High Frequency Water Level Responses to Natural Signals

Abstract

To date, there are few examples of the ambient monitoring of water levels from multiple nearby wells collected at high frequency (one second or less). High frequency monitoring is typically reserved for short term hydraulic tests, however, the natural groundwater system is constantly in flux leading to a multitude of changes suitable for analysis. This thesis focuses on data analysis methods and characteristics of water levels collected at high frequency from 48 wells at a fractured rock site. By using high frequency monitoring and focusing on natural signals with known inputs, thousands of detectable responses were collected and analyzed during this thesis. Each of these responses can be considered a naturally induced hydraulic test useful to better understand the subsurface.

This thesis is arranged as three papers suitable for future publication using datasets collected with high precision and high frequency. The first paper introduces an efficient time-domain method for calculating barometric response functions on large datasets. The method overcomes computational limitations associated with high frequency data and extends the possible range of transmissivity estimates from barometric/loading response functions. The second paper applies time and frequency domain methods to examine the spatial and temporal aspects of water level responses to Earth tides and barometric pressure changes for the purpose of parameter estimation and understanding subsurface mechanisms. The third paper evaluates water level responses to regional and teleseismic earthquakes. This thesis highlights the benefits of using high frequency monitoring and provides software tools to improve the analysis of water level responses to natural signals for researchers and practitioners.