Solar Activity from 2006 to 2014 and Short-term Forecasts of Solar Proton Events Using the ESPERTA Model

To evaluate the solar energetic proton (SEP) forecast model of Laurenza et al., here termed ESPERTA, we computed the input parameters (soft X-ray (SXR) fluence and similar to 1 MHz radio fluence) for all >= M2 SXR flares from 2006 to 2014. This database is outside the 19952005 interval on which ESPERTA was developed. To assess the difference in the general level of activity between these two intervals, we compared the occurrence frequencies of SXR flares and SEP events for the first six years of cycles 23 (1996 September2002 September) and 24 (2008 December2014 December). We found a reduction of SXR flares and SEP events of 40% and 46%, respectively, in the latter period. Moreover, the numbers of >= M2 flares with high values of SXR and similar to 1 MHz fluences (>0.1 J m(-2) and >6 x 10(5) sfu x minute, respectively) are both reduced by similar to 30%. A somewhat larger percentage decrease of these two parameters (similar to 40% versus similar to 30%) is obtained for the 20062014 interval in comparison with 19952005. Despite these differences, ESPERTA performance was comparable for the two intervals. For the 20062014 interval, ESPERTA had a probability of detection (POD) of 59% (19/32) and a false alarm rate (FAR) of 30% (8/27), versus a POD = 63% (47/75) and an FAR = 42% (34/81) for the original 19952005 data set. In addition, for the 20062014 interval the median (average) warning time was estimated to be similar to 2 hr (similar to 7 hr), versus similar to 6 hr (similar to 9 hr), for the 19952005 data set.