The prediction did not begin on the day of the earthquake. It began in 1970, after a series of large earthquakes from 1966 to 1969 prompted a national meeting to assess where the next major seismic threat lay. Scientists concluded that the pattern was migrating northeast, and that a magnitude 5-6 earthquake was likely along the Bohai Gulf. The Jinzhou, Zhuanghe, and Yalu River faults were identified as the primary dangers. Liaoning province saw a surge of scientific activity: new seismic stations were built, monitoring equipment was deployed, and a public education campaign taught residents how to recognize earthquake precursors and respond to evacuation orders.

By June 1974, four years of data collection led to a refined prediction: the earthquake would likely occur within two years in the northern Bohai Gulf. Microearthquake activity was increasing. Leveling surveys showed ground deformation. Then, in the days before February 4, 1975, the foreshock sequence escalated dramatically. Dozens of small earthquakes rattled the Haicheng area, each one louder than the last. It was this pronounced acceleration -- tremors growing more frequent and more powerful -- that prompted the final evacuation orders, issued just hours before the main shock. Scientists would later debate whether this constituted a genuine prediction or whether officials simply reacted to unmistakable warning signs, but the distinction mattered less to the people who survived.

The earthquake struck in the dead of a Manchurian winter. With a Modified Mercalli Intensity of IX at the epicenter, the damage was catastrophic: buildings collapsed across Haicheng, infrastructure was destroyed, and the cold compounded the suffering. Three hundred seventy-two people froze to death in the aftermath, and 6,578 suffered frostbite -- casualties that speak to the brutal reality of surviving an earthquake when nighttime temperatures plunge far below zero. Total fatalities were kept below 2,500, with 27,500 injuries. Without the evacuation, estimates suggest more than 150,000 people would have been injured. The damage reached as far as Seoul, where buildings swayed and transformers tripped from the long-period seismic waves.

The Haicheng prediction's legacy was complicated just eighteen months later. On July 28, 1976, a magnitude 7.8 earthquake struck Tangshan, 500 kilometers to the southwest, killing over 240,000 people. None of the foreshock patterns, ground deformation signals, or other precursors that had preceded Haicheng were observed before Tangshan. There was no prediction, no evacuation, and the death toll was one of the highest in recorded history. The contrast was devastating to the idea that earthquake prediction was becoming a reliable science. Many seismologists concluded that the Haicheng evacuation had been, at best, a fortunate response to an unusually clear foreshock sequence -- not a repeatable methodology. The debate continues. What is not debatable is that on one winter evening in Liaoning, the right call was made at the right time, and a city was saved.