Hydrological signals in borehole strain data

Identifying and modelling hydrological signals in borehole strain data

Borehole strainmeters observe time-dependent deformation just below the surface of the Earth with remarkable precision.  They let us probe sub-Angstrom changes in borehole diameter.  And those changes in diameter can be created by a range of processes: by slip on faults at depth, by the downward force of river water on the Earth’s surface, and by increases in fluid pressure at depth created by rapid rainfall.  In this project, you will identify and model some of the hydrological signals: the results of rivers, rainfall, irrigation, and groundwater flow.   

Hydrology-induced strainmeter signals have been largely unexamined by the geodetic community simply because the instruments were installed to focus on tectonic signals: on small earthquakes and slow fault slip.  But the hydrological signals are often the largest signals in the data, and they have much to teach us as well.  We may, for instance, observe how groundwater flows to 100-m depths and identify changes in that fluid flow following an earthquake.  Further, if we can successfully identify and model hydrological signals, we will be able to remove them from the data and more precisely probe the tectonic signals.   

The challenge in observing and modelling hydrologic and tectonic strainmeter signals is that they all happen at the same time.  You will separate the signals by (1) visually identifying some large hydrological signals, (2) using a random forest approach to model the observed strain as a function of local hydrological and weather data, and (3) examining features of the model that could reflect physical processes.  Depending on time and interest, you may focus on one task more than the others or probe other physical models or consider more physical models of groundwater flow. 

Number of students: 1

If interested in this project, please contact jessica.hawthorne@earth.ox.ac.uk