Algorithm Research & Development

ShakeAlert is a distributed system, enabling the independent development of individual components. It has also been designed to use multiple algorithms, whose information is combined to provide single alerts to end users. This framework facilitates ongoing development and updates of the system, as new research is conducted and methods developed.

Scientists at Caltech have worked on a range of EEW research problems, with interest in the concept stretching back decades. Areas of past, present and future research include:

  • Single-station, ‘onsite’ early warning, such as the τc-Pd algorithm that is one of the prototype system algorithms
  • Finite-fault algorithms that can well characterize large magnitude, long rupture events
  • The use of Bayesian statistics to probabilistically consider problems such as combining different algorithm output, incorporation of prior seismicity information and event waveform characterization.
  • Ways to incorporate new data streams, such as low-cost, high-density seismic sensors, GPS data etc.

Further information on specific algorithms or methods can be found in the additional pages within this section.