Ensemble prediction model of solar proton events associated with solar flares and coronal mass ejections

An ensemble prediction model of solar proton events （SPEs）, combining the information of solar flares and coronal mass ejections （CMEs）, is built. In this model, solar flares are parameterized by the peak flux, the duration and the longitude. In addition, CMEs are parameterized by the width, the speed and the measurement position angle. The importance of each parameter for the occurrence of SPEs is estimated by the information gain ratio. We find that the CME width and speed are more informative than the flare’s peak flux and duration. As the physical mechanism of SPEs is not very clear, a hidden naive Bayes approach, which is a probability-based calculation method from the field of machine learning, is used to build the prediction model from the observational data. As is known, SPEs originate from solar flares and/or shock waves associated with CMEs. Hence, we first build two base prediction models using the properties of solar flares and CMEs, respectively. Then the outputs of these models are combined to generate the ensemble prediction model of SPEs. The ensemble prediction model incorporating the complementary information of solar flares and CMEs achieves better performance than each base prediction model taken separately.

Evidence for a Strong Correlation of Solar Proton Events with Solar Radio Bursts

A statistical analysis is made on the correlation between solar proton events with energies >10Mev and solar radio bursts during the four-year period from 1997 November to 2000 November. We examine 28 solar proton events and their corresponding solar radio bursts at 15400, 8800, 4995, 2695, 1415, 606, 410 and 245 MHz. The statistical result shows that there is a close association between solar proton events and ≥3 solar radio bursts occurring at several frequencies, one or two days before. In particular, it is noteworthy that proton events occurring in pairs within the same month are preceded 1-2 days by individual radio bursts and most of the radio bursts of solar flares occur at all eight frequencies. Those 245 MHz radio bursts associated with proton events have intense peak fluxes (up to 67000 sfu). Solar proton events are preceded 1 or 2 days by≥ 3 radio bursts at several frequencies and proton events occurring in pairs within the same month are preceded 1 or 2 days by some individual radio bursts. These correlations may be used for providing short-term or medium-term prediction of solar proton events.

Short-term solar eruptive activity prediction models based on machine learning approaches:A review

Solar eruptive activities,mainly including solar flares,coronal mass ejections(CME),and solar proton events(SPE),have an important impact on space weather and our technosphere.The short-term solar eruptive activity prediction is an active field of research in the space weather prediction.Numerical,statistical,and machine learning methods are proposed to build prediction models of the solar eruptive activities.With the development of space-based and ground-based facilities,a large amount of observational data of the Sun is accumulated,and data-driven prediction models of solar eruptive activities have made a significant progress.In this review,we briefly introduce the machine learning algorithms applied in solar eruptive activity prediction,summarize the prediction modeling process,overview the progress made in the field of solar eruptive activity prediction model,and look forward to the possible directions in the future.

The space environment monitor aboard FY-2 satellite

The space environment monitor(SEM)aboard FY-2 satellite consists of the high energy particle detector(HEPD)and the solar X-ray flux detector(SXFD).The SEM can provide real-time monitoring of flare and solar proton event for its operation at geostationary orbit and is also the first Chinese space system for monitoring and alerting solar proton event.During the 23rd solar maximum cycle,almost all the solar proton events that took place in this period are monitored and some of them are predicted successfully by analyzing the characteristics of X-ray flare monitored by the SEM.Some basic variation characteristics of particle at geostationary orbit are found such as day-night periodic variation of particle flux,the electron flux with energy>1.4 MeV in the scope from 10 to 200/cm^(2).s-sr and the proton flux with energy>1.1 MeV in the scope from 600 to 8000/cm^(2)-s.sr during the time with no magnetic storm and solar eruption.