Long-term monitoring of permafrost-affected rock walls

Subsurface monitoring of permafrost conditions at depths up to 20-30 m is crucial to assess the safety and reliability of mountain infrastructure, because permafrost degradation critically affects rock slope stability in high mountains. Thus, developing methods to provide information on thermal and hydrostatic subsurface properties is essential, especially as boreholes provide one-dimensional (1D) thermal data in a complex 3D terrain and usually cannot be installed in actively unstable rock masses. Electrical resistivity tomography (ERT) has been proven to be a straightforward monitoring tool for near surface bedrock permafrost for monthly or longer intervals. But as rockfalls are often prepared over periods of days or hours, ERT for early warning purposes should also detect short-term triggering events such as pressurised water flow. During the course of this Ph.D.- project, an open air laboratory (OpAL) for long-term monitoring of permafrost and mass movements was established in the summit region of the Kitzsteinhorn (3,203 m a.s.l.) and is now the best instrumented long-term monitoring site for permafrost and mass movements in Austria. The systemic and scale-oriented approach includes a combination of automated weather stations, remote sensing techniques, temperature measurements in shallow and deep boreholes, geophysical and geotechnical investigations to identify potentially critical thresholds for rock slope instabilities. The OpAL was developed and instrumented to provide answers to the following main research questions: (i) Is an automatic electrical resistivity tomography (AERT) capable to continuously monitor the state of frozen, steep permafrost affected bedrock? (ii) On which temporal and spatial scales can AERT monitor stability-relevant hydrostatic and thermal changes? (iii) Do iButtons provide sufficient accuracy to perform spatially and temporally highly resolved near-surface rock temperature (NSRT) measurements in complexly structured permafrost-affected rock walls? (iv) Is it possible to observe a temperature-resistivity (T-ρ) relationship under field conditions, what are the characteristics and are these T-ρ gradients comparable to laboratory results? To provide stability-relevant hydrostatic and thermal information with ERT, the first permanently installed AERT system was established on a steep, unstable permafrost rock wall to monitor subsurface electrical resistivity changes continuously. ERT is measured every 4h at the up to 67° steep rock wall below the Kitzsteinhorn cable car. Wenner datasets (n=996) were analysed from February 2013 to February 2014 in terms of data stability, raw data characteristics and measurement errors coinciding with potential disturbing factors. Strong resistivity changes coincided with rapid freezing or water inundating rock fractures. Automatically detected periods with large resistivity changes produce ERT time series with low resistivities extending from the bottom upwards during times of snowmelt. Fracture inventories, visual observations of cleftwater and NSRT measurements, provide indications that the flow of pressurised water in fractures warms the surrounding rock in an upward direction. Complementary to the AERT measurements, the spatial and temporal variations of NSRT are measured in a hitherto unknown coverage using a newly developed low-cost alternative based on miniature temperature loggers (iButtons). This includes an innovative adaptation procedure including preparation, installation and maintenance of the system. To meet the requirements for spatially and temporally distributed rock temperature measurements, the measurements were analysed regarding reliability and accuracy. The combination of AERT and NSRT measurements leads to the first comprehensive field evidence of a T-ρ relationship. The T-ρ relationship can be used as a proxy for the thermal state and in further consequence for stability-relevant information on the mechanical state of permafrost rock walls. Based on 1485 Wenner and rock temperature datasets, the T-ρ characteristics were analysed considering different depths of investigation. Linear regression modelling shows that analysed T-ρ relationships are highly significant with a p-value < 2.2e-16 and a R² of up to 0.64. The measured T-ρ gradients of 22.9 to 27.9 %/°C under permafrost conditions, are in the range of the laboratory results of 29.8±10.6 %/°C (Krautblatter, 2009). This Ph.D.-thesis presents (i) the first approach to monitor permafrost rock walls quasi-continuously with AERT where high apparent resistivity changes in 4h intervals may precede critical hydrostatic events confined by permafrost rocks including (ii) a newly developed strategy for spatially and temporally high resolution NSRT measurements. The combined methodological approach (AERT and NSRT measurements) results (iii) in the first comprehensive field evidence of the T-ρ relationship which has been postulated in the laboratory a few years ago. OpAL instrumentation and field monitoring techniques developed, tested and analysed in this PhD provide a framework that can be used for early warning systems in unstable permafrost-affected rock walls.