A 24-year-old self-confessed "earthquake freak" from MIT has become a catalyst for a solution to a great transpacific riddle -- what makes the ground lurch and people die, a subject of interest in China and America?

During four months in China, U.S. scientistts Lucile Jones has punched a mound of unusual Chinese earthquake data into an American computer to produce a scientific paper slated to be the first such Sino-American effort published in a Chinese journal. She has been to three Chinese earthquake sites, looking for clues to the fate of her own fault-threatened California home.

"The great advantage the Chinese have in predicting earthquakes is 3,000 years of records, which they have combed very carefully" looking for events that recur just before large tremors, said Jones. "They don't have the money to do it any other way." She plans to leave behind a small America-made computer she brought with her for her Chinese colleaques to use.

Jones, who speaks Chinese, is one of the first five American researchers to come here under the new governmental scientific exchange program. The Chinese have given her a warm welcome; earthquake predicition is an area in which they know they have something to teach the Americans.

Earlier in this decade, Jones said, some U.S. scientists said anyone who claimed ability to predict earthquakes "was a quack." When the Chinese did just that at Haicheng in 1975, saving many lives, "it was a slap in our face," she said.

Jones' rapid entrance into China's earthquake-watching communnity has been aided by her unusual combination of training in geophysics at Massachusetts Institute of Technology, where she is working toward her doctorate and her command of Chinese, which she acquired as an undergraduate at Brown and during a year in Taiwan.

She has been obsessed with earthquakes for a long tme. "The first, I hear of a big quake in California, I'm taking the next plane to L.A.," she said.

The Chinese have been able to predict earthquakes not only because of their excellent historical records, but because of enormous manpower they have been able to employ in areas where they suspect tremos are about to occur. Local seismology bureaus look for sudden changes in well-water level, animal behavior and more scientific measurements such as electrical resistance in the soil.

Suspicious measurements in known fault areas lead local officials to set up observation teams, the great strength of the Chinese system. Before an August 1976 quake in Sichuan (Szechwan) Province, local officials organized 1,200 observation groups. The mass of data coming in led officials to order all residents out of their homes four days before the major shock, up to 7.2 on the Richter scale, hit the area.

"Only 30 people were killed in the area, most very old people who did not believe the prediction and refused to leave their homes," said Jones.

Sometimes, however, like the enormous Tanghshan earthquakes of 1976 that killed a reported 750,000, there are too few clues and the Chinese fail to predict a quake.

"Tangshan occurred on a dippy little fault that nobody paid any attention to," Jones said.

In Sichuan, attention was so intense that seismologists were able to observe personally the fireballs that untrained observers had reported before previous earthquakes.

The scientist concluded the phenomenon was caused by methane gas, released and ignited by underground events leading up to the main tremor.

Jones' analysis of the data from the 1975 Haicheng quake in Liaoning Province focussed on the foreshocks which in a way may have pointed toward the site of the final tremor. About 500 foreshocks occurred in all, a break for predictors who had no foreshocks to work with at Tangshan. Jones' analysis showed "foreshocks occurred on two different plane then one perpendicular to that, then back to the plane where the main shock occured."

Jones coauthored her report, scheduled for publication in the journal "Works of Chinese Geophysics," with Xu Jibjian, head of a state geophysics institute's research laboratory, and Wang Giquan, a femal researcher at the lab. Jones said Xu has "treated me sort of like a granddaughter," while Wang has accompanied her in all her travels about the country.

Despite their progress, the Chinese remain apologetic about Tangshan.

"Theey say, 'Yeah, we've done more earthquake predictions than anyone else, but we didn't predict Tangshan so we can't pat ourselves on the back,'" Jones said.

Jones says she does not think California can employ the mass observation techniques used by the Chinese, but may find some of the pre-shock clues, the Chinese have unearthed as useful. No matter how much data there is, she said, "this is still more of an art than a science."