Centre de documentation
scientifique

Regional gravity field modeling using airborne gravimetry data / Bas Alberts (2009)  

Public Titre : Regional gravity field modeling using airborne gravimetry data Type de document : Thèse/HDR Auteurs : Bas Alberts, Auteur Editeur : Delft : Netherlands Geodetic Commission NGC Année de publication : 2009 Collection : Netherlands Geodetic Commission Publications on Geodesy, ISSN 0165-1706 num. 70 Importance : 180 p. Format : 17 x 24 cm ISBN/ISSN/EAN : 978-90-6132-312-9 Note générale : Bibliographie Document en téléchargement sur le site de NCG : lien dans la notice Langues : Anglais (eng) Descripteur : [Vedettes matières IGN] Géodésie physique [Termes IGN] champ de pesanteur local [Termes IGN] Chili [Termes IGN] espace de Hilbert [Termes IGN] gravimétrie aérienne [Termes IGN] gravimétrie en mer [Termes IGN] levé gravimétrique [Termes IGN] méthode des moindres carrés [Termes IGN] modèle de géopotentiel local [Termes IGN] Nord, mer du [Termes IGN] Ontario (Canada) [Termes IGN] pondération [Termes IGN] processus [Termes IGN] traitement automatique de données Index. décimale : 30.42 Gravimétrie Résumé : (Auteur) Airborne gravimetry is the most efficient technique to provide accurate high-resolution gravity data in regions that lack good data coverage and that are difficult to access otherwise. With current airborne gravimetry systems gravity can be obtained at a spatial resolution of 2 km with an accuracy of 1-2' mGal. It is therefore an ideal technique to complement ongoing satellite gravity missions and establish the basis for many applications of regional gravity field modelling. Gravity field determination using airborne gravity data can be divided in two major steps. The first step comprises the preprocessing of raw in-flight gravity sensor measurements to obtain gravity disturbances at flight level and the second step consists of the inversion of these observations into gravity functionals at ground level. The preprocessing of airborne gravity data consists of several independent steps such as low-pass filtering, a cross-over adjustment to minimize misfits at cross-overs of intersecting lines, and gridding. Each of these steps may introduce errors that accumulate in the course of processing, which can limit the accuracy and the resolution of the resulting gravity field. For the inversion of the airborne gravity data at flight level into gravity functionals at the Earth's surface, several approaches can be used. Methods that have been successfully applied to airborne gravity data are integral methods and least-squares collocation, but both methods have some disadvantages. Integral methods require that the data are available in a much larger area than for which the gravity functionals are computed. A large cap size is required to reduce edge effects that result from missing data outside the target area. Least-squares collocation suffers much less from these errors and can yield accurate results, provided that the auto-covariance function gives a good representation of data in- and outside the area. However, the number of base functions equals the number of observations, which makes least-squares collocation numerically less efficient. In this thesis a new methodology for processing airborne gravity data is proposed. It combines separate preprocessing steps with the estimation of gravity field parameters in one algorithm. Importantly, the concept of low-pass filtering is replaced by a frequency-dependent data weighting to handle the strong colored noise in the data. Frequencies at which the noise level is high get a lower weight than frequencies at which the noise level is low. Furthermore, bias parameters are estimated jointly with gravity field parameters instead of applying a cross-over adjustment. To parameterize the gravity potential a spectral representation is used, which means that the estimation results in a set of coefficients. These coefficients are used to compute gravity functional at any location on the Earth's surface within the survey area. The advantage of the developed approach is that it requires a minimum of preprocessing and that all data can be used as obtained at the locations where they are observed. The performance of the developed methodology is tested using simulated data and data acquired in airborne gravimetry surveys. The goal of the simulations is to test the approach in a controlled environment and to make optimal choices for the processing of real data. For the numerical studies with simulated data, the new methodology outperforms the more traditional approaches for airborne gravity data processing. For the application of the developed methodology to real data, three data sets are used. The first data set comprises airborne gravity measurements over the Skagerrak area, obtained as part of a joint project between several European institutions in 1996. This survey provided accurate airborne gravity data, and because good surface gravity data are available within the area, the data set is very useful to test the performance of the approach. The second data set was obtained by the GeoForschungsZentrum Potsdam during a survey off the coast of Chile in 2002. This data set, which has a lower accuracy than the first data set, is used to investigate the estimation of non-gravitational parameters such as biases and scaling factors. The final data set that is used consists of airborne gravity data acquired by Sander Geophysics Limited in 2003. The survey area is located near Timmins, Ontario and is much smaller than the area of the other data sets. The small size of the area and the high accuracy of the data make it a challenging data set for regional gravity modeling. The computational experiments with real data show that the performance of the developed methodology is at the same level as traditional methods in terms of gravity field errors. However, it provides a more flexible and powerful approach to airborne gravity data processing. It requires a minimum of preprocessing and all observations are used in the determination of a regional gravity field. The frequency-dependent data weighting is successfully applied to each data set. The approach provides a statistically optimal solution and is a formalized way to handle colored noise. A noise model can be estimated from a posteriori least-squares residuals in an iterative way. The procedure is purely data-driven and, unlike low-pass filtering, does not depend on previous experience of the user. The developed methodology allows for the simultaneous estimation of non-gravitational parameters with the gravity field parameters. A testing procedure should be applied, however, to avoid insignificant estimations and high correlations. For the Chile data set a significant improvement of the estimated gravity field is obtained when bias and scale factors are estimated from the observations. The results of the computations with the real data sets show the high potential of using airborne gravimetry to obtain accurate gravity for geodetic and geophysical applications. Note de contenu : 1 Introduction 1.1 Background 1.2 Objectives 1.3 Outline 2 Airborne gravimetry 2.1 Historical overview 2.2 The principle of airborne gravimetry 2.3 Mathematical models 2.4 Applications and opportunities 3 Processing of airborne gravity data 3.1 Pre-processing 3.2 Inversion of airborne gravity data 3.3 Discussion 4 Combined data processing and inversion 4.1 Gravity field representation 4.2 Inversion methodology 4.3 Regularization and parameter choice rule 4.4 Frequency-dependent data weighting 4.5 Estimation of non-gravitational parameters 4.6 Edge effect reduction 4.7 Combination with prior information 5 Application to simulated data 5.1 Description of the data 5.2 Computations with noise-free data 5.3 Computations with data corrupted by white noise 5.4 Computations with data corrupted by colored noise 5.5 Bias and drift handling 5.6 Summary of the optimal solution strategy 6 Application to airborne gravimetric survey data 6.1 Skagerrak data set 6.2 Chile data set 6.3 Timmins, Ontario data set 6.4 Summary and discussion 7 Conclusions and recommendations 7.1 Conclusions . 7.2 Recommendations. A Pre-processing of airborne gravity data A.1 GPS processing A..2 Gravity processing B Coordinate transformation C Least-squares collocation and Hilbert spaces C.1 Definition of a Hilbert space and some properties C.2 Reproducing kernel Hilbert spaces C.3 Least-squares collocation D Derivation of the ZOT regularization matrix E Modification of the base functions Numéro de notice : 15494 Affiliation des auteurs : non IGN Thématique : POSITIONNEMENT Nature : Thèse étrangère DOI : sans En ligne : https://www.ncgeo.nl/downloads/70Alberts.pdf Format de la ressource électronique : URL Permalink : https://documentation.ensg.eu/index.php?lvl=notice_display&id=62736

Exemplaires(1)

Code-barres	Cote	Support	Localisation	Section	Disponibilité
15494-01	30.42	Livre	Centre de documentation	Géodésie	Disponible