Generating consistent triangular terrain elevation data for noise modelling

Generating consistent triangular terrain elevation data for noise modelling: Master thesis Geomatics

van Rijssel, Laurens (TU Delft Architecture and the Built Environment)

Dukai, B. (mentor)
Stoter, J.E. (graduation committee)
Kok, Arnaud (graduation committee)

Delft University of Technology

Geomatics

2022-01-13

In 2016, a collaboration between the TU Delft, RIVM and a few other companies started to automate the generation of environmental input data for noise modelling.
As a part of that cooperation a promising triangle based data structure is researched to represent terrain elevation.
Due to the properties of its generation it can provide certainty over the presence of objects while the format allows for efficient cross sections extraction.
Yet a persistent challenge remains, the large file size makes exchange and processing of terrain data troublesome.
This research aims to optimize the terrain elevation data to decrease the file size while ensuring accurate noise calculations.

Using an in-depth analysis of the usage of terrain elevation in noise modelling in a Dutch and European context, a environmental based approach for a triangulated terrain generation is developed.
This approach is conceptually implemented to facilitate the calculation of noise values using terrain generated with the provided approach.
Empirical tests have been conducted to measure the influence of different terrain elevation settings over a variety of terrains.

The study concludes by providing an advice on an approach for elevation model generation.
Several insights are discovered regarding the impact of terrain elevation accuracy on noise propagation calculations.
Compared to a currently publicly available terrain data set a size reduction of 50 to 90% can be obtained without quality reduction.

Geomatics
Noise modelling
TIN
Triangulation
terrain modelling
noise propagation

http://resolver.tudelft.nl/uuid:f77cd8e1-682b-4812-9652-25b23b5319a4

Student theses

master thesis

© 2022 Laurens van Rijssel