWorkshop, 233–247, Darmstadt, February 9–11.

Schaer, S.: 1999, PhD Thesis, Bern University.Van Dierendonck, A.J., Klobuchar, J. and Hua, Q.: 1993, Ionospheric Scintillation Monitoring Us-

ing Commercial Single Frequency C/A Code Receivers, Proceedings of the Sixth InternationalTechnical Meeting of Satellite Division of the Institute of Navigation ION-GPS 93, 1333–1342.

Wild, U., Beutler, G., Gurtner, W. and Rothacher, M.: 1989, Estimating the Ionosphere Using One orMore Dual Frequency GPS Receivers, Proceedings of the Fifth International Geodetic Symposiumon Satellite Positioning, 724–736, Las Cruces, New Mexico, 13–17 March.

Wilson, B.D., Manucci, A.J. and Edwards, C.D.: 1995, Radio Sci. 30(3), 639–648.

claudio brunini , amalia meza , francisco azpilicueta …

Documents