Predictive Distributions as a Data-Driven Indicator for Phase-Pick Confidence

Yongsoo Park, Nathan T. Stevens, & Brent G. Delbridge

Submitted August 30, 2026, SCEC Contribution #15124, 2026 SCEC Annual Meeting Poster #TBD

Reliable uncertainty estimates are essential for using neural-network phase pickers in automated seismic monitoring and catalog construction. Using the Pacific Northwest Seismic Network dataset, we show that the spread of predicted probability mass functions (PMFs) tracks relative picking difficulty even when models are trained only with one-hot-arrival-time labels, which assign probability one to the catalog pick time and zero to all other times, and without explicit quality information. Controlled label-perturbation experiments indicate that the predictive distributions inherit both systematic bias and label variability from the training targets, allowing probabilistic structure to emerge under deterministic supervision. We also find that disagreement across separately trained models increases for more difficult samples but can be substantially reduced by training on ensemble-averaged soft labels. Together, these results show that PMF-derived measures such as interquartile range provide objective, reproducible indicators of picking difficulty, with practical value for both seismic monitoring and retrospective catalog construction.

Citation
Park, Y., Stevens, N. T., & Delbridge, B. G. (2026, 08). Predictive Distributions as a Data-Driven Indicator for Phase-Pick Confidence. Poster Presentation at 2026 SCEC Annual Meeting.

Related Projects & Working Groups
Seismology