11:32:04	 From  Hazel Bain  : Good morning/afternoon/evening :) 
11:35:23	 From  Kathryn Whitman : Agenda https://docs.google.com/document/d/1Nl_r-FrRzdvTyWEUNMR5Nn-TriS2-YnUHDqtwGrEUVg/edit?usp=sharing 
11:39:53	 From  Kathryn Whitman : Download Graham's slides: https://drive.google.com/file/d/1kGF-H00wA-oKZt6x22wD7fvS7uxJoroj/view?usp=sharing
11:42:55	 From  Kathryn Whitman : Please everyone be sure to mute your mics when you are not speaking
11:56:20	 From  Phil Quinn : I think it's a good idea to get away from asking the generalizing question, "is my model good?", and instead ask the more specific question, "what is my model good at and what is it bad at?"
11:57:44	 From  Hazel Bain  : @phil, I agree. Especially if we are looking to use models for forecasting. It is understood that no model is perfect and that's ok. But what is important is to know what situations a model would perform well or not so well
12:01:30	 From  Leila : Hi all, The H2-01 CME arrival splinter is starting now (1700 UT)
12:24:45	 From  Hazel Bain  : The SWPC scale was defined based on user needs
12:25:13	 From  Hazel Bain  : (probably also a little to do with what observations were available on an ops spacecraft such as GOES)
12:25:22	 From  Kathryn Whitman : For SRAG operations, we've chosen >100 MeV exceeds 1 pfu, because that is the energy and intensity threshold that will cause an increase in dose inside of a shielded spacecraft
12:26:21	 From  Phil Quinn : and the >10 MeV above >10 pfu since that's the energy that protons can penetrate spacesuits
12:26:33	 From  Hazel Bain  : SWPC is interested in > 10 MeV exceeding 10 pfu. And >100 MeV exceeding 1 pfu (because users like SRAG care about that)
12:29:01	 From  Junxiang Hu : Is there a specific reason to choose integral flux over differential flux? Early rise in >10 MeV flux could be totally from >100 MeV contributions
12:32:55	 From  Piers Jiggens  : The event flux threshold (just like the probability threshold) should be user-inspired. I think the important step for now is to allow us to question different definitions and people not to say "this is (not) an event" (based only on >10 MeV and pfu) without putting this in context of the definition being used. Note that there is an impact here on forecast horizon as well (especially for poorly-connected SPEs)..
12:33:29	 From  Phil Quinn : We use >10 MeV above 10 pfu because that's when protons can penetrate a spacesuit.  But if >10 MeV flux is from >100 MeV contributions, then the protons can definitely penetrate the suit.  And the action we take because of that is the same
12:33:55	 From  Kathryn Whitman : Mertens and Slaba: characterization of SEP contribution inside spacecraft https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2019SW002363
12:35:11	 From  Piers Jiggens  : 10 pfu is arbitrary (and based more on instrument background than anything else) regardless of your justification based on energy (what it can penetrate).
12:35:29	 From  Hazel Bain  : @Piers, I think that contributes to the idea using all the information that a model produces...i.e. not thresholding based on a probability or threshold and boiling down to a yes/no on a very narrow definition of an event
12:39:56	 From  Bill Swalwell : List of false alarm events in this paper: https://link.springer.com/article/10.1007%2Fs11207-017-1196-y
12:41:30	 From  Kathryn Whitman : Thanks @Bill
12:41:46	 From  Athanasios Papaioannou : Interesting paper that brings together several ML techniques and examines a pretty nice set of metrics https://link.springer.com/article/10.1007%2Fs11207-021-01837-x
12:41:47	 From  Phil Quinn : Yes thanks Bill.  Also hi :)
12:42:36	 From  Bill Swalwell : Hi Phil! Good to hear from you!
12:43:47	 From  Athanasios Papaioannou : Also this new paper https://agupubs.onlinelibrary.wiley.com/doi/epdf/10.1029/2021SW002794 discusses the inherently imbalanced problem of the SEP prediction and the implication of this for FAR
12:46:59	 From  Phil Quinn : Thanks Thanasis
12:47:13	 From  Hazel Bain  : Thanks for the interesting papers folks
12:55:50	 From  Alessandro Bruno : Unfortunately the SOHO detector has saturation issues during large SEP events, in addition to several data gaps
12:57:46	 From  Alessandro Bruno : These issues affect the estimate of both peak and event-integrated intensities (for large events)
12:58:25	 From  Junxiang Hu : Do STEREOs have the similar issue?
12:58:49	 From  Kathryn Whitman : Thanks @Alessandro
12:59:33	 From  Kathryn Whitman : @Junxiang from my understanding, STEREOs are not affected by the problems of GOES, but doesn't go up to as high energies. @Ian Richardson knows more
12:59:59	 From  Athanasios Papaioannou : For operational or for scientific use? A recent paper by Patrick https://arxiv.org/pdf/1611.03289.pdf has shown that SOHO can identify reliably SEP events up to 500 MeV, after that there is contamination by electrons
13:01:17	 From  Athanasios Papaioannou : However, SOHO is not to be used in an operational mode (as we speak). Nonetheless, I agree with Arik. We need better (more reliable) data and in an operational mode.
13:06:03	 From  Ian Richardson : I'm not aware of any problems with the STEREO HETs.  The electron intensities need to be multiplied by ~15 however.