Regional quasi-geoid modelling beyond remove-compute-restore
20/09/2016 | 16:45 | Session 3: Local/regional geoid determination methods and models
Author(s): Roland Klees, Cornelis Slobbe and Hassan Hashemi Farahani
Roland Klees, Cornelis Slobbe and Hassan Hashemi Farahani
The stochastic models of data used in regional quasi-geoid modelling have improved significantly over the last years. Today, noise covariance matrices are available for all datasets including latest GRACE/GOCE-based global gravity field models (GGMs), and data from radar altimetry, and terrestrial and airborne gravimetry. Missing variance/covariances can be estimated using variance component estimation, and long-wavelength errors in terrestrial gravity datasets can be parameterized. At the same time, there is an increasing interest of users in information about the accuracy of estimated height anomalies for quality control and further data processing, which goes beyond a validation using GPS/levelling data.
Here we present an overall methodology for regional quasi-geoid modelling using least-squares, which goes beyond the traditional remove-compute-restore procedure. We identify the main challenges of such an approach. Based on extended numerical experiments, we provide answers to the most important research questions related to among others the choice of a suitable parameterization of the regional gravity field (single scale versus multi-scale), the choice of the functional model of the GGM dataset, and the strategy to deal with the different spectral bandwidth of the various datasets.