2.2 Subduction interface sections
We assign fault sections to broadly characterize earthquake recurrence and slip deficit rates along strike of the AASZ (Figure 1). Our selection of the term ’section’ rather than ’segment’ is intentional, as our analysis is not meant to imply that each section is a fixed rupture segment. Previous segmentation models of the AASZ focused primarily on the historical pattern of ruptures, mainly inferred from aftershock zones (Davies et al., 1981; McCann et al., 1979; Sykes, 1971) (Figure 2). Here we define fault sections in part by historical and prehistoric rupture patches, but also geodetic data, geologic observations of land-level changes and tsunami recurrence, and structural observations. The sections are modified from those presented in the 2007 NSHM (Wesson et al., 2007) and many closely follow those defined by a coordinated effort by the National Tsunami Hazard Mitigation Program to devise tsunami source models for Alaska (Ross et al., 2023). The sections vary in length, ranging from from ~95 km (Barren Islands section) to ~540 km (Komandorski section) (Figure 1).
We defined sections based on observed along strike changes in criteria: changes in slip magnitude and/or depth range of slip in historical great earthquakes, changes in the interseismic slip deficit distribution (rate or spatial pattern) based on geodetic studies, or changes in earthquake recurrence estimated from paleo-earthquake or paleo-tsunami studies. We required along-strike changes in at least two of these quantities to define a section boundary. For example, the Barren Islands section, the shortest in our model, was defined (see Section 3 for details) based on a significant narrowing of the region of slip deficit observed in geodetic models (Suito & Freymueller, 2009; 2020), plus a corresponding narrowing of the region of slip in 1964 in the model of Ichinose et al. (2007).
The hazard modeling approach that will use our model values does not allow for variations in slip deficit with depth, but only updip and downdip limits of the seismogenic zone and a slip deficit rate/moment accumulation rate. The downdip limit is estimated from slip in known great earthquakes or from geodetic estimates. The updip limit is difficult to estimate because geodetic studies have very limited model resolution near the trench, and usually no more than one large or great earthquake has a known slip distribution. Even where earthquake slip distributions have been modeled, the extent of shallow slip may be poorly constrained without near-trench observations (e.g., Brooks et al., 2023). The tsunami record can demonstrate evidence of past slip to the trench (or near it), but if we lack clear evidence for the slip behavior of the megathrust at shallow depth, or if previous coupling models lack clear updip limits, we assume that the updip limit of slip deficit extends to the trench.