sPHENIX is a new detector at RHIC.

[Anthony Frawley <afrawley AT fsu.edu>]

Hello Craig,

Regarding the Upsilon case specifically, I can point out that we did not use fiducial cuts on the PHENIX EMCal for J/psi measurements.

But I take your point. We should compare the maximum eta value of the descoped EMCal with the original maximum eta of 1.1.

Thanks

Tony

---

From: Craig Woody <woody AT bnl.gov>
Sent: Tuesday, November 7, 2017 7:59 AM
To: Marzia Rosati
Cc: Anthony Frawley; sphenix-l AT lists.bnl.gov; Gunther Roland
Subject: Re: Question about upsilon statistics in descoping document

 

Hi Marzia, Tony, Gunter and All,
   I'd just like to clarify the point that I was trying to make in my original message to Marzia. In the descoping document, it claims a loss in upsilon statistics of 25% for the |eta| < 0.85 scenario compared with the "nominal" configuration, which I take to be |eta|<1.0. However, this is comparing the nominal acceptance with a fiducial cut to the reduced acceptance without a fiducial cut. In our baseline design, we are building both calorimeters out to |eta|=1.1 with the assumption that we would be using the region of |eta|< 1.0 for our physics analysis. This 0.1 eta cut for the fiducial region was just a guess we made a number of years ago, which was mainly for jet analysis, and I'm sure it will need to be refined with both Monte Carlo and with real data when the time comes. It is also presumably a cut that applies mainly to hadronic jets and not necessarily for electrons from upsilon decays, which may allow us to make a smaller fiducial cut in the EMCAL (i.e., allowing larger acceptance) for electrons near the edge of EMCAL where we don't care about hadronic leakage.
   Nevertheless, the point remains the same. If we're proposing to only build the physical EMCAL out to eta|<0.85, we will need to make some fiducial cut around the edge for any physics analysis. For upsilons, it may be closer to the edge (maybe only a few cm), but if that's the case, then we could have measured out to close to eta = 1.1 in the baseline design. In either case, we should be comparing the fiducial acceptance in the descoped design to the fiducial acceptance in the baseline design.
   I think it's also worth pointing out that aside from the statistics for the upsilon, the region at the edge of the EMCAL is going to give a very non-uniform response for jet measurements. We already have a very complicated three compartment, depth segmented calorimeter (well, maybe only two compartments now with the IHCAL gone...), and the hadronic response is going to be very different near the edges of the reduced acceptance EMCAL. I think for clean jet measurements, we'll probably want to restrict ourselves to something like|eta|< 0.75 where the hadronic response is at least uniform in depth. For events near the edges of the EMCAL, there will be hadronic showers starting in the EMCAL which will have substantial side leakage, as well as some very unusual e/pi response which I have a hard time imagining. I have to admit that I haven't been following all of the recent HCAL simulations, but hopefully someone is looking at this.
   Anyhow, I think the bottom line is that relative to our baseline design, the loss in upsilon statistics with a reduced physical acceptance the EMCAL of |eta|<0.85 will be larger than 25% when we compare the fiducial regions in both cases (whatever that would be). Also, just for the record, at no time was I asking anything about cutting the EMCAL acceptance to |eta|<0.6.

Cheers,
Craig     

On 11/6/2017 8:58 PM, Marzia Rosati wrote:

Craig and sPHENIX list,
since this is a general interest question I am responding to the list.

The plot I sent to Gunther was generated with Pythia Upsilon(1S) particles decaying into dielectrons, so the Upsilon and electrons have a realistic pt distribution and account for the decay kinematics.

I made simple acceptance cuts on the two electrons from Upsilon decays as stated in the legend: |eta|<1, |eta|<0.85, |eta|<0.6. The reduced eta coverage implies a Upsilon signal loss of 24% for |eta|<0.85 and 59% loss for |eta|<0.6, relative to eta<1.

If the actual detector coverage is different than what stated in the Figures, I can easily generate new numbers but those were the eta values requested by Gunther and Dave. I assume the |eta|<0.72 coverage will imply an acceptance loss very close to the half way value in between the 0.6 and 0.85 values i.e.  40-45% loss.

