SEEDS INTERNSHIP: Hydroseismology: Monitoring Groundwater Changes Using Railway Quakes
This internship project investigates how repetitive anthropogenic seismic sources, specifically passing trains, can be used to monitor near-surface hydrological processes. Trains act as highly repeatable seismic sources that traverse the same locations many times per day, providing a unique opportunity to regularly probe shallow subsurface properties. The project will focus on identifying temporal variations in seismic attenuation associated with changes in groundwater level, soil moisture, and rainfall. Using a single-station spectral analysis approach, the intern will process continuous seismic data to detect and catalogue repeatable train-generated signals, compute event-based power spectra, and quantify attenuation changes from the frequency-dependent decay of spectral amplitudes relative to a reference state. These seismic observations will be compared with independent environmental datasets, such as rainfall and river-level records, to evaluate the sensitivity of attenuation to hydrological processes occurring within the upper tens of metres of the subsurface. A key motivation is the remote monitoring of soil moisture, which is critical for assessing the stability of railway lineside slopes. The intern will gain hands-on experience deploying seismometers near high-speed railway lines in Berkshire (west of London) and develop practical skills in passive seismic analysis, frequency-domain signal processing, and hydro-seismic interpretation using Python-based workflows. The project provides a foundation for extending attenuation-based monitoring to other sites, noise sources, and dense sensor networks, with relevance to groundwater monitoring, environmental geophysics, and geotechnical applications.
Internship project work plan: Full-time
Week 1 – Project induction and background
The intern will be introduced to the scientific background of hydroseismology and anthropogenic noise sources, with a focus on railway-generated seismic signals. They will receive training in seismic instrumentation, data formats, and Python-based analysis tools, and review relevant literature on seismic attenuation and hydrological monitoring.
Week 2 – Field deployment and data familiarisation
The intern will assist with deploying one or more seismometers near active railway lines in Berkshire and learn best practices for field logistics and data quality control. They will begin exploring continuous seismic datasets, visualising recordings, and identifying characteristic train-generated signals.
Week 3 – Signal detection and event cataloguing
The intern will develop and apply methods to detect repeatable train passages in continuous data and build an event catalogue. They will extract event windows, compute basic power spectra, and assess repeatability and signal stability through time.
Week 4 – Spectral analysis and attenuation estimation
The intern will implement single-station spectral analysis techniques to quantify frequency-dependent amplitude decay and estimate temporal changes in seismic attenuation. They will explore sensitivity to processing choices and establish a reference state for comparison.
Week 5 – Environmental data integration and interpretation
The intern will compile relevant environmental datasets (e.g. rainfall, river level, groundwater indicators) and compare them with observed attenuation variations. They will interpret results in terms of near-surface hydrological processes, including soil moisture and groundwater fluctuations.
Week 6-7 – Synthesis, validation, and communication
The intern will refine analyses, evaluate uncertainties, and assess the robustness of observed relationships. They will synthesise findings into figures and a short written report or presentation, discussing implications for railway slope stability and future monitoring applications.
Internship project work plan: Part-time
Week 1 (Part-time) – Introduction and data familiarisation The student will be introduced to the project aims, background literature, and available seismic and environmental datasets. They will set up the Python-based analysis environment, explore continuous seismic recordings, and learn to recognise characteristic train-generated signals through visual inspection and basic signal processing. Week 2 (Part-time) – Event detection and spectral analysis The student will identify and extract repeatable train-passage events from the seismic data and compute event-based power spectra. They will begin tracking temporal changes in spectral amplitudes and assess signal repeatability and data quality. Week 3 (Part-time) – Attenuation estimation and environmental comparison The student will estimate relative changes in seismic attenuation using frequency-dependent amplitude decay and compare these results with simple environmental indicators, such as rainfall time series. They will interpret observed patterns in terms of shallow hydrological processes. Final Half-Week – Synthesis and reporting The student will consolidate results, generate summary figures, and produce a short written report describing the methods, findings, limitations, and potential extensions of the work.
