Search

A new reference attenuation model for calculating local magnitude (ML) for New Zealand earthquakes is derived. An earlier reference model, denoted NZ16, was developed in 2016 as a possible alternative to the classical reference model used by GeoNet for earthquakes located by the SeisComP processing system since January 2012. It aimed to provide a logA0 attenuation relation of the amplitude with the hypocentral distance applicable to the New Zealand region and to make ML more consistent with moment magnitude (Mw). For the present update, denoted NZ20, a much larger set of regional moment tensor solutions and corresponding station amplitudes is available. Residual analysis is used to screen the individual station amplitude readings and then to exclude observations from periods of time when the station residuals were consistently anomalous. Effects of other possible sources of bias on the attenuation relation are also examined, including unreliable locations of earthquakes with large azimuthal gaps and the saturation of ML in large earthquakes. An extra explanatory variable is added to the attenuation relation to account for dependence on hypocentral depth, so the updated local magnitude is consistent with Mw across the full range of depths at which earthquakes occur in New Zealand. Furthermore, to counteract possible sources of bias, the data set for analysis is restricted. Earthquakes above the approximate amplitude saturation threshold of Mw 6.6, as well as those with azimuthal gaps greater than 270°, are excluded.

Few studies have been conducted on the variance of scattering (![Graphic][1] ) and intrinsic attenuation (![Graphic][2] ) with depth in the crust. Furthermore, these values, routinely estimated using multiple lapse time window analysis (MLTWA), were seriously affected by different earthquake depths. Based on earthquakes recorded by dense seismic networks in South Korea, we used the MLTWA method on single or small clusters of events with similar depths. The depths of 54 earthquakes, relocated using seven velocity models, provide a typical seismogenic layer with the brittle–plastic transition (BPT) depth at 11 km, which correlated well with the seismicity of Korea. The depth was also used for the theoretical model based on the direct‐simulation Monte Carlo method. As the result of MLTWA, events with shallow focal depths exhibit higher ![Graphic][3] correlated with dense scatter, whereas those with deeper focal depths exhibit lower ![Graphic][4] . This depth dependency of the ![Graphic][5] value is indicated by the significant difference between the upper and lower parts of the BPT depth in the seismogenic layer. We also found that ![Graphic][6] values do not show depth dependency, whereas ![Graphic][7] values do. Thus, this study demonstrates for the first time that the crustal depth dependency of total (![Graphic][8] ) attenuation originates from ![Graphic][9] rather than from ![Graphic][10] . The approach using single or few events may be helpful for a regional study of ![Graphic][11] and ![Graphic][12] in seismically inactive regions for which insufficient data exist and may provide insight into the seismogenic layer related to potential seismic hazards. Online Material: Tables of residuals and range of multiple lapse time window analysis (MLTWA) parameters and Q −1 values for studied events. [1]: /embed/inline-graphic-1.gif [2]: /embed/inline-graphic-2.gif [3]: /embed/inline-graphic-3.gif [4]: /embed/inline-graphic-4.gif [5]: /embed/inline-graphic-5.gif [6]: /embed/inline-graphic-6.gif [7]: /embed/inline-graphic-7.gif [8]: /embed/inline-graphic-8.gif [9]: /embed/inline-graphic-9.gif [10]: /embed/inline-graphic-10.gif [11]: /embed/inline-graphic-11.gif [12]: /embed/inline-graphic-12.gif

New Zealand is one of the more seismically active countries in the world with over 15,000 earthquakes each year in New Zealand and the adjacent offshore region. Routine regional moment tensor (RMT) analysis was implemented by GeoNet in 2007, and nearly 1300 RMT solutions have been calculated for the New Zealand region from August 2003 through early September 2012. The New Zealand RMT catalog contains events with moment magnitude ( M w ) 3.2–7.3 and centroid depth 2–375 km, and includes RMT solutions from aftershock sequences for two major earthquakes: 15 July 2009 M w 7.8 Dusky Sound and 3 September 2010 M w 7.1 Darfield. The RMT solutions provide important constraints on the active tectonics of New Zealand including stress and strain, and the RMT catalog, along with 155 Global Centroid Moment Tensor Project solutions, is used to examine the relationship between M w and local magnitude ( M L ). For shallow focus (≤33 km depth) the M L / M w relationship is found to be similar across New Zealand except for the Fiordland region. For all of New Zealand except Fiordland, M L =(0.93±0.06) M w +(0.54±0.24), and for Fiordland, M L =(0.83±0.04) M w +(1.09±0.13). For deep‐focus earthquakes (>33 km depth) M L is consistently larger than M w by more than a full magnitude unit for some events. The discrepancy between M L and M w is found to be depth dependent with ( M L − M w )=−0.94exp(− h /75.37)+0.80, where h is focal depth in km.

The mechanisms that permit earthquake rupture at the high pressures occurring below 100 km depth remain enigmatic and controversial. The most distinct difference between deep and shallow earthquakes is a scarcity of aftershocks below 100 km. We report strong changes in aftershock productivity with depth. From 100 to 300 km, aftershock productivity normalized to account for the effect of main- shock magnitude is more than an order of magnitude lower than that of shallow events. Between 300 and 550 km, aftershock productivity is particularly low, falling to about one-third the level between 100 and 300 km. Below 550 km, however, aftershock productivity abruptly increases to more than twice as great as that between 100 and 300 km. A similar pattern is seen in alternate measures of aftershock productivity, including the raw numbers of aftershocks, the fraction of mainshocks that have aftershocks, and the fraction of mainshock seismic moment released in aftershocks. The depth dependence of aftershock productivity is consistent with a previously proposed change in earthquake generation mechanism near 300 km depth. Notably, the abrupt increase in aftershock productivity below 550 km corresponds with abrupt decreases in earthquake duration and complexity and sug- gests a major change in the rupture mechanism at that depth, as well. Online Material: Catalog of mainshocks and related aftershocks used in this study.