Millenary M_w > 9.0 earthquakes required by geodetic strain in the Himalaya

Abstract

The Himalayan arc produced the largest known continental earthquake, the Mw ≈ 8.7 Assam earthquake of 1950, but how frequently and where else in the Himalaya such large‐magnitude earthquakes occur is not known. Paleoseismic evidence for coseismic ruptures at the front of the Himalaya with 15 to 30 m of slip suggests even larger events in medieval times, but this inference is debated. Here we estimate the frequency and magnitude of the largest earthquake in the Himalaya needed so that the moment released by seismicity balances the deficit of moment derived from measurements of geodetic strain. Assuming one third of the moment buildup is released aseismically and the earthquakes roughly follow a Gutenberg‐Richter distribution, we find that Mw > 9.0 events are needed with a confidence level of at least 60% and must return approximately once per 800 years on average.

A Geodesy‐ and Seismicity‐Based Local Earthquake Likelihood Model for Central Los Angeles

“…Some studies iteratively estimate M max in cumulative magnitude-frequency space (Stevens & Avouac, 2016, 2017; others estimate it using total fault areas and scaling relations (Field et al, 2014) or assume a value for the maximum earthquake's recurrence interval (Hsu et al, 2016). First, M max is unknown due to the short history of observation.…”

Determination of Mmax from Background Seismicity and Moment Conservation

“…We take into account earthquakes that occur up to 150 km north perpendicular to the surface trace of the Main Frontal thrust, and we have excluded the 25 April 2015 Gorkha earthquake and its aftershocks. Figure 6 shows the results and a comparison with the historical and paleoseismic record for the past 1000 years, summarized by Stevens and Avouac (2016).…”

“…Our method yields a 66% chance that M max has to be over M w 8.7 (the largest historical earthquake) and about a 95% chance that it has to be more than M w 8.5 and less than M w 9.5. For comparison, Stevens and Avouac (2016), who carried out a similar analysis with the addition of historical and paleoseismic data, assess there to be a 66% chance that M max has to be 9 or over, higher than M w 8.7 here. Uncertainties in the b-value lead to uncertainties that will be larger as the predicted M max becomes larger, because this means that the data have to be extrapolated further.…”

“…It has produced large earthquakes in the past and poses a large threat to densely populated areas. We use a moment buildup rate of 15:1 1 × 10 19 N·m=yr (Stevens and Avouac, 2015) and allow 33% of the moment to be released nonseismically (by afterslip), following Stevens and Avouac (2016). We take into account earthquakes that occur up to 150 km north perpendicular to the surface trace of the Main Frontal thrust, and we have excluded the 25 April 2015 Gorkha earthquake and its aftershocks.…”

“…There is, however, a physical limit, due to the fact that strain due to seismicity must match the strain resulting from geological motion across the fault on average over geological time (Brune, 1968). If the fault-slip rate is known, additional constraints on the seismicity model can thus be derived based on the seismic moment budget (e.g., Molnar, 1979;Ader et al, 2012;Kagan and Jackson, 2013;Rong et al, 2014;Avouac, 2015;Bird et al, 2015;Stevens and Avouac, 2016). There is now abundant evidence that seismic slip occurs in locations where faults were locked previously; this information can be gathered from geodetic observations (e.g., Avouac, 2015).…”