Difference between revisions of "IC SG3"
| Line 1: | Line 1: | ||
| − | <big>''' | + | <big>'''JSG 0.3: Comparison of current methodologies in regional gravity field modelling'''</big> |
| − | + | Chairs: ''M. Schmidt (Germany), Ch. Gerlach (Germany)''<br> | |
| − | Affiliation: ''Comm. | + | Affiliation: ''Comm. 2, 3'' |
__TOC__ | __TOC__ | ||
===Introduction=== | ===Introduction=== | ||
| − | + | Traditionally the gravitational potential of the Earth and other celestial bodies is modelled as a series expansion in terms of spherical harmonics. Although this representation is technically possible for ultra-high expansions, it is well-known that spherical harmonic approaches cannot repre-sent data of heterogeneous density and quality in a proper way. In order to overcome these and other deficiencies regional modelling comes into question. | |
| − | + | In the last years many groups have developed sophisticated approaches for regional modelling, e.g. the expansion of the gravity field or functionals of the field in terms of spherical (radial) base functions. Analogously to spherical harmonic approaches, also in regional modelling the un-known model parameters, i.e. the coefficients of the series expansion, can be either determined by means of numerical integration or as the solution of a parameter estimation process. Numerical integration techniques are widely used in the mathematical community and provide efficient and stable solutions. However, numerical integration tech-niques suffer from important disadvantages. Among others these methods (1) require the input data to be given on a spherical integration grid, (2) cannot provide estimated error variances and covariances of the model parameters and (3) have difficulties to handle the combination of data from different measurement techniques. Due to these dis-advantages, parameter estimation is the preferred strategy in the geodetic community. Although solutions in regional modelling based on parameter estimation are generated by several groups since many years, a large number of un-solved problems and open questions still remain. They mostly arise from the condition of the normal equation system and are therefore directly connected to the para-metrization of the gravity field, the type and distribution of observation data, the choice and location of base functions, possible regularisation schemes, etc. | |
| − | + | The aim of the proposed SG is to find guidelines on suit-able strategies for setting up the parameter estimation of regional gravity field modelling. This includes appropriate strategies for the combination of satellite, airborne and terrestrial data. The focus of the SG is on the methodologi-cal foundation of regional gravity field modelling based on series expansions in terms of localizing base functions. Therefore numerical studies will be concentrated on simu-lations based on synthetic data. It is not the aim of the SG to process and compare solutions from real data. | |
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | The | ||
| − | |||
| − | |||
===Objectives=== | ===Objectives=== | ||
| − | * | + | The main objectives of this SG are: |
| − | * | + | * to collect information of available methodologies and strategies for regional modelling, including |
| − | * | + | ** the type of base functions (splines, wavelets, Slepian function, Mascons, etc.), |
| − | * | + | ** the point grids for placing the functions (standard grid, icosaeder, Reuter grid, etc. on a sphere, ellipsoid, etc.), |
| − | * | + | ** the choice and establishment of an appropriate adjust-ment model (combination strategy, variance component estimation, rank deficiency problems, e.g., due to downward continuation, etc.), |
| − | * | + | ** the consideration of model errors (truncation errors, edge effects, leakage, etc.), |
| − | * | + | ** the specific field of application, |
| − | * | + | * to analyze the collected information in order to find spe-cific properties of the different approaches and to find, why certain strategies have been chosen, |
| − | * | + | * to create a benchmark data set for comparative numerical studies, |
| − | * | + | * to carry out numerical comparisons between different solution strategies for estimating the model parameters and to validate the results with other approaches (spheri-cal harmonic models, least-squares collocation, etc.), |
| − | * | + | * to quantify and interpret the differences of the compari-sons with a focus on detection, explanation and treatment of inconsistencies and possible instabilities of the differ-ent approaches, |
| − | * | + | * to create guidelines for generating regional gravity solu-tions, |
| − | * | + | * to outline standards and conventions for future regional gravity products. |
| − | * | + | * Comparable work outside gravity field determination, e.g. in the mathematical communities and in geomagnetic field determination will be taken into account. |
| − | * | + | * To achieve the objectives, the SG interacts and collabo-rates with other ICCT SGs as well as IAG Commission 2. As a matter of fact the outcomes of the SG can be also used by other IAG commissions, especially in Commission 3. |
| − | + | * The SG's work will be distributed to IAG sister associa-tions through respective members. | |
| − | * | ||
| − | + | ===Program of Activities=== | |
| − | + | The SG’s program of activities will include | |
| + | organization of SG meetings and of one or more scientific workshops on regional modelling | ||
| + | participation in respective symposia (EGU, AGU, etc.), publication of important findings in proper journals, | ||
| + | maintaining a website for general information as well as for internal exchange of data sets and results, | ||
| + | supporting ICCT activities | ||
| − | + | ===Members=== | |
| − | |||
