12:36:20 From Hazel Bain : I think there are some people in the main room that might be planning to come to this breakout 12:38:22 From Hazel Bain : will do 12:43:51 From Leila : https://ccmc.gsfc.nasa.gov/assessment/topics/SEP/campaign2020.php 12:58:32 From Arik Posner : HESPERIA-REleASE development was led by Olga Malandraki. 13:05:29 From Janet Luhmann : Katie, Leila et al. Can you clarify whether these models were all run in on-th-fly forecast mode versus retrospectively run (allowing experimental adjustments?)? 13:09:31 From Mark Dierckxsens : Do others still hear Katie? 13:09:37 From Arik Posner : No 13:09:37 From Christina Lee (SSL/UCB) [she/her] : No 13:09:39 From Manolis Georgoulis : I cannot 13:09:41 From Ian Richardson : nO 13:09:41 From Anastasios Anastasiadis : NO 13:09:41 From Christina Lee (SSL/UCB) [she/her] : I thought it was just me 13:09:43 From Junxiang Hu : Blacked out 13:09:44 From Leila : I can’t hear Katie either 13:09:49 From Manolis Georgoulis : Yes, same here 13:12:35 From Kathryn Whitman : oh my. When did I drop off? I did the whole presentation! 13:12:52 From Christina Lee (SSL/UCB) [she/her] : Several minutes ago 13:19:10 From Piers Jiggens : Would be nice to include a time stamp of when a forecast could be made (i.e. no data after that time was used) 13:20:44 From Christina Lee (SSL/UCB) [she/her] : Will this presentation be made available to this group? There was A LOT of (very useful!!) material to digest. :) 13:21:09 From Leila : yes we will post it for everyone on the team website 13:22:00 From Leila : https://iswat-cospar.org/H3-01 13:22:10 From Leila : https://ccmc.gsfc.nasa.gov/assessment/topics/helio-sep.php 13:22:21 From Piers Jiggens : sorry, I have to leave now. I'll check the outcome and hope to be here for the whole session tomorrow. Thanks all! 13:23:45 From Kathryn Whitman : thank you Piers! 13:36:22 From Leila : Here are links to real-time forecasts on the SEP scoreboard: https://sep.ccmc.gsfc.nasa.gov/probability/
https://sep.ccmc.gsfc.nasa.gov/intensity/
https://ccmc.gsfc.nasa.gov/challenges/sep.php 13:36:26 From Kathryn Whitman : Validation presentation: https://drive.google.com/file/d/1-KKKbQtLYemRZ75SADL56M3pfui6I5gR/view?usp=sharing 13:37:02 From Erika Palmerio : should we just request access to the presentation? 13:44:25 From Christina Lee (SSL/UCB) [she/her] : Junxiang’s comment about relying on CME input got me thinking of this: @Katie @Leila, do you assess the performance of the SEP models separately for those that rely (solely?) on CME input vs other kinds of inputs? 13:44:40 From Manolis Georgoulis : Apologies, I have to run. Thanks for the very interesting discussion - I will make sure to join tomorrow, as well. 13:45:37 From Hazel Bain : @christina we've provided some CME inputs for modelers to use. To removed that problem of CME fitting subjectivity 13:47:51 From Leila : Christina, inputs to some SEP models are listed in slide 2 of this presentation:
https://ccmc.gsfc.nasa.gov/challenges/sepinfo/202010_ESWS_Mays_SEP_Scoreboard.pdf 13:48:17 From Christina Lee (SSL/UCB) [she/her] : Thanks for your additional comments, @Hazel and @Leila! 13:48:24 From Leila : inputs are listed in the SEP model table at the bottom of this page https://ccmc.gsfc.nasa.gov/challenges/sep.php (compiled by Mike Marsh) 13:49:02 From Nariaki Nitta : What is the definition of supra thermal injection efficiency? Is it readily available in near real time? 13:50:07 From Anastasios Anastasiadis : @Katie thank you for sharing the presentation. I have to go. I will join in tomorrow. 13:53:28 From Hazel Bain : @Junxiang the other thing to think about is that by validating models that use CMEs as inputs, if we find that they could really improve SEP forecasts, that is reason for us and folks at the ops centers e.g. NOAA SWPC, to advocate for coronagraph images to be made in as close to real time as possible. That would be suggested as an observational gap 13:58:03 From Junxiang Hu : Having coronagraph early will be quite helpful. CME speed and longitude have a strong effect on SEP flux magnitude and the increase/decay speed for physics based models afaik 14:00:09 From Alessandro Bruno : Having the uncertainties on the CME speed/direction would help as well (for error propagation) 14:00:10 From Hazel Bain : @junxiang there will be operational coronagraphs on the next GOES spacecraft and on the SWFO-L1 mission 14:00:26 From Leila : Some folks aat NCAR and GSFC are also looking into early CME info from the MLSO K-Cor ground-based coronagraph and building a SEP all clear based