Successes, misses and false alarms in predicting global/seasonal equatorial plasma bubble occurrences with the ROCSAT data

The density variations at a constant height such as observed by the circularly orbiting ROCSAT-1 spacecraft are studied to construct a “prediction model” for the occurrences, no-occurrences, misses and false alarms of global/seasonal equatorial plasma bubble (EPB) occurrences. This global/seasonal prediction model is different from previous studies that merely provide the seasonal EPB occurrence probability at a local time region or globally. The current model uses the density increment above the seasonal mean to predict the EPB occurrences by constructing a contingency table that counts the numbers of successes, fails, misses, and false alarms in predicting the EPB occurrences. Different thresholds of density increment are used for the criteria in prediction to obtain an optimal result in the prediction model. The success of the model for predicting the global EPB occurrences or no-occurrences varies between 75% and 85% for any season between 1999 and 2004 during the moderate to high solar activity period. Studies of the causes for the misses and false alarms in the existing prediction result lead us to propose that additional observational parameter besides the density increment in the prediction model should be included. This additional parameter should be the seeding agent for the RTI process that related to the atmospheric/ionospheric background conditions in the lower ionosphere.