[Olympus] Acceptance Simulations

Thu Nov 22 11:42:56 EST 2012

Dear Michael,

I see. Thanks for the explanation. 

Maybe what I wrote was partially misleading. Of course, I know that MC studies will be the key to understand most systematic uncertainties. I just personally think that neither the gaps between the TOFs nor their real particle efficiency will drive the uncertainty. In particular, if you consider the ideal case where they are symmetrically arranged. I'm just afraid that it may send the wrong message if we were to publish results of a toy study with input which cannot (and should better not) be true. I have never seen a 2cm gap between scintillators and I still worry about what should cause strongly inefficient scintillators. In both cases the actual fact will not change your result, only the level of precision to which you know it. I'm afraid, I have no good idea how to determine the TOF efficiency even roughly without using full tracking information and lots of statistics.  

So, my concern is two-fold:
1) You may waste precious time as the two-fold ratio of a small uncertainty will be very small, indeed. At least for the gaps I cannot see how this could possibly cause a large effect, in particular, if you leave everything symmetric.
2) If you are not careful, you may risk leaving the impression that we have left 2cm gaps between scintillators and they would work at 60% efficiency.

My very crude idea if you are speaking of a "toy experiment" simulating a possible scenario would read along the following lines: Assume that we are inefficient to a certain level in finding a track in a certain sector of Olympus and you misjudge your actual efficiency by a certain level. This would be independent on what the cause is and should be left-right asymmetric. One could define it for fiducial volumes or even simpler for angles (but not very realistic). If you then put in a 5, 10% e+/e- asymmetry what would the reconstructed asymmetry be? As I said, just an idea....

This is, of course, only my view and I hope that such thoughts help though I have no time to work on the details and will be drowned in night shifts when I'll be at the experiment soon.

Cheers,
Inti

On 22 Nov 2012, at 16:19, Michael Kohl wrote:

