Did you all know there is a paper on the Sept 10, 2017 solar particle event on the ISS (which is seen in the Environmental Data App and RadLab Rodent Research-9 data spike)? I did not. Curious if anyone has any thoughts, responses, ideas.
https://doi.org/10.1029/2018SW001920
@kirill @svcostes @j_miller @RLWG @xmao @jgong @lauren.sanders @mdelp @anarayan09
Yeah, that’s a cool paper. How’d you run across it? All the authors have been and/or will be involved in acquiring data for RadLab. In particular, Thomas Berger is providing DOSTEL data.
Some thoughts/responses/ideas:
Those are some of the first data I worked on for GL, back in 2018.
There were no active radiation monitors in the US Lab in that period, so I first compared data from Columbus from the RR-9 dates with data from US Lab from a few months prior. The average doses were comparable so I used the US Lab data from June-July 2017, on the theory that it was preferable to use data from the module where the animals were located. But this assumes that doses on the ISS are relatively constant. Over weeks or even several months they are, but on a time scale of hours (considering passages through the SAA), years (the solar cycle), or days (if there’s an SPE) they aren’t. So the original RR-9 dose report missed the SPE. Now in practice this doesn’t matter for ISS payloads, because the doses are still low–the 2017 SPE spike didn’t noticeably affect the RR-9 average compared to other RR missions–but it’s still something that the investigators should be aware of. (For example suppose the crew had noticed the rodents freaking out on 9/10/17…) Of course it’s up to the investigators to determine whether a short duration spike at low dose matters on the time scale of how radiation affects their organism, or at least ask for expert help. In the ISS radiation environment that’s unlikely, but BLEO it may well not be.
The take home message is that when archiving time-dependent data one needs to consider the optimal time interval. We’ve discussed this in the RadLab team: many of the detectors provide rather fine grained data–5 min or even 1 min in some cases–and we need to give thought about what delta_t to use. For ISS data we have at least daily doses for most detectors, and we should report it. Back in 2018 I wasn’t doing that. (What can I say–I was young then…) RadLab does. If one searches on Sept. 2017, this is the result:
Jack
Wow, these are fascinating points. At the next RadLab AWG mtg @RLWG , I’d love to hear more on this flushed out. Plus to compare EDA & RadLab
Nicholas Stoffle (now at Axiom), just joined the AWG today (not yet on the Forum), based upon an exciting radiation RadLab/Directed Acyclic Graph project sorta started/stumbled upon at the @HUMANawg through a fantastic presentation by @padilla - on different DAG approach, including Bayesian network probabilities
I also met Nicholas for the first time last week, so perused his publications
Pinging few others on this topic @robert.j.reynolds-2 @eantonse @Dr.Overbey @kirill @svcostes
Here is RR9 in EDA:
The accumulated dose plot illustrates my point about the SPE not having a big effect on dose: there’s just a slight kink on Sept. 10.
On a related note, in OSDR, the two tabs “Environmental Data” and “Environmental Data App” under Data&Tools could be confusing to users looking for radiation data, as both provide radiation data in slightly different formats. I think it’d be good to reconcile the two, maybe incorporating a pointer to RadLab as well. The “Environmental Data” tab points to the old GeneLab site which has some useful summaries and definitions but also some outdated tables. In particular, Sylvain and I agreed to do away with the GCR vs SAA breakdown, which is a relic of the days when almost all our data were coming from SRAG. That distinction is specific to SRAG data and is not useful for biology. Also, in the RR table, having only the daily average and the total misses transient peaks due to SPE. If that table is to be retained it should be modified to combine GCR and SAA, and add maximum and minimum daily values for the mission, as we did for the Shuttle data. (Speaking of which the STS_radiation_data link points to the RR table rather than the Shuttle data; that needs to be fixed.)
The distinction between between GCR and SAA is useful for model validation. You may not want to actually get rid of it unless there’s a cost to maintain. I think there’s still useful things that could be done to validate physics models in LEO although I don’t know if there are any activities currently ongoing.
Good point Steve. Our motivation was to simplify the db for biology users who (through GeneLab) were the original intended audience and who are principally interested in total dose, but we do want RadLab to be useful to the wider community. Some of the non-SRAG detectors have more detail than shown, as well. I’ll be interested to hear what other users think.
@j_miller of note, thanks for pointing out that mis-directing link for Shuttle.
It now has been fixed. Thanks @danielle.lopez91 & @lovorkad
