Earthquake characteristics, including spatial size and perimeter of cells generated by tessellating earthquake epicenters and interevent times and distances are investigated using data around Los Angeles, California between 1984 and 2004 from the Southern California Earthquakes Data Center. The tapered Pareto distribution is found to fit substantially better than competing models such as the Pareto, lognormal, and exponential distributions. These results are confirmed by repeated sampling of
earthquakes over different time periods and magnitudes.