A Cellular Seismology Investigation of the Caribbean Plate Region
A CELLULAR SEISMOLOGY INVESTIGATION OF THE CARIBBEAN PLATE REGION Jacqueline Rose Cinella Dr. Alan Kafka, thesis advisor The Caribbean Plate region (CPR) provides a natural laboratory for a Cellular Seismology (CS) investigation due to its diversity of plate boundaries and the associated seismotectonic environments. Study of the CPR is of interest due, in part, to the occurrence of devastating earthquakes, such as the 2010 earthquake in Haiti. Using data from this region, this study addresses fundamental questions about CS forecasts: to what extent do the time period covered by earthquake catalogs, the threshold magnitude of the "before" and "after" catalogs, and the type of plate boundary affect the CS success rates? In general, CS success rates are quite high for the CPR as a whole, with hit percentages often at least 86%. Results suggest that the type of plate boundary affects the success of the CS forecasts, with differences between CS characteristics of broad plate boundary zones that include intraplate type deformation, versus along transform boundaries, spreading centers or along subduction zones. Analysis of subregions surrounding Haiti show that the 2010 Haiti earthquake was not unequivocally identified by CS as a likely location of a future large earthquake, although detailed analysis of small earthquakes and historical earthquakes did uncover past seismicity in the vicinity of that event. Thus, continued monitoring of smaller earthquakes will be important for understanding the strengths and weaknesses of seismicity-based earthquake forecasting. Hit percentages in subregions surrounding Haiti were very low compared to other subregions outlining the boundaries of the CPR, again suggesting plate boundary affects CS results. Although CS does, in general, predict locations where future earthquakes are likely to occur, in seismicity-based forecasting we need to always be cautious of cases like the Haiti earthquake where the relationship between past seismicity and large, damaging earthquakes is complex. In general, the results of this study show that CS is a useful tool for furthering our understanding of the extent to which past seismicity delineates zones where future earthquakes are likely to occur.