Optimising Deep-Learning First-Break Picking

Accurate first-break picking is a key requirement for high-quality velocity model building, but manual picking remains
time-intensive and error-prone. Existing single-station automatic approaches neglect spatial coherence across neigh-
bouring traces and other deep-learning–based first-break picking approaches sacrifice temporal resolution through
data downsampling, limiting their reliability. We present DeepFB, a U-Net–based neural network designed for robust
automatic first-break picking in active-source seismic data. The model operates on overlapping chunks of multiple
traces, preserving temporal resolution while exploiting spatial correlations. Automated hyperparameter optimisation
using the evolutionary algorithm Propulate yielded optimal model configurations without manual tuning. Among
other hyperparameters, we tested whether model performance improves when training is performed with a reduced
traveltime dataset or when noise augmentation techniques are used to improve first-break picking, particularly in noisy
ocean-bottom seismometer records. Application to the HIPER2 experiment at the coastline of Ecuador demonstrates
... mehr
that DeepFB achieves picking accuracy comparable to manual picking, with residuals close to the manual picks from
our test dataset. Tomographic inversions based on automatic and manual picks produce consistent velocity models,
with only localised discrepancies near complex geological structures. DeepFB thus enables accurate, efficient, and
scalable first-break picking, reducing manual workload while maintaining reliability