Parkfield and San Andreas offer lessons in earthquake prediction research.

The Parkfield Experiment on the San Andreas Fault forecasted a quake for 1985–1993 but the magnitude-6.0 event occurred in 2004; scientists continue to deploy instruments, satellites, laboratory studies and computing methods to study faults and improve probabilistic forecasts.

Scientists continue to study how and why earthquakes occur and whether they can be anticipated. The Parkfield Experiment on the San Andreas Fault became a focal test case because Parkfield had a historical pattern of significant quakes about every 22 years. From the mid-1980s researchers installed extensive monitoring instruments and issued a probabilistic forecast for a quake between 1985 and 1993. A magnitude 6.0 earthquake did occur in Parkfield in 2004, later than the forecast window, and the episode influenced subsequent research priorities. Key facts: - Parkfield's recorded seismograms for 1922, 1934 and 1966 were notably similar, and some events showed foreshocks about 17 minutes before the main shock. - From the mid-1980s scientists installed seismographs, strainmeters (at about 650 feet / 200 meters), magnetometers, creepmeters and other instruments along the fault. - Researchers forecasted, with roughly 90–95% confidence, that the next Parkfield earthquake would occur between 1985 and 1993; the main shock occurred in 2004. - Predicting specific earthquakes is difficult because many faults differ in shape and structure, subsurface properties are often unknown, and earthquake records do not extend far enough in time. - Ongoing advances include seismic hazard mapping, physics-based models of ground behavior, satellite methods (InSAR, LIDAR, GPS), laboratory experiments, machine learning, and P-wave early warning systems that can provide seconds of advance notice. Summary: The Parkfield case illustrated limits of deterministic forecasting and reinforced the need for probabilistic approaches. Research continues with broader instrumentation, remote sensing, laboratory work and improved computational methods; timelines for precise forecasting remain undetermined.