Catalogue en ligne IGN

Troposphere modeling and filtering for precise GPS leveling / Frank Kleijer (2004)

Public

Titre :	Troposphere modeling and filtering for precise GPS leveling
Type de document :	Thèse/HDR
Auteurs :	Frank Kleijer, Auteur
Editeur :	Delft : Netherlands Geodetic Commission NGC
Année de publication :	2004
Collection :	Netherlands Geodetic Commission Publications on Geodesy, ISSN 0165-1706 num. 56
Importance :	262 p.
Format :	17 x 24 cm
ISBN/ISSN/EAN :	978-90-6132-284-9
Note générale :	Bibliographie
Langues :	Anglais (eng)
Descripteur :	[Vedettes matières IGN] Rayonnement électromagnétique [Termes IGN] correction troposphérique [Termes IGN] filtre de Kalman [Termes IGN] modèle stochastique [Termes IGN] nivellement par GPS [Termes IGN] positionnement par GNSS [Termes IGN] propagation troposphérique [Termes IGN] simulation [Termes IGN] traitement de données GNSS [Termes IGN] troposphère
Index. décimale :	24.30 Rayonnement électromagnétique
Résumé :	(Auteur) Precision : In the Netherlands, a precision of 5 rnm for estimated GPS height differ ences is required to achieve comparable accuracy as geoid height differences. This precision can be achieved for 24-hour data sets when applying a proper modeling. Precise levelling with GPS requires a judicious modeling of tropospheric delays, which has physical, functional, and stochastic aspect. Modeling aspects : The physical modeling comprises zenith delays for the hydrostatic and wet component and zenith angle dependent mapping functions. Because the amount of water vapor in the atmosphere fluctuates widely and because the water-vapor aspect induced signal delays affect the height component strongly, a-priori modeling of these delays results in an insufficient precision of height differences. Parameterizing the tropospheric delay in the functional model is therefore necessary, at least for medium and long baselines. The observation model can further be strengthened by pseudo-observations. These pseudo-observations may be spatiotemporal constraints on tropospheric delay differences, or constraints on residual slant delays. With the latter type of constraint the isotropy assumption is loosened. An existing theoretical model is revised to obtain the corresponding covariance matrix. The stochastic modeling of both types of constraints is based on the assumption of Kohnogorov turbulence. Filtering : The observation models can be implemented in a recursive filter like the Kalman Filter or the SRIF Several variations of these filters are described. For fast computations the most suitable recursive filtering technique is the Kalman-Cholesky Filter with pre-elimination of temporal GPS parameters, such as clock errors and ionospheric delays. Some tests and reliability descriptions are worked out for this filter. For practical implementation the temporal behavior of the zenith wet delay is to be assumed a random-walk process, which gives a fair description. The zenith wet delay can be estimated every epoch or every pre-defined batch of epochs. Impact of model components : The effect of several model components on mainly the height is analyzed by simulation software. Special attention is given to the residual-slant-delay model because it is potentially precision and reliability improving. The impact of this model does however depend on the precision level of the observations and it still needs to be validated. Although observations to low-elevation satellites have a large contribution to the precision of the height, the residual-slant-delay model implies a strong down weighting of observations to satellites below ten degrees elevation. The highest accuracy can be obtained when the phase ambiguities are fixed. Even for long observation time spans this makes a difference of up effect on the formal precision of the height, but they have, a large influence on the precision of the filtered zenith delays. Because overconstraining (with constraints that are too tight) can have a large precision-deteriorating effect, spat iotemporal constraints are not recommended for GPS leveling. Further, short batches are preferred to avoid biases. To prevent the presence of near rank deficiencies, the zenith delays of one station are often fixed. From a precision point of view this is not necessary and is not recommended because this can also introduce biases. Even larger biases can be introduced when the zenith delays of all stations are fixed to their a-priori values. For very short baselines (< ±1 kin), this model is however justifiable because the formal precision improves considerately, keeping the effect of the biases in balance. to 15-20%. Batch size and spatiotemporal constraints turned out to have little effect on the formal precision of the height, but they have, a large influence on the precision of the filtered zenith delays. Because overconstraining (with constraints that are too tight) can have a large precision-deteriorating effect, spat iotemporal constraints are not recommended for GPS leveling. Further, short batches are preferred to avoid biases. To prevent the presence of near rank deficiencies, the zenith delays of one station are often fixed. From a precision point of view this is not necessary and is not recommended because this can also introduce biases. Even larger biases can be introduced when the zenith delays of all stations are fixed to their a-priori values. For very short baselines (< ±1 kin), this model is however justifiable because the formal precision improves considerately, keeping the effect of the biases in balance.
Note de contenu :	General introduction I) Troposphere delay modeling for space geodetic measurements Symbols and units in Part I 3 Introduction to Part I 4 Physics of the atmosphere 5 Zenith-delay models 6 Slant-delay models 7 Azimuthal asymmetry and gradient parameters 8 Conclusions of Part 1 A Temperature lapse rate B Saastamoinen integrals C Effective height Bibliography II) Paramerization of the tropospheric delay in GPS observation models Symbols of Part II 9 Introduction to Part II 10 GPS observation equations 11 Eliminating rank deficiencies in troposphere-fixed models 12 Troposphere-float and weighted models 13 Near rank deficiencies 14 Pre-elimination transformations 15 Conclusions of Part II III) Stochastic modeling of (slant) tropospheric delays observed by GPS Symbols of Part III 16 Introduction to Part III 17 Power-law processes 18 Stochastic modeling of troposphere constraints 19 Conclusions of Part III IV) Recursive GPS data processing Symbols of Part IV 20 Introduction to Part IV 21 Kalman filtering with pre-elimination 22 Pre-elimination in a recursive SRIF 23 Implementation aspects 24 Condition equations, testing, and reliability 25 Conclusions of Part IV V) Simulations Symbols of Part V 26 Introduction to Part V 27 Software implementation 28 Means of quality assessment 29 Simulation scenarios 30 The troposphere-fixed model 31 Conclusions of Part V 32 Conclusions and recommendations
Numéro de notice :	16056
Affiliation des auteurs :	non IGN
Thématique :	IMAGERIE
Nature :	Thèse étrangère
En ligne :	https://ncgeo.nl/downloads/56Kleijer.pdf
Format de la ressource électronique :	URL
Permalink :	https://documentation.ensg.eu/index.php?lvl=notice_display&id=55149

Exemplaires(2)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
16056-01	30.64	Livre	Centre de documentation	Géodésie	Disponible
16056-02	24.30	Livre	Centre de documentation	Physique	Disponible

Documents numériques

Troposphere modeling and filtering

Adobe Acrobat PDF

IGN

Centre de documentation
scientifique

Accueil

Sélection de la langue

Adresse

Se connecter

Actualité

L'actu ! Horaires du CDOS

Informations pratiques

Exemplaires(2)

Documents numériques

IGN / ENSG

L'IGN a pour vocation

Accès directs

2014 -2024 IGN

IGN

Centre de documentationscientifique

Accueil

Sélection de la langue

Adresse

Se connecter

Actualité

L'actu ! Horaires du CDOS

Informations pratiques

Exemplaires(2)

Documents numériques

IGN / ENSG

L'IGN a pour vocation

Accès directs

2014 -2024 IGN

Centre de documentation
scientifique