Satellite and Ground-Based Observations of Solar Energetic Particle Event Onsets: a Comparison of Two Critical Observational Platforms

Solar energetic particles (SEPs) propagating to Earth cause airline communication blackouts, radiation hazards for astronauts and airline passengers, and upsets in as well as degradation of spacecraft electronics. The NOAA Space Weather Prediction Center uses observations from the Geostationary Operational Environmental Satellites (GOES) to monitor SEPs. This satellite-based system for monitoring SEP events is complemented by a network of ground-based neutron monitors (NMs). Detections of SEPs by NMs, also called ground level enhancements (GLEs), are particularly important for mitigating radiation dosage to airline passengers and crew during a SEP event.  A previous study by Kuwabara et al. (Space Weather, v. 4, S10001, 2006) concluded that a NM network could detect GLE onsets ~10-30 minutes earlier than GOES. However, their study did not account for differences in the instruments, the time resolution of the data, and the alert protocols.

Using newly-available high-resolution data from GOES 8-12 and NMs, we analyzed 16 GLEs from 1997-2006 and conducted a more ‘apples-to-apples’ comparison of these two systems’ relative GLE detection times using 1-minute-cadence data. We reproduced Kuwabara et al.’s results for NM onsets using their real-time onset-detection technique (which involves a moving trailing average), then applied the same technique to the GOES observations, tuned to the different background noise levels such that false alerts are minimized. We compared the NM and GOES onset times and discovered that they did not differ as drastically as Kuwabara et al. claimed. Moreover, there was no significant local time dependence of the GOES onset times. The median difference between GOES and NM onsets was 1 minute (GOES lagging NM), with the 80th percentile of the differences (GOES onset – NM onset) being 3 minutes. Weaker GLEs tended to have more varied onset times, whereas stronger GLEs yielded smaller onset differences. Based on the expected travel times of relativistic protons and on the nature of the technique used, we conclude that the differences between GLE onset times derived from NM and GOES observations using a similar real-time technique are mainly due to noise.