Previous work

Reconstructing earthquakes from historical tsunamis.

This work used historical records of tsunamis—particularly accounts recorded by Dutch soldiers in Indonesia—to identify the earthquakes that caused them and strengthen modern seismic and tsunami risk assessment.

Historical tsunami inference

Connecting eyewitness accounts to modern hazard assessment.

Bayesian statistical techniques, including extensive use of Markov chain Monte Carlo (MCMC), made it possible to compare competing earthquake scenarios and quantify uncertainty from sparse historical observations.

The GeoClaw software package was used extensively to simulate tsunami generation and propagation, allowing modeled wave heights and arrival times to be compared with the historical record.

Jared Whitehead was supported by a Fulbright Faculty Award to visit Bandung Institute of Technology in the winter of 2025, strengthening research collaboration in Indonesia.

Selected publications

Published result

Sparse, anecdotal historical accounts can still yield meaningful earthquake estimates when their uncertainty is modeled explicitly.

Embracing Uncertainty in ‘Small Data’ Problems: Estimating Earthquakes From Historical AnecdotesJournal of Geophysical Research: Machine Learning and Computation 2 (2025)

Read the publication

Published result

Nearly 200,000 tsunami simulations helped compare competing source hypotheses for the 1820 Flores Sea earthquake.

A tale of two faults: statistical reconstruction of the 1820 Flores Sea earthquake using tsunami observations aloneGeophysical Journal International 237, 419–435 (2024)

Read the publication