The central limit theorem applies to sample means, not individual events. I’m not sure it has any relevance to predicting the probability of a single large outlier tsunami