| − | + | '' '''Michael Schmidt (Germany), chair<br />Christian Gerlach (Germany), chair'''<br />Katrin Bentel (Norway)<br />Annette Eicker (Germany)<br />Indridi Einarsson (Denmark)<br />Junyi Guo (USA)<br />Majid Naeimi (Germany)<br />Isabelle Panet (France)<br /> | |
| − | '' ''' | + | Judith Schall (Germany)<br />Uwe Schäfer (Germany)<br />Frederick Simons (USA)<br />C.K. Shum (USA)<br />Matthias Weigelt (Germany) |
| + | <br />Gongyou Wu (China)<br />'' | ||
Revision as of 11:08, 29 June 2012
JSG 0.3: Comparison of current methodologies in regional gravity field modelling
Chairs: M. Schmidt (Germany), Ch. Gerlach (Germany)
Affiliation: Comm. 2, 3
Introduction
Traditionally the gravitational potential of the Earth and other celestial bodies is modelled as a series expansion in terms of spherical harmonics. Although this representation is technically possible for ultra-high expansions, it is well-known that spherical harmonic approaches cannot repre-sent data of heterogeneous density and quality in a proper way. In order to overcome these and other deficiencies regional modelling comes into question. In the last years many groups have developed sophisticated approaches for regional modelling, e.g. the expansion of the gravity field or functionals of the field in terms of spherical (radial) base functions. Analogously to spherical harmonic approaches, also in regional modelling the un-known model parameters, i.e. the coefficients of the series expansion, can be either determined by means of numerical integration or as the solution of a parameter estimation process. Numerical integration techniques are widely used in the mathematical community and provide efficient and stable solutions. However, numerical integration tech-niques suffer from important disadvantages. Among others these methods (1) require the input data to be given on a spherical integration grid, (2) cannot provide estimated error variances and covariances of the model parameters and (3) have difficulties to handle the combination of data from different measurement techniques. Due to these dis-advantages, parameter estimation is the preferred strategy in the geodetic community. Although solutions in regional modelling based on parameter estimation are generated by several groups since many years, a large number of un-solved problems and open questions still remain. They mostly arise from the condition of the normal equation system and are therefore directly connected to the para-metrization of the gravity field, the type and distribution of observation data, the choice and location of base functions, possible regularisation schemes, etc. The aim of the proposed SG is to find guidelines on suit-able strategies for setting up the parameter estimation of regional gravity field modelling. This includes appropriate strategies for the combination of satellite, airborne and terrestrial data. The focus of the SG is on the methodologi-cal foundation of regional gravity field modelling based on series expansions in terms of localizing base functions. Therefore numerical studies will be concentrated on simu-lations based on synthetic data. It is not the aim of the SG to process and compare solutions from real data.
Objectives
The main objectives of this SG are:
- to collect information of available methodologies and strategies for regional modelling, including
- the type of base functions (splines, wavelets, Slepian function, Mascons, etc.),
- the point grids for placing the functions (standard grid, icosaeder, Reuter grid, etc. on a sphere, ellipsoid, etc.),
- the choice and establishment of an appropriate adjust-ment model (combination strategy, variance component estimation, rank deficiency problems, e.g., due to downward continuation, etc.),
- the consideration of model errors (truncation errors, edge effects, leakage, etc.),
- the specific field of application,
- to analyze the collected information in order to find spe-cific properties of the different approaches and to find, why certain strategies have been chosen,
- to create a benchmark data set for comparative numerical studies,
- to carry out numerical comparisons between different solution strategies for estimating the model parameters and to validate the results with other approaches (spheri-cal harmonic models, least-squares collocation, etc.),
- to quantify and interpret the differences of the compari-sons with a focus on detection, explanation and treatment of inconsistencies and possible instabilities of the differ-ent approaches,
- to create guidelines for generating regional gravity solu-tions,
- to outline standards and conventions for future regional gravity products.
- Comparable work outside gravity field determination, e.g. in the mathematical communities and in geomagnetic field determination will be taken into account.
- To achieve the objectives, the SG interacts and collabo-rates with other ICCT SGs as well as IAG Commission 2. As a matter of fact the outcomes of the SG can be also used by other IAG commissions, especially in Commission 3.
- The SG's work will be distributed to IAG sister associa-tions through respective members.
Program of Activities
The SG’s program of activities will include organization of SG meetings and of one or more scientific workshops on regional modelling participation in respective symposia (EGU, AGU, etc.), publication of important findings in proper journals, maintaining a website for general information as well as for internal exchange of data sets and results, supporting ICCT activities
Members
Michael Schmidt (Germany), chair
Christian Gerlach (Germany), chair
Katrin Bentel (Norway)
Annette Eicker (Germany)
Indridi Einarsson (Denmark)
Junyi Guo (USA)
Majid Naeimi (Germany)
Isabelle Panet (France)
Judith Schall (Germany)
Uwe Schäfer (Germany)
Frederick Simons (USA)
C.K. Shum (USA)
Matthias Weigelt (Germany)
Gongyou Wu (China)
