Given the challenges of increasing urbanization, land use, infrastructure, resource protection, and climate change, the management of underground space has gained significant importance. This is relevant for cities built on unconsolidated sediments like Munich, which serves as a paradigmatic example driving this dissertation. Additional study areas include Starnberg and the Augsburg – Low Lech Valley area. The geological heterogeneity is crucial for subsurface planning. However, the geological conditions are often unknown, with data frequently scarce and underutilized.
The dissertation addresses these challenges by enhancing the predictive capability of unconsolidated sediments by focusing on: i) 3-D prediction of fundamental geological characteristics, ii) integration and quantification of uncertainties, and iii) ensuring multipurpose usability. The findings have been published in four articles in international journals.
3-D geo-models rarely consider the intergranular constituents of unconsolidated sediments, leading to an underestimation of the geological composition. The first publication addresses this in Munich by predicting the fundamental characteristics of these sediments in 3-D, i.e., the grain-size distribution (GSD), using a novel approach called the Di models method. This approach employs sequential stochastic simulation to simulate grain fractions for various cumulative frequencies of the fictive GSD, allowing for a flexible definition of grain-size-based heterogeneities. The method has been successfully applied to Munich, the Starnberg and U9 Tunnels, Augsburg, and the Low Lech Valley.
The input data used in the approach, direct soil descriptions from drilled materials, are often inaccurate, affecting uncertainty quantification (UQ). The second publication addresses this by reinterpreting the soil observations to account for lithological noise. This is done by determining all soil components, quantifying their proportions, and capturing the fictive GSD range. Using entropy and joint entropy for UQ allows realistic outcomes and enables the comparison of model reliability. The uncertainty measures have proven extremely valuable in optimizing the planned U9 Subway Tunnel drilling campaign and analyzing the uncertainty regarding the aquifer interconnectedness in Augsburg.
Geological 3-D models are often purpose-specific and may not fully capture the geological complexity. The third and fourth publications explore the need for multipurpose usability and methods for estimating material properties at the level of facies structures. Combining geological and numerical models with hydrogeochemistry data supports sustainable subsurface planning. GSD patterns define lithological geometries and reservoir architectures. In Munich, the results revealed a gravelly Quaternary aquifer and Miocene aquifers, confirmed by hydrogeochemical data. In Augsburg and the Low Lech Valley, the 3-D geo-models showed a multi-layered aquifer system with vertical lithological zoning. Estimating fictive GSD allows for predicting the hydraulic conductivity. The results were used to calibrate a city-wide numerical groundwater model in Munich and to assess groundwater withdrawal and geothermal potential for groundwater heat pumps. In the Low Lech Valley, a restriction on the drilling depth was implemented to protect the underlying aquifers.
In conclusion, this thesis provides valuable insights into the characterization of 3-D geology in unconsolidated sediments, supporting sustainable spatial planning of the subsurface as a multifunctional resource. The main objective of the dissertation, to improve the 3-D predictive capability of unconsolidated sediments, has been successfully achieved.
«
Given the challenges of increasing urbanization, land use, infrastructure, resource protection, and climate change, the management of underground space has gained significant importance. This is relevant for cities built on unconsolidated sediments like Munich, which serves as a paradigmatic example driving this dissertation. Additional study areas include Starnberg and the Augsburg – Low Lech Valley area. The geological heterogeneity is crucial for subsurface planning. However, the geological c...
»