Nicholas Van Der Elst
Andrea L. Llenos
2019
<p><span>Earthquake swarms present a challenge for operational earthquake forecasting because they are driven primarily by transient external processes, such as fluid flow, the behavior and duration of which are difficult to predict. In this study, we develop a swarm duration model to estimate how long a swarm is likely to last based on actuarial statistics of previous swarms in a given region. We demonstrate this approach using swarms that have been identified in the Salton trough in southern California, finding that swarms last an average of </span><span class="inline-formula no-formula-id"><span id="MathJax-Element-1-Frame" class="MathJax" data-mathml="<math xmlns="http://www.w3.org/1998/Math/MathML"><mo xmlns="">&#x223C;</mo><mn xmlns="">7</mn></math>"><span id="MathJax-Span-1" class="math"><span><span id="MathJax-Span-2" class="mrow"><span id="MathJax-Span-3" class="mo">∼</span><span id="MathJax-Span-4" class="mn">7</span></span></span></span><span class="MJX_Assistive_MathML">∼7</span></span></span><span> days and have a relatively constant 15%–16% chance of terminating each day for the first 14 days of the swarm. Cataloged swarm durations are exponentially distributed, so we use a Poissonian model for swarm termination to encapsulate and extend the actuarial statistics. We then show how using the swarm duration model would have affected the earthquake forecast that was released during the 2016 Bombay Beach swarm. The earthquake forecast is substantially improved by incorporating a probabilistic model for how long the swarm is likely to last.</span></p>
application/pdf
10.1785/0120180332
en
GeoScienceWorld
Improving earthquake forecasts during swarms with a duration model
article