[Clascomment] OPT-IN: Near Threshold Neutral Pion Electroproduction at High Momentum Transfers and Generalized Form Factors

[Stepan Stepanyan]

Hi Punit,

It is clear (as I stated in the question) that these two vertical bands 
are on top of pre-
and post-radiative value of \Delta\theta, but it is not clear why you 
name them "extension"
when there is clear disconnect. Then can you explain they appear at such 
high values of MM2.

As for the answer to the second question, I do not think you got the 
point here. You assumed
that after your cuts, still some BH events are left in the spectrum and 
you use 2 Gaussian fit to
extract the number pi0 events from the MM2 distributions. What I am 
arguing is that due to the
MM2 dependence on \Delta\theta, in particular shifting of the MM2 
towards smaller values with
increase of \Delta\theta, in the projection you see a shoulder on the 
left of the main peak, which
you attributed to BH contribution. It might still be some BH are left 
there, but attributing the whole
shoulder to BH events and fitting it with a Gaussian function and 
subtract it is incorrect in my mind.
Looking through your plots on MM2 (with fits) it looks to me the 
systematics for this procedure is
more close to 8-12%. Unfortunately I can read fitted values of 
parameters from the plots, but from
two example plots you have in the paper I can see that peak of the "pi0" 
Gaussian is bellow the
expected nominal value, while it looks like peak of the second Gaussian, 
which according the claim
is BH, is above the 0, instead of being below 0.

In short, to be constructive, I think that 6% systematics must go up to 
10%.

Regards, Stepan

On 11/13/12 1:33 PM, Puneet Khetarpal wrote:
> Hi Stepan,
>
> Please see my responses to your questions below:
>
> On Nov 12, 2012, at 4:03 PM, Stepan Stepanyan <stepanya at jlab.org> wrote:
>
>> Hello Punit et al.,
>>
>> I have a question on figure 8 and cuts applied.
>> First simple questions just for understanding, what are these two vertical bands above the red line that are at the pre- and post-radiative \Delata\theta.
> -- The two vertical bands above the red line are the extensions of the pre and post-radiative tails.
>
>> The second is more serious in my mind, your "pi0" band (horizontal) has a \Delta\theta dependence. How is the MM2 dependence on other kinematic parameters, like proton momentum, or electron angle etc. I could not find analysis note to see if there are any studies on MM@ dependence on kinematics.
>
> The proton kinematics in the CM frame is complementary to the pion as they are produced back to back (Fig. 1). The electron information is encoded in Q^2 and W, the kinematics over which we have reported the results. We performed the necessary
> corrections for the electron and proton angle and momentum. The results after the correction can be seen in the following plots:
>
> http://clasweb.jlab.org/rungroups/piprod/wiki/index.php/Corrections
>
>> What worries me is the after using elliptical and linear cuts, the projections of MM2 does not look truly Gaussian and you assumed that it is due to reaming BH events. It might be that there are still some BH events left there, but when you have kinematic dependence of MM2, it never can be treated as a Gaussian and fitted as is to get the number of events. This will boil down to what the systematic uncertainty will be due to this type of extraction. I want to know what the contribution of the second Gaussian is in the spectrum relative to your extracted pi0 number, and I think that should be quoted as a systematic uncertainty.
>> Another related question, if there are kinematic dependences of MM2, than some of parameters are not measured (corrected) well. So, what are the level of uncertainties in [absolute value of] measured parameters. Since measurements are close to threshold, small changes in kinematics may bring  big changes in cross section.
>>
>> Stepan
> The MM^2 dependence plots for W and Q2 and cos theta of the pion in CM bins can be seen here:
>
> http://www.jlab.org/Hall-B/secure/e1-6/puneetk/semi-inclusive-epx-all/feb10/mm2_epx_fit_1.png
>
> http://www.jlab.org/Hall-B/secure/e1-6/puneetk/semi-inclusive-epx-all/dec09/ct_wbin_3.png
> (please ignore the black curves in this plot)
>
> Of course, the MM2 is not truly Gaussian and there may indeed be some BH events present after subtraction, which is why we have a systematic uncertainty of +-6% (on average) assigned to this
> procedure. This procedure includes the difference in the cross section and structure functions after applying this Gaussian background subtraction and without applying this Gaussian background subtraction. As such, the possibility of contamination is appropriately accounted for.
>
> Thank you,
> Puneet
>
> http://about.me/puneetk
>
> "No word is true until it is eaten" - 36 Lessons of Vivec, Sermon 27
>