Sam McColl - PhD Summary
Deglaciation and related events, such as glacier lake outburst floods and landslides, have resulted in 30,000 deaths and caused billions of dollars of damage around the globe (Evans and Clague, 1992; at the time of publication). Projected climate changes indicate a general increase in the potential for many geohazards (McGuire, 2010). Thus there is a pressing need for better estimation of the remaining potential for geohazards impacting human activities arising from climatic changes in mountainous areas.
It has long been recognised that glaciation, deglaciation and deglacierization can adversely affect slope stability in mountainous terrain. Much research in the literature to date has focussed on establishing the temporal and spatial relationships between deglaciation and slope failures. Less research has focussed on the mechanisms linking these two phenomena, and there have been too few studies on the response of rock slopes to contemporary glacier retreat. This has resulted in many postulations about cause and little substantive evidence (McColl et al., 2010). My research attempts to help reconcile this by bringing together an understanding of the physical properties of glaciers, rock mass properties, rock slope deformation/failure processes, and climatic and non-climatic variables. The project addresses five areas:- Slope instabilities and glacier ice masses. Can glaciers accommodate rock slope deformation, i.e. can slopes fail prior to glacier retreat? To answer the question I am searching the Southern Alps of New Zealand for examples of rockslides deforming into glaciers – so far I have found several! To better understand the process I am using mathematical models, as well as monitoring an active landslide in a glaciated valley (McColl & Davies, in prep).
- Seismicity and post-glacial rock slope failure. In North America and Europe researchers have linked periods of landslide activity to seismicity enhanced by isostatic rebound of ice sheets. In New Zealand, many large pre-historic landslides, including the giant (27 km3) Green Lake Landslide seem to be late-glacial to early Holocene in age and due to their large size are likely to have been triggered by large earthquakes. Along with others in the department I am searching through existing paleoseismic records and searching for other evidence of enhanced seismic activity during the late-glacial to Holocene period in NZ. Is it possible that isostatic rebound from mountain glaciers can significantly influence seismicity and landslide patterns in an active plate boundary setting?
- Glaciers and seismic topographic amplification. As well as influencing seismicity, glaciers may also influence earthquake shaking intensity. The erosion by glaciers results in steep, high relief topography, which is a recipe for intense coseismic ground motions on hilltops, a phenomenon called topographic amplification. Until glaciers retreat however, the occupation of valleys and the inundation rock masses by ice may lessen the shaking that the rock would otherwise experience during an earthquake. To test this notion I have used numerical modelling software (FLAC 6.0) and have shown that indeed there is likely to be a significant increase in shaking intensities as a result of deglacierization (McColl et al., 2012).
- Post-glacial rock mass-degradation. Many landslides that have occurred in deglaciated valleys have failed on surface-parallel stress-release joints. Stress-redistributions causing the development of stress-release joints are considered to arise from glacial erosion, tectonics or glacial unloading. I am using numerical models to investigate the role of stress-release joints and the conditions that may have lead to the development of these joints in the 13 M m3 Young River landslide of Mount Aspiring National Park that catastrophically failed in 2007.
-
Reliability of landslide and glacier distribution data. In linking slope failure distributions to deglaciation patterns one has to assume that the data provides an accurate representation of the deglaciation and slope-failure history. However, I along with others have recognised several problems with these data. For example, landslides occurring in the early post-glacial period are less likely to be represented in the records because there is less chance of preservation. Another problem is that glacial and landslide deposits are often mis-identified. As part of my project I have investigated one such deposit that was originally considered to be recessional glacier moraine, but which I now reinterpret as being an early Holocene landslide deposit (McColl and Davies, 2011).