on that 14:00:39 From Hazel Bain : SWFO-L1 will be launched in early 2025. I'm not certain when the GOES sat with coronagraph is scheduled for 14:01:16 From Kathryn Whitman : @Alessandro - noted 14:01:27 From Leila : Good point Alessandro, that could be used to provide a SEP forecast ensemble, error bar, or confidence 14:01:36 From Alessandro Bruno : right 14:02:25 From Hazel Bain : @Alessandro, Katie, Leila and the folks at M2M have looked at some error propagation from CME inputs 14:02:37 From Hazel Bain : a sensitivity study of the input parameters of sorts 14:03:38 From Kathryn Whitman : @Hazel, yes we have done a study using uncertainties on CME parameters with SEPMOD. It was valuable to test out model performance within the range of uncertainty. We can provide information like that for all the challenge events 14:04:27 From Christina Lee (SSL/UCB) [she/her] : @Hazel: does input parameters of sorts include the accuracy of the modeled background solar wind? (I assume yes, but want to confirm) 14:05:57 From Nariaki Nitta : Coronagraph observations are essential for SEP nowcast, but Erika’s example of SEP starting before the CME reaches the Enlil inner boundary suggests the importance of information on flares and low coronal dynamics. 14:07:14 From Hazel Bain : @Nariaki yes we need to understand the CME propagation and SEP production below 21Rsun. BUT all is not lost. it is important to not just forecast the onset, we also need to know how long we expect the event to last and how intense it will be 14:08:44 From Hazel Bain : @christina you mean which input parameters were provided for the challenge? Or in general I don't think if anyone has paired the accuracy the solar wind in the model with the SEP forecast on top of that 14:09:21 From Hazel Bain : but it's a good point. in CME propagation models some of the error comes from CME fitting subjectivity, some from the solar wind model that the CME is moving through 14:12:38 From Ian Richardson : We (inc. Chris, St Cry, Joan Burkepile) have characterized MLSO CME parameters for ~100 events with SEPs, but we're still figuring out how the CME and SEP parameters are related and whether there is a possibility of using ground-based CME observations as an element of SEP prediction. 14:13:10 From Ian Richardson : St Cyr! 14:13:26 From Christina Lee (SSL/UCB) [she/her] : @Hazel: I was asking that question “in general”, but now I am curious about whether the accuracy of the background solar wind was considered in the challenge 14:13:27 From Christina Lee (SSL/UCB) [she/her] : :) 14:13:58 From Rui F. Pinto : @erika (et al) I see a good candidate here to help us out on the connectivity team (S2-O5) 14:15:50 From Hazel Bain : @christina, not that I am aware of 14:15:57 From Leila : Christina - the H2-01 CME arrival team does consider background solar wind. For SEP validation it hasn’t been done because physics based SEP models using MHD models have never been directly compared yet, so we are not there yet 14:16:10 From Hazel Bain : @christina, I'm trying to think how to do that best... 14:16:56 From Hazel Bain : @leila I'd be interested in hearing about how the CME arrival team handles that 14:17:46 From Erika Palmerio : @Ian the study you described sounds super interesting! Are there plans to present the related results anywhere in the near future? 14:17:52 From Kathryn Whitman : @Christina, your point is right that the performance of SEP models that pair with background solar wind is really the performance of the two models together 14:19:12 From Kathryn Whitman : and no, we haven't tried to assess those separately so far 14:23:01 From Leila : The CME arrival team also hasn’t figured it out completely, beyond correlations between background wind error and CME arrival time/impact error. It did spur on some numerical experiments by Christina Kay year, which provided some nice info. 
https://ui.adsabs.harvard.edu/abs/2020SpWea..1802382K/abstract
https://ui.adsabs.harvard.edu/abs/2021JGRA..12628966K/abstract

Recently, a new ISWAT team was formed to explore this further https://iswat-cospar.org/H2-03

So to answer this question, collaborating with other teams will help. Alessandro also mentioned the magnetic connectivity team S2-05 which is also important https://iswat-cospar.org/S2-05 14:25:34 From Hazel Bain : Thanks @leila 14:29:17 From Mark Dierckxsens : I'll need to leave now, see you tomorrow! 14:31:50 From Leila : Hazel’s paper https://ui.adsabs.harvard.edu/abs/2021SpWea..1902670B/abstract 14:33:20 From Bill Swalwell : Thanks everyone. Hugely interesting.