The use of Terrestrial Laser Scanning in
characterizing active tectonic processes
from postseismic slip to the long term
growth of normal faults

Abstract

This thesis investigates two main hypotheses regarding uncertainty in the measurement of paleoseismic offsets used to estimate fault activity and paleoearthquake magnitudes on normal faults: (1) That variations in fault geometry have a significant effect on throw-rates and fault offsets; and (2) that postseismic deformation can be a significant component of the total fault slip for moderate magnitude earthquakes. These hypotheses are tested using high resolution terrestrial laser scan datasets of normal fault topographic offsets and surface ruptures. The first hypothesis is addressed by studying the crustal scale Campo Felice active normal fault in the Central Apennines, Italy. Variation in throw-rate along strike since the last glacial
maximum (15 ka ±3) is measured from an offset periglacial surface at two hundred and fifty sites using cross sectional data derived from a high resolution terrestrial laser scan (TLS) dataset. The measurements are used to create a detailed throw-rate profile. Field measurements of fault geometry (strike, dip and kinematic slip direction) are also gathered. Variation in fault throw-rate is found to correlate with fault strike. A study of weathered band thickness on the exposed Miocene limestone bedrock fault
scarp, thought to have been created by single past slip events on the fault also appears to correlate with fault strike. A strain-rate profile is calculated using the throw-rate profile and the field measurements of kinematic slip. In contrast to throw-rate, strainrate is independant of changes in fault strike and dip. It is suggested that strain-rate in comparison to throw-rate provides a more robust measure of fault activity as it is unaffected by changes in fault geometry. The outcome of this study is that
paleoseismic studies on active faults should take into account fault geometry before choosing sites which may have anomalously high or low paleoseismic offsets. Fault geometry introduces significant uncertainty into the estimation of inferred paleoearthquake magnitudes from paleoseismic offsets and hence seismic hazard analysis.
The second hypothesis is addressed through the study of near-field postseismic deformation (surface rupture afterslip) following the 6th April 2009 6.3 Mw L’Aquila
earthquake, created by slip on the Paganica normal fault in the Central Italian Apennines. A novel use of TLS technology allowed the postseismic deformation at
four sites along the L’Aquila surface rupture to be measured between 8 – 126 days after the earthquake. Complimentary measurements of postseismic deformation at a
fifth site using a robotic total station were combined with the TLS datasets to describe the along strike variation in postseismic deformation. The near-field postseismic deformation measured occurred mostly in the immediate hangingwall of the surface rupture and increased with decreasing rate over time. The postseismic deformation
measured is comparable to theoretical and empirical models which have been used to describe afterslip for previous earthquakes. The magnitude of near-field postseismic
deformation was up to 60% that of the coseismic offset in the near-field and suggests that postseismic deformation can form a significant component of paleoseismic offsets of moderate magnitude. Postseismic deformation was also found to be greatest above regions of the fault zone where a high coseismic slip gradient existed, suggesting that postseismic deformation occurs at the periphery of the coseismic slip patch within the fault zone. Regression relationships which relate surface offset to moment magnitude are populated by field observations of surface offsets where earthquake magnitude is known. These regression relationships are then used to infer paleoearthquake magnitudes from paleoseismic offsets. The field studies used to populate regression relationships do not routinely take into account the potential effects of fault geometry and significant postseismic slip. As a result paleoearthquake magnitudes inferred from such regression relationships are maybe over estimated. It is suggested that future regression relationships of surface offset and moment magnitude should factor in the effects of fault geometry and postseismic deformation in order to produce a relationship in which surface offset (both coseismic and postseismic) is described for a range of magnitudes and, where possible, any local effects of fault geometry are removed from the input dataset. The production of such a relationship will allow paleoseismologists to measure combined coseismic and postseismic offsets from field studies and to infer paleoearthquake magnitude with decreased uncertainty.