NAP to EVRS

NAP to EVRS: a re-evaluation of the Dutch input and its impact on the realization of the European Vertical Reference System

Speth, K.A.J.

Klees, R. (mentor)

Civil Engineering and Geosciences

Geoscience and Remote Sensing

Applied Geophysics and Civil Engineering

2016-10-27

Double degree of Master of Science in Applied Geophysics at Delft University of Technology, ETH Zürich and RWTH Aachen University. - In 2004, Rijkswaterstaat provided geopotential differences between the benchmarks of the Normaal Amsterdams Peil (NAP) network to the Bundesamt für Kartographie und Geodäsie (BKG) for the realization of the European Vertical Reference System (EVRS). This data was found to be incorrect, resulting in inaccurate transformations between EVRF2007 and the national height systems of the participating countries. To correct the realization of the EVRS and obtain accurate transformations between the height systems, it is essential that the BKG is provided with correct geopotential information on the NAP network. In this thesis, a computational procedure has been developed to obtain correct geopotential differences. The final procedure was implemented in a MATLAB software package that was provided to Rijkswaterstaat for future computations. The old results were compared with the newly computed results, in order to obtain insight into the origin of the errors made in 2004. The impact of these errors on realizations of the EVRS was examined as well. Finally, the connection of the NAP levelling network to the neighbouring countries, and thus the Unified European Levelling Network, (UELN) was investigated. Due to the available data, most steps in the computational procedure were already predefined. Therefore, the main development was a method for predicting gravity at the levelling benchmarks. Observed gravity was reduced to surface gravity anomalies to remove height dependency. This makes accurate 2-D interpolation possible. Four methods were tested for the interpolation of these gravity anomalies: ordinary Kriging, least-squares collocation, cubic spline and biharmonic spline. Biharmonic spline interpolation was selected as the gravity prediction method for the final procedure. For potentially further improvement of the gravity prediction, two gravity corrections were investigated as well: residual terrain modelling and correcting for a global gravity field model. Both corrections were found to cause negligible improvement to the gravity prediction with respect to the uncertainty of the levelling observations. Therefore, these corrections were not used in the final procedure. Large regional differences were observed between the newly obtained results and those of Rijkswaterstaat. These differences could predominantly be explained by a mistake in sign convention of the surface gravity anomalies when restoring the observed gravity. In an updated realization of the EVRS, using the new NAP geopotential information, a tilt of several millimetres within the Netherlands and Belgium is observed. When updated data from other countries are also considered, datum points heights varied in the centimetre range. Finally, there are a total of 28 Dutch-German cross-border observations at 13 locations, evenly distributed along the border. This lead to the conclusion that the connection between the NAP and the German levelling network is strong. The connection of the NAP with the Belgian network is much weaker, with only a single observation currently known at the BKG. Although, from the data used here it seems that more connections should be readily available.

NAP
EVRS
EVRF2007
Geopotential
Gravity Interpolation
RTM

http://resolver.tudelft.nl/uuid:93703012-035f-4c28-918b-132831256fad

Student theses

master thesis

(c) 2016 Speth, K.A.J.