If I follow your logic, I assume the 0.6 descoping scenario would result in slightly different coverage too. I can generate new plots if needed we just have to settle on the detector plans and actual coverage for various descoping scenario first....

Best regards
Marzia

On 11/6/17 6:11 AM, Craig Woody wrote:

Hi Marzia,
   I was looking at the descoping document that they turned in and there is something in there that bothers me. It says we need to cut down the EMCAL acceptance from eta = 1.1 to eta = 0.85 and that this would lead to a reduction of 25% in upsilon statistics (see p.6), which I presume is based on the plot you made shown in Fig. 2.4 . I know the 0.85 number came from Ed who just scaled the total cost of the EMCAL to save the $1.1M that was needed to get within the budget limit. I did a more accurate estimate by looking at the items we could actually cut, keeping the fixed costs the same, and I came with a number of 0.82. This was surprisingly close to the scaling estimate, but the 0.82 (which is probably more like 0.8 in reality) needs to have a fiducial cut applied for doing any sort of physics analysis, since we can't measure right to the edge of the detector. That's why the baseline design goes out to eta =1.1 in order to measure within eta = 1.0. Therefore, you need to reduce the 0.82 to 0.72 when comparing to the statistics that we would measure out to eta = 1.0. If you look at Fig 2.4 in the descoping document, the loss in statistics for this is clearly more than 25%. Take Pt = 4 GeV for the upsilon, and the acceptance for eta = 1 is about 0.32. It looks like the acceptance for eta ~ 0.7 would be about 0.17, which would give a ratio to eta = 1.0 of 0.53, or almost 50% loss in statistics. This ratio is more or less constant across the Pt spectrum, so it's pretty clear to me that we're going to loose a lot more than 25% in upsilon statistics. Am I missing something here, or do these losses seem to be underestimated ?

Thanks,
Craig

P.S. I'm also copying Jin on this in case he has any comments.



On 10/23/2017 8:57 PM, Marzia Rosati wrote:

Gunther et al.,
attached is a the upsilon efficiency vs pt for all 3 eta ranges to be used as the right hand side plot in figure 1.4.

I am not sure what is meant by line 124-126. "While these losses are not catastrophic, it is worth noting that within the sPHENIX run plan a 25% loss is equal to the statistics collected in one of the RHIC run periods".

As far as I understand this statement is not correct. The acceptance in PHENIX is about 18 times smaller than sPHENIX so we our stats are much smaller. STAR has a similar acceptance to sPHENIX but readout much slower and limited PiD so I don't think they ever collected upsilon samples in any beam species (pp AuAu) which are 1/4 of the expected sPHENIX stats.

Best regards
Marzia

On 10/23/17 5:54 PM, Gunther M Roland wrote:

A friendly reminder that comments on this important document are still being accepted! Thanks to everyone who shared their thoughts/corrections already. We will work on an updated draft tomorrow and then take it from there.

Best,

Gunther and Dave

On Oct 21, 2017, at 12:02 AM, Gunther M Roland <rolandg AT mit.edu> wrote:


Friends,

As discussed at the General Meeting today, we are forwarding draft 1 of our document outlining the detector scope for a $32M cost cap. The pdf file can be found at
https://www.dropbox.com/s/5wyutbndogozm5q/sPH-GEN-2017-002_v1.pdf?dl=0
(we'll provide another link tomorrow for those that can't access dropbox)

Please send your comments in reply to this mail, keeping the [sPH-GEN-2017-002] tag in the subject line***. Comments received by close-of-business on Monday, 10/23, will be most useful.

Cheers,

Gunther and Dave

***we will move future reviews to an sphenix-notes-l AT bnl.gov list, but the list couldn't be generated in time for this note.
_______________________________________________
sPHENIX-l mailing list
sPHENIX-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-l

_______________________________________________
sPHENIX-l mailing list
sPHENIX-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-l
.

_______________________________________________
sPHENIX-l mailing list
sPHENIX-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/sphenix-l