The real problem are the occasional "coronal mass ejections", large masses of charged particles thrown off irregularly by the sun that can be quite energetic.
You could set yourself up for some quote mining here. CMEs are not that occasional, they occur quite frequently. It's a certain type of CME that creates the occasional energetic solar storms akin to the August 1972 event and the 2003 Halloween storm. The most recent literature that I have read from the SOHO research describes a correlation with 'CME speed' and the probability of a solar storm.
If this speed is greater than that of the solar wind then a favourable charge distribution occurs in the solar plasma that shock drives protons to GeV energies. My main desk top computer is in another country right now, and that is where all the literature I have on this subject is stored (bad planning on my part). If I recall correctly, 99% of CMEs have a speed less that that of the solar wind, so the conditions to drive protons to GeV energies are rare.
Another issue, and this is me getting on my hobby horse, is that solar flares are different to these rare CME events. A certain antipodean hoaxer attributes proton radiation levels from such CME events to every H-alpha event in the NOAA record. There are many more issues with his calculations and how he treats flares, but he never got past base 1 with Jay and others at the IMDB regarding the correlation between proton fluxes and H-alpha events.
Since our friend at YouTube is allowed to invoke NOAA data, I'm still waiting for him to invoke this page from the NOAA source he is so keen on. It gives a really good overview of the rarity of such events. There's plenty of interesting data in the NOAA archives, if you are prepared to dig, and it shows that the proton flux levels in space are not awash with GeV protons that often.
NOAA Space Weather CyclesAn S5 event is the most extreme event and may not even occur in a cycle. The 1859 Carrington event is one such example and proves to be an interesting case study when considering our antipodean friend and the IMDb debacle.
After that public mauling he claimed that it was impossible to correlate H-alpha events with proton fluxes before 1972 because there was no proton data before 1972, and Jay was just being a big poopy face again by asking questions that he could not possibly answer. Of course, he was royally found out by Jay, and once he realised that he could not correlate H-alpha data with proton fluxes he attacked Jay's line of questioning.
This of course begs the question:
What data has our antipodean been using to calculate radiation exposures if there is no proton data before 1972?There is proton data from before 1972, and our antipodean friend knows this because he is using data from a pre-1972 CME event which is based on proton data gathered with weather balloons. He has royally dodged the IMBd question by accusing Jay of asking for the impossible, but conveniently forgets the data he uses is derived from proton data.
Which is why I invoke the Carrington event, a pre-space age, pre particle-physics age CME event. How do we know that this was a powerful CME event? The answer can be from ground based records and correlating them with ice core samples.
Sorry for the digression.