09:15 〜 09:30
[G02-1-04] MRR and LSC – A mutual benefit for advanced regional gravity field modeling
Various regional gravity field modeling approaches exist next to each other – everyone revealing individual strengths and weaknesses depending on the methods themselves and on the fields of application. For instance, in order to establish precise height systems in developing countries with sparse data availability and/or low-quality, Multi-Resolution Representation (MRR) using spherical basis functions and Least Squares Collocation (LSC) are suitable approaches. Drawing mutual benefit from both is the innovative key aspect of our study and enables an advanced modeling of the gravity field in the specific regions by optimally combining heterogeneous data sets.
The common theoretical base of MRR and LSC is that in- and output functionals are described by series expansion in terms of Legendre Polynomials. For formulating observation equations, LSC incorporates full covariance information of the observations, and thus, provides coherent error estimates of the resulting models; MRR manages a well-balanced relative weighting of heterogeneous input data by variance component estimation, and a consistent spectral combination by appropriate filtering.
We study in detail the differences of LSC and MRR, learn from each other and finally bring together the strengths of both approaches: by (1) matching the mathematical fundamentals, (2) analyzing closed-loop scenarios with simulated data, and finally (3) setting up an optimal regional modeling strategy. We focus on a study area in South America with different topographic features and combine heterogeneous data sets obtained in varying measurement heights and with different quality, spectral and spatial resolutions. From the findings of these comprehensive studies, we plan to derive a generalized concept which provides developing countries the opportunity of establishing physical height systems. Those have a high socio-economic impact for various applications in science, infrastructure and administration.
The common theoretical base of MRR and LSC is that in- and output functionals are described by series expansion in terms of Legendre Polynomials. For formulating observation equations, LSC incorporates full covariance information of the observations, and thus, provides coherent error estimates of the resulting models; MRR manages a well-balanced relative weighting of heterogeneous input data by variance component estimation, and a consistent spectral combination by appropriate filtering.
We study in detail the differences of LSC and MRR, learn from each other and finally bring together the strengths of both approaches: by (1) matching the mathematical fundamentals, (2) analyzing closed-loop scenarios with simulated data, and finally (3) setting up an optimal regional modeling strategy. We focus on a study area in South America with different topographic features and combine heterogeneous data sets obtained in varying measurement heights and with different quality, spectral and spatial resolutions. From the findings of these comprehensive studies, we plan to derive a generalized concept which provides developing countries the opportunity of establishing physical height systems. Those have a high socio-economic impact for various applications in science, infrastructure and administration.