Adapting a regional geoid model to national geodetic infrastructure and conventions
20/09/2016 | 17:30 | Session 3: Local/regional geoid determination methods and models
Author(s): Tõnis Oja, Silja Märdla, Artu Ellmann and Andres Rüdja
Tõnis Oja, Silja Märdla, Artu Ellmann and Andres Rüdja
Within the frames of the Nordic Geodetic Commission (NKG) geoid modelling project NKG2015 five computation centres in the Nordic-Baltic region performed extensive computations of gravity anomaly field and gravimetric geoid models. The geoid computations by these centres were made using different computation methods, strategies and software packages (cf. GGHS2016 abstract submission no. 143 by Ågren et al.). The present contribution focuses on specific issues related to merging the resulting geoid model to existing geodetic infrastructure and national conventions within a participating country. More specifically, the approaches jointly elaborated and results achieved by the Tallinn University of Technology and Estonian Land Board (together referred to as the Estonian computation centre) will be tackled in detail. The computation area included the Nordic and Baltic countries with the surroundings within the geographical limits of 53...74 °N, -2...36 °E. The updated NKG gravity database (with about half a million points) and new NKG digital elevation model (NKG DEM2014) were used to fill the computation area with data in subsequent gravity anomaly field and geoid modelling computations. The database and the DEM included significantly updated, high quality, gravity and height data from gravimetric measurements and airborne laser scanning done in Estonia. Another objective of this contribution is to introduce the different approaches used for gravity field and geoid modelling by the Estonian computation center in the NKG2015 project. The results, including almost 40 concurrent gravimetric geoid models, were evaluated based on the analysis of residuals (model versus observation), differences between the models and other information collected in the modelling process. These models were further evaluated by using the joint GNSS/levelling dataset, which contained accurate geodetic and normal heights (transformed to the common reference frames and reduced to the common epoch 2000.0) from all participating countries. The NKG2015 geoid model has a standard deviation of 3.0 cm from a 1-parameter fit with GNSS/levelling data over the whole Nordic-Baltic region. Using the Estonian GNSS/levelling dataset submitted to the NKG project, the standard deviation of 1.1 cm after a 1-parameter fit was determined. However, the latest adjustment (dated as May 2016) of the national levelling network yielded also updated normal heights for the national GNSS-levelling points. After converting the heights to the common epoch and including the corrections due to the permanent Earth tide, the resulting standard deviation from 1-parameter fit was reduced to 0.8 cm (min, max residuals were -2.1 and +2.4 cm). Despite such a good fit, the geoid models and GNSS-levelling data show clear systematic differences for some areas in Estonia (especially alongside the coastline) which need to be analyzed further.