[Clascomment] OPT-IN: A search for baryon- and lepton-number violating decays of Lambda hyperons using the CLAS detector at Jeff erson Laboratory

Wed Jun 3 17:05:02 EDT 2015

Dear Daniel,
	Thanks much for these *very* helpful comments.  Matt and I have discussed them, and we feel that all of your suggestions will make for a stronger presentation of our analysis.  It’s going to take a few days for us to make the changes, but we’ll post a new version of the paper when it’s ready.
	I’ll attempt to answer your question regarding the number of MC events.  There are a few features to note.  First, since we are not calculating a differential cross section or similar quantity, we are not as concerned with extrapolating into regions of the detector where the acceptance is small.  Rather, we’re performing a bulk search over the “entirety” of the detector and kinematics.  We’ve made an effort to match the MC to the physics of the gamma p —> K+ Lambda reaction and to the g11 photon energy distribution; the acceptance comparison( global calculation vs binned calculation) for the standard model decay suggests that the match is adequate.  
	Because we’re only quoting final results to one significant digit, it’s unlikely that increasing the number of MC events for the BNV channels will have any effect on the results. In addition, we calculate and propagate the uncertainty from the number of MC events, so even if it could be made smaller, we see that it's not the limiting factor on the uncertainties/upper-limits.  It might also be worth noting that the Lambda itself is not terribly forward going, so we can expect less of an issue with *its* decay products going into regions of small acceptance.
	The other analysis step in which the MC plays an important role is in the optimization of cuts using the Punzi figure of merit (FOM).  The two-dimensional histograms of the figure of merit (see fig. 5a) suggest that the limiting factor is the number of data events that pass the cuts on MM^2 and MM(K+).  There is a jagged/dramatic difference in FOM near the optimal bin, transitioning from red to orange or yellow as the cuts widen.  This suggests that there are one or two data events that are right on the cusp of passing cuts of these widths, an effect consistent with what we see in the signal identification histogram (fig. 5b).  
	There is also an underlying structure in fig. 5a (and similar) that is due to the cuts’ efficiency in rejecting MC.  This variation is smooth (it’s a bit difficult to tell with this full spectrum shading), which means that we have enough MC to determine an optimal cut (i.e., a small change in cut width does not cause a large change in efficiency).
	The combination of this smooth variation of the FOM with the correspondence of MC to the K+Lambda kinematics leads us to believe that we have enough MC.

	I think that the only other comment that required a response was that on the photon energy spectrum.  The relevant range of the brem. spectrum is from K+Lambda threshold (E =  0.909 GeV) up to the maximum tagged photon energy (3.86 GeV).  I’ll add this to the text.

	We’ll work to make the other corrections.  Thanks again!
	
Best regards,
Mike


-------------------------------------------
Michael E McCracken, Ph.D.
Physics Department
Washington & Jefferson College
Washington, PA 15301
724.223.6148
-------------------------------------------


