Strain accumulation and episodicity of fault movements in Otago

Executive Summary

We present the results of a pilot study of episodicity of fault movement among a number of parallel reverse faults in Central and East Otago. South of Jackson Bay, the Alpine Fault exhibits very low dip-slip values, and active deformation east of the fault suggests a high proportion of the orthogonal convergence is partitioned onto structures across Otago. Previous investigations of the Pisa-Grandview and Dunstan Faults in Central Otago, and the Titri and Akatore Faults in coastal Otago produced evidence suggesting periods of enhanced activity followed by long periods of quiescence, with a possible temporal switching of activity between structures. This study places constraints on the timing of the most recent activity on intervening structures by relating surface deformation to datable deposits. In addition, the current rate of regional strain accumulation is established enabling constraint of the total amount of displacement required within the region, and a comparison with geologically determined longer-term rates. This project involved three principal avenues of investigation:

• Obtaining reconnaissance observations of tectonic geomorphology along the southeast range fronts between the Dunstan Range and the east coast and producing more detailed studies of a few selected sites on individual range fronts.
• Obtaining OSL ages on critical horizons that can place age constraints on the timing of the most recent surface deformation associated with rangefront faulting.
• Occupation of two high-precision GPS monitoring sites and comparison with Dunedin continuous base station to establish current strain rate and distribution across the region.

All investigations were carried out successfully and a wealth of new data was obtained which has some interesting implications for seismic hazard assessment and the mechanics of deformation within Otago. The results from the three lines of investigation may be summarised as follows:

• All rangefronts exhibit distributed deformation with no continuous range-scale surface ruptures evident. On a large scale, rangefront deformation is revealed by lineaments formed of stream course deviations and slope breaks. Small-offset range-parallel faulting and folding found both within the ranges and within basin sediments are the norm. These observations are consistent with a model of fold propagation above buried reverse faults. In map view, some of the rangefronts are distinctly sigmoidal and their departure from linearity appears to be structurally controlled. The inflection points are commonly the sites of anomalously-striking stream incision and along-strike extrapolation of known fault segments. Many of the ranges appear to have formed by the amalgamation of en-echelon segments rather than growth on a single continuous structure.
• The Holocene ages of all but one of the sites within deformed rangefront sediments implies the current level of seismic activity in Otago is perhaps higher than previously perceived. This perception is no doubt due in part to the subtle nature of the distributed deformation and lack of prominent continuous fault scarps, combined with a low historical seismicity.

The amalgamation of OSL ages, geomorphologic and geological data has enabled tentative estimation of uplift rates of some of the ranges. If it is assumed that the faults dip at around 45˚, then heave equals throw, or uplift rate equals shortening rate. These figures are:

1. Taieri Ridge – 0.6 ± 0.04 mm/yr averaged over 12.9 ± 1.1ka
2. Rock and Pillar Range – 0.15 ± 0.01 mm/yr averaged over 8.42 ± 0.83ka
3. Rough Ridge – 0.18 mm/yr to 0.44 mm/yr averaged over 2.5-1.0Ma (minimum-maximum limits); 0.86 mm/yr averaged over 130ka; and 0.27 mm/yr averaged over 4ka. These figures for Rough Ridge are comparable with rates obtained in a companion study using 10Be isotopes within sarsen stones (quartzite bounders) to date the timing of peneplain exposure on South Rough Ridge (Jackson et al, 2002).
4. GPS data indicate a contraction rate of 1.58 ± 0.18 mm/yr between Dunedin and Hyde Rock, and 2.2 ± 0.88 mm/yr between Dunedin and Makarora. Both estimates are subject to uncertainty. These are linear contraction rates which are roughly perpendicular to the strike of the faults in this study, so comparison with shortening determined by dip-slip on these faults is justified. The rate between Dunedin and Makarora is consistent with estimates of long-term contraction across the region based on fault displacement data, but the Dunedin-Hyde Rock rate is faster than might have previously been expected. However, the GPS-derived short term contraction rate across east Otago is compatible with the range-front deformation data and is consistent with the conclusion that east Otago is currently an area of increased activity.

The new data presented raise the likelihood for the contribution of structures other than the Dunstan and Akatore Faults to seismic hazard in Otago. Clearly, Holocene activity has occurred on other structures across the province, which offer potential for sizeable earthquakes. Assuming a reasonably constant contraction rate of 1.5 mm/yr between Hyde Rock and Dunedin, at least 15m of shortening is required within the region for the Holocene. Documented offsets on the Dunstan and Akatore faults can account for almost half of this requirement and, as this study shows, at least some of the balance has resulted in activity on intervening structures. Currently therefore, all faults in east Otago except for the Titri Fault must be viewed as potentially active.

Whereas there are faults which exhibit long periods of quiescence strongly indicating episodic behaviour, the new data presented here show that several faults in east Otago have been active simultaneously during the Holocene and that a simple model of episodicity is unlikely. More detailed work is required on each of these structures to determine their individual seismic history. OSL dating techniques have proved invaluable in placing age constraints on deformation within materials traditionally difficult to date accurately.