THE ABILITY TO PREDICT earthquakes, which geologists believe may be achieved within a few years, undoubtedly will be a mixed blessing. It will provide a tool with which human lives can be saved every year. But it will also create enormous problems for individual citizens, as well as public officials. What do you do, for instance, if the prediction says - as weather forecasts now do - that there is a 50 per cent chance of an earthquake today? Or, what do you do if an earthquake is said to be highly likely within the next two weeks?

We raise these questions because Congress is well on its way toward doubling support for research into earthquakes and how to predict them. Legislation authorizing these new programs as well as the creation of a national plan to reduce earthquake hazards has already passed the Senate and is on its way through the House. Appropriations for earthquake programs already authorized have been increased in next year's budget.

These are, in our view, worthwhile actions. Geologists are close enough to some answers about earthquakes to justify a substantially greater effort. It is fortunate, as well as remarkable, that Congress has become interested in the problem without waiting for another great quake to hit the United States. Undoubtedly it was influenced by the heavy death toll from earthquakes in 1976, particularly in China, and from reports that the Chinese are deeply into - and far ahead of us in - earthquake research.

There have been some warning signals flashed about these research programs, both by geologists and social scientists. The geologists are worried about the usefulness of the predictions and the public reaction if they turn out to be either too generalized or wrong. The social scientists are worried about the impact on whole populations of too much knowledge about earthquakes. They point out that for years people have been moving into areas known to be earthquake-prone and seem quite prepared to run the risk of losing their property or their lives or both - someday. But the geologists wonder what would happen if, instead of saying that "someday" could be any time within the next century, they were able to say it was likely to come within the next 12 months. Would cities be deserted? Property values fall? Insurance rates skyrocket? Suck questions are well worth thinking rather urgently about. But the fact that they will arise is no reason to delay the research programs. Knowledge about earthquake prediction needs to move from the art it now is to the more exact science it can become.