Roger's right.

The first and apparently unsurmountable obstacle is for you to write a report showing your method works. Which quakes, which precursory data, which patterns worked, in a reproducible and and statistically valid way.

Such a report would generate interest. Unless the patterns are crystal clear, this would need to be followed by showing the method works for blind predictions.

The first paper could be just ten pages long, but if you cannot write it, I seriously doubt you have any idea whether your method works.