> Dear Inti,
> in the four-fold (double) ratio, acceptance and efficiency effects drop 
> out to first order, in the single ratio they don't. They do partially if 
> the same detector elements are hit. Back-of-the-envelope estimates are not 
> quite possible due to large acceptance, toroidal field and extended 
> target. We need to assess for now how big the systematic effects on e+/e- 
> can be in the worst case, and the only way is to use MC. This shall serve 
> as a basis to determine what control measurements should still be done 
> before the run ends.
> 
> As Doug pointed out, the analysis of the single ratio will ultimately 
> require very accurate accounting of the real detector properties, and 
> still it is not quite clear how well imperfections have to be known (not 
> only their size but also their uncertainty), in order to achieve the goal 
> of 1% systematics for e+/e-, hence not only the size of the effects on 
> e+/e- but also the sensitivities of e+/e- to uncertainties of geometric 
> details must be studied with MC. 
> In addition, there has not been established an accurate method to 
> determine and monitor the TOF efficiencies in situ. Understanding of the 
> WC is of course very important, too.
> 
> Please be invited (and other people, too!) to get engaged in this process.
> 
> Best regards
>   Michael
> 
> 
> On Thu, 22 Nov 2012, Inti Lehmann wrote:
> 
>> Hi Axel and Michael,
>> 
>> I'm lost now: What are you looking for or trying to prove, Michael?
>> 
>> In the corners of the TOFs we have overlap. In the flat sections we have small gaps of possibly 2mm. Everybody can do a back-of-the-envelope calculation to realise that the loss of acceptance due to this gaps will be at the 1% level. It is the worst in small angles as the gaps are most parallel to particle tracks.The gaps may be slightly different from place to place, but there is nothing one can learn with MC about a reality which is not modelled.
>> 
>> Though I repeat myself, I say it again. Any scintillator of this size which exhibits an efficiency below 99% is simply broken! If I were you I would not show Joachim Mnich or anyone else plots which insinuate that you expect scintillators with lower efficiency. Anyone would tell you to go and replace them rather than discussing systematic effects or changing the field.
>> 
>> My feeling is that the only systematic effect due to the TOFs is due to the different e+ e- acceptance at the forward end. This can probably modelled rather well. Efficiency effects may be serious but I'm quite sure they will be driven by the WC. Independently on the cause one could just study scenarios with changing efficiencies in different regions. If I were you I would at doing that over the WC active area rather than the TOFs.
>> 
>> Hope to have helped!
>> 
>> Cheers,
>> Inti
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> On 21 Nov 2012, at 05:18, Michael Kohl wrote:
>> 
>>> Hi Axel,
>>> thank you.
>>> Now these double ratios look like they were done with the white generator?
>>> However, the latest plot file "tof_gaps.pdf" in my possession for the
>>> single-polarity ratios still has the large errors at large angles. Could
>>> you update them, too?
>>> And add a plot for the yields themselves (from which the ratios are
>>> formed, maybe all four with different colors in one panel, i.e. three
>>> panels for 0, 1, and 2cm gaps).
>>> 
>>> Also it would be good for any of the ratios to zoom into the y range to
>>> 1.0 +-10% in order to see the deviations from one in more detail.
>>> 
>>> So it looks like there are some good washout effects at work, in
>>> particular if the efficiency structures are smaller than the "washout
>>> scale", which could likely be of order of the extended target length. What
>>> would be interesting next is to study along the lines of Stanislav's
>>> proposal to see what happens when entire TOF bars or combinations of TOF
>>> bars are less efficient than others. Here, the benefit
>>> from washing out structures may be less so, because the localized
>>> efficiency structures would be of similar order like the "washout scale".
>>> Similarly, this could be the case for inefficient cells in the WC. How
>>> much bigger are the corrections for single ratios than for the fourfold
>>> ratios?
>>> 
>>> The list of things to look at can obviously be extended for many aspects.
>>> What we shall think about is, and this was requested by J. Mnich, what is
>>> the best set of measurements other than B+ running that should be done
>>> toward the end of the OLYMPUS run that can eventually help us verifying
>>> and tuning the MC, in order to achieve better than 1% systematics
>>> everywhere.
>>> 
>>> Thank you,
>>>  Michael
>>> 
>>> 
>>> 
>>> On Tue, 20 Nov 2012, Axel Schmidt wrote:
>>> 
>>>> Hi Michael,
>>>> 	I'm attaching the plot of the acceptance double ratio with all four configurations as was initially proposed.  One can see that even with gaps in the ToFs, the ratio remains very close to 1.
>>>> 
>>>> 	I have done some studies with different magnetic field strengths (all with positive polarity).  None of the results were so different than those at 5000 A, so I did not find them so intriguing.
>>>> 
>>>> Cheers,
>>>> Axel
>>>> 
>>>> 
>>> 
>>> +---------------------------------------------------------------------
>>> | Dr. Michael Kohl, Assistant Professor and Staff Research Scientist
>>> | Physics Department, Hampton University, Hampton, VA 23668
>>> | Jefferson Lab, C117, 12000 Jefferson Avenue, Newport News, VA 23606
>>> | Phone: +1-757-727-5153 (HU), +1-757-269-7343 (Jlab)
>>> | Fax:   +1-757-728-6910 (HU), +1-757-269-7363 (Jlab)
>>> | Email: kohlm at jlab.org
>>> | Cell:  +1-757-256-5122 (USA)
>>> |
>>> | Deutsches Elektronen-Synchrotron DESY, Bd. 66, Rm. 6,
>>> | Phone: +49-40-8998-6406, Cell: +49-171-101-1967
>>> +---------------------------------------------------------------------
>>> _______________________________________________
>>> Olympus mailing list
>>> Olympus at mit.edu
>>> http://mailman.mit.edu/mailman/listinfo/olympus
>> 
>> 
> 
> +---------------------------------------------------------------------
> | Dr. Michael Kohl, Assistant Professor and Staff Research Scientist
> | Physics Department, Hampton University, Hampton, VA 23668
> | Jefferson Lab, C117, 12000 Jefferson Avenue, Newport News, VA 23606
> | Phone: +1-757-727-5153 (HU), +1-757-269-7343 (Jlab)
> | Fax:   +1-757-728-6910 (HU), +1-757-269-7363 (Jlab)
> | Email: kohlm at jlab.org
> | Cell:  +1-757-256-5122 (USA)
> |
> | Deutsches Elektronen-Synchrotron DESY, Bd. 66, Rm. 6,
> | Phone: +49-40-8998-6406, Cell: +49-171-101-1967
> +---------------------------------------------------------------------