From: James Chiang Date: Wed, Aug 13, 2008 at 5:33 PM Subject: Re: ASP release plans To: Steve Ritz Cc: Julie McEnery , "Peter F. Michelson" , "Cameron, Robert" , Chris Shrader , jchiang87@gmail.com Hi Everyone, I've had a few exchanges with Julie offline regarding Steve's suggestions below. So I propose we go with the following bands: 100MeV-300GeV, 300MeV-1GeV, 1GeV-300GeV plus the hardness ratio for the top two bands. Unless I hear any objections by the COB Aug 14 PT, this is what we will use. Thanks, -Jim On Tue, Aug 12, 2008 at 1:24 PM, Steve Ritz wrote: > Hi Julie, > I *think* it's OK. I agree the scientific value of the extra bins is > marginal for the extra effort, but I have a few questions: > > 1) The plan was 5 bands (plus the sum of all, for a total of 6): 100-300 > MeV, 300-1000 MeV, 1-3 GeV, 3-10 GeV, >10 GeV. If the logic is that we > don't have enough statistics to populate the bins, I can understand > combining at least the upper two bins, which is the high bin in the > proposal. Why not instead combine the upper 3 bins (>1 GeV)? The plots on > slide 8 show marginal statistics for >3 GeV. > 2) Since 100-300 MeV is still dicey, why don't we instead keep the longterm > plan at the low side, but for now hold back the lowest bin (100-300 MeV)? > That would result in two bands (300 MeV - 1 GeV, and >1 GeV) with 1 GeV as > the demarkation point. We'd probably all be more comfortable with that, and > then we can have a planned path to get to 3 bins (add one below 300 MeV when > we're ready) in a way that allows longterm comparisons back to this period. > 3) For sources above a higher flux threshold, longer term, might we > re-instate the split at the high energy (3-10 GeV, and >10 GeV), or is that > never significant? The TeV people might be very interested in a shorter > lever arm to their regime. > 4) There are factorizable errors on the fluxes: the overall normalization > (relatively large at this point) and the day-to-day relative variation. I'm > not wild about saying the fluxes are in arbitrary units, and I instead > suggest we quote an overall normalization error that is comfortable, pulled > out of the plots as an overall error. Over time, we can then just reduce > that error. That puts on a path toward continuous improvement, and allows > at least some quantitative use of the data. [The approaches are equivalent > if we were to say the normalization error is infinite, and I think we're in > better shape than that already!] > > Let me emphasize that I don't wish to derail forward progress, and I would > go along with the plan as proposed. > > Regards, > Steve > > On Aug 11, 2008, at 9:06 PM, Julie McEnery wrote: > >> Dear Peter, Steve, Chris, and Rob: >> >> I would like to ask if Jim's proposal for ASP data on monitored sources is >> acceptable to everyone. >> >> These were described in his presentation last Friday >> (http://confluence.slac.stanford.edu/x/oYMxAQ) >> >> In a nutshell for each of the monitored sources: >> Flux >100 MeV >> two energy bands: 100 MeV - 3 GeV and 3 GeV to 300 GeV >> Hardness ratio: Calculated from the energy bands >> Errors on all the above quantities. >> >> Produce all data on day and week long timescales (this will result in many >> upper limits). [alternative is to release only detections] >> >> Add a statement saying that the absolute flux calibration is not yet >> complete, so that the fluxes should be considered to be in arbitrary units. >> >> Regards, >> Julie > > -- James Chiang SLAC, MS 29 home: (650) 964-7597 GLAST ISOC 2575 Sand Hill Rd office: (650) 926-2930 Menlo Park CA 94025 FAX: (650) 926-5566