High-resolution record of displacement accumulation on an active normal fault: implications for models of slip accumulation during repeated earthquakes

The spatial and temporal accumulation of slip from multiple earthquake cycles on active faults is poorly understood. Here, we describe a methodology that can determine the time period of observation necessary to reliably constrain fault behaviour, using a high-resolution long-timescale (the last 17 kyr) fault displacement dataset over the Rangitaiki Fault (Whakatane Graben, New Zealand). The fault linked at ca. 300 ka BP and analysis of time periods within the last 17 kyr gives insight into steady-state behaviour for time intervals as short as ca. 2 kyr. The maximum displacement rate observed on the Rangitaiki Fault is 3.6 +/- 1.1 mm yr(-1) measured over 17 kyr. Displacement profiles of the last 9 ka of fault movement are similar to profiles showing the last 300 ka of fault movement. In contrast, profiles determined for short time intervals (2-3 kyr) are highly irregular and show points of zero displacement on the larger segments. This indicates temporal and spatial variability in incremental displacement associated with surface-rupturing slip events. There is spatial variability in slip rates along fault segments, with minima at locations of fault interaction or where fault linkage has occurred in the past. This evidence suggests that some earthquakes appear to have been confined to specific segments, whereas larger composite ruptures have involved the entire fault. The short-term variability in fault behaviour suggests that fault activity rates inferred from geodetic surveys or surface ruptures from a single earthquake may not adequately represent the longer-term activity nor reflect its future behaviour. Different magnitude events may occur along the same fault segment, with asperities preventing whole segment rupture for smaller magnitude events. (c) 2006 Elsevier Ltd. All rights reserved.