> On Jun 3, 2015, at 11:32 AM, Daniel Carman <carman at jlab.org> wrote:
> 
> Dear Mike and Matt,
> 
> I have read through your draft of the "rare" Lambda decay branching ratios based on the draft
> of May 26. I have enjoyed watching this analysis proceed from your Working Group presentations,
> especially as this is definitely a bit of a novel idea and not one of our "usual" CLAS papers.
> My comments are included below. If you have any questions, let me know.
> 
> 
> 					        Regards,
> 						
> 						Daniel
> 
> ************************************************************************************
> Physics/Presentation Comment.
> - The title of this paper is "... baryon- and lepton-number violating decays". The "BNV" aspects
>   are quantified and discussed, but the "LNV" aspects are not mentioned after the first page.
>   In my reading I found this lack of discussion of lepton-number violation problematic given the
>   title and the introduction. I think this needs some attention both in the analysis discussion 
>   and in the summary/conclusions section.
> 
> General.
> - Your color figures should have "(Color Online)" in the caption.
> - You have tense issues throughout. What I mean is that the analysis steps tend in your manuscript
>   to be discussed in the present tense when the past tense is most appropriate. For example, you
>   use "For this analysis we make use of the CLAS drift chambers ..." when it is more appropriate
>   to use "For this analysis we made use of the CLAS drift chambers ...". The tense issues began
>   in Section II and continued through the analysis/results section.
> 
> Page 1.
> - Abstract. Line 2. I suggest "... using the CLAS detector at Jefferson Laboratory."
> - Line 14. I think you mean "conception" not "inception".
> - Line 18. Here you use Refs.[4,5] (publication dates in 2004 and 1976) as data that motivated
>   Sakharov's work in Ref.[6] published in 1967. This violates causality.
> - Line 26. Use "... interactions that violate ...".
> - Line 48. Use "... calculations that constrain ...".
> - Line 79. Use "... \bar{\Lambda}$), which then undergoes ...".
> 
> Page 2.
> - Line 135. Use "... decays, the expected backgrounds ...".
> - Line 148. Use "... by CEBAF (Continuous ...".
> - Line 160. Use "4 m".
> - Line 166. Use "... can be found in [28]." If there are other references that are relevant for the reader
>   beyond [28], list them explicitly.
> 
> Page 3.
> - Line 183. 3e6 MC events seems too small for a reasonable study by a factor of 10 to 50 in my opinion for
>   such a large phase space. Why such a small number?
> - Line 183. Use "... \Lambda$ events, and weighted ...".
> - Line 191. Use "... model the CLAS acceptance, ...".
> - Line 213. I suggest "... through the magnetic field of CLAS) ...".
> - Line 216. I suggest "... $d$ is the path length of the track from the vertex to the TOF system, ...".
> - Eq.(3). Use a period after this equation, not a comma.
> - Line 228. Use "... Monte Carlo events is shown ...".
> 
> Page 4.
> - Line 237. Use "... hyperons that include ...".
> - Line 239. Use "... apply geometrical fiducial ...".
> - Line 257. For your number $N_{rec}$, what is the relevant photon energy range for the analysis? Seems like
>   this should be listed here with the mention that it is a brem spectrum.
> - Line 266. Use "... and kinematics, we separate ...".
> - Line 281. Use "... MM^2$, as well as ...".
> 
> Page 5.
> - Line 288. Use "... for a potentially small ...".
> - Line 300. Use "We have tuned our cuts ...".
> - Line 302. Use "... plots that would ...".
> - Line 305. Use "... of all of our data points until ... of the analysis cuts, ...".
> - Line 311. Use "... to the measured ...".
> - Line 326. I suggest "... order as the dimensions of CLAS and ...".
> 
> Page 6.
> - Fig. 4. Your x-axis labels overlap the axis values. Also the axis values at the ends of the plots
>   overlap and are partially cut off.
> - Fig. 4 caption. Line 1. Use "... boundaries of the PID ..."
> - Eq.(10). Wrong units.
> - Line 396. I suggest "... there are significant numbers of ...".
> 
> Page 7.
> - Fig. 5 caption. Line 4. Use "... of the optimal cuts, ...".
> 
> Page 8.
> - Line 447. Missing units on $w_2$.
> - Line 450. Use "... to identify the signal.".
> - Line 456. Missing units on $w_2$.
> - Line 462. Use "... of the background in the signal ...".
> - Line 469. 5e5 again seems way too low of a number of generated events.
> 
> References:
> - Problem with formatting of nearly all references as there are no spaces between the different
>   parts of the journal names (e.g. Phys.Rev.Lett. vs. Phys. Rev. Lett.).
> - I would have liked to have seen the inclusion of a recent reference for the M/AM ratio based on
>   data.
> - I would have liked to have seen a more recent reference instead of [7] which is from 1985. Surely
>   there is something from your lifetime.
> - For published papers, do not include the preprint numbers.
> - Ref.[20]. With a long author list, use first author and then "et al.".
> - Ref.[21] has some kind of problem. Who is S.-K. Collaboration?
> - Ref.[22] is incomplete.
> - Ref.[27] is not a CLAS Collaboration paper. First author is "B.A. Mecking".
> - Ref.[28]. First author only and then "et al.".
> - Ref.[30]. I have no idea what eConf C030908 MODT002 is. This reads like some kooky hexadecimal code!
> 
> _______________________________________________
> Clascomment mailing list
> Clascomment at jlab.org
> https://mailman.jlab.org/mailman/listinfo/clascomment