OPT-IN: Electromagnetic decay of the Σ*0 to Îγ

[Elton Smith]

Hi Dustin and Ken,

I have several comments on the original draft. Please refer to those page numbers.

Overall the decay width is a nice result and should start constraining models. 

General notation: 

1) You seem to use Sigma0(1385) and Sigma*0 interchangeably. You should certainly use a single notation, as Sigma*0 sometimes seems to imply higher lying states. I suggest you consistently use Sigma0(1385), including in the title to be specific.

2) You use 'BR' and 'R' for branching ratio (e.g. eqs 10, 11 vs Table III. Chose one (I suggest Br) and use consistently.

p. 2 right column first paragraph. 2% energy resolution -- what happened to the e-counters?

p. 2-3 Event Selection.

Here you make essentially no mention of tracking, but later in the kinematic fitting section you talk about the covariant matrix V. We all assume this is the covariant matrix derived from tracking reconstruction.  

I suggest that in this section you explicitly mention that events are selected with 2 positive tracks (one satisfying the kaon hypothesis and one satisfying the proton hypothesis) and 1 negative track (satisfying the pion hypothesis). [I might have missed it, but although implied, I don't think this is stated explicitly]. A good place to insert this would be preceding the last paragraph left column p 3.
You can follow this clarification  to state that the covariant matrix is obtained for each track and what the matrix includes in anticipation of the discussion to kinematic fitting.

p. 3 left column last paragraph:

For this paragraph to be understood, you need the preceding statement about selecting 2 pos and 1 neg tracks. My understanding of the discussion is that Fig 1 contains the entire sample of events selected so far (not just gp->p pi+pi-(X), which is what the text and the figure caption says). The text and caption must then clarify that events in Fig 1 above the cut are further selected as a cleaner sample of gp->pK+pi- events).

p. 4 Kinematic Fitting.

This section is premature and should be moved later to around the middle of the section on 'analysis procedure' end of page 5.
The discussion of constraint equations (top p. 5 left)  should be married with the actual constraints used (eq. 5).

p. 5 Eq 4

where are the unmeasured variables u on the left of the equation (in L or F) and which are the unmeasured variables for a at least one of the cases?

p 5 middle first paragraph:

The following sentence is confusing to me: "The chi2 minimization occurs by differentiating the chi2 with respect to each of the variables, while linearizing the constraint equations and obtaining improved measured values from the fit". Usually the linearization of equations is simply an approximation tool to allow inversion of the matrix equations. If this is the case, simply remove that phase to avoid confusion. If the sentence implies more than this, it needs to be expanded/explained.

p. 5 paragraph before simulations.

As suggested earlier, move this to the discussion of event selection.

p. 5 left bottom, analysis procedure.

I don't understand the phrase "double bremsstrahlung". Are you referring to the accidental coincidence between to interactions in the target? If so, use the word accidental and specify which two interactions are in time coincidence.

p.5-8 Analysis procedure.

This section should be broken down into two sections: One on extracting the number of  signal events (via kinematic fitting), i.e. n_gamma and n_pi0, and one on the estimation of background reactions (could be called 'Backgrounds'). 

In the section on extracting the number of signal events, you need to summarize early on what you plan to do. I believe the key paragraph is 2nd paragraph right column on p. 6. It begins by stating that "A two-step kinematic fitting procedure is used." I would add to this "...to significantly reduce the large pi0 background before attempting to extract the number of radiative events." You don't have to spend a page "proving" that this is the optimum (or required) procedure. 

Some specifics:

p. 5 second paragraph right column:

In this paragraph, it sounds like you are performing a preliminary kinematic fit to constrain the p and pi- to the Lambda mass. I assume this is part of the kinematic procedure which is described subsequently, so its description needs to be moved down after the two reactions are listed and before the sentence on "the constraint equations are".

p.6 Eq 6

Is this a parameterization to account for a) non-Gaussian errors b) backgrounds in the event sample, both? What is the expectation for a "pure" sample? Is the P0 term  for backgrounds?

p. 6 right column, middle paragraph

The "recovery uncertainty" is not clear. Is this simply a trade off between enhancing signal and minimizing background? If so, this may be an easier way to explain it.

p. 7 Fig 9

Remove zero suppression in fig a).

'New' Background section would start around bottom of p. 7, starting with "Table II lists..."


This section would be helped considerably with a TABLE containing the estimated number of events coming from each of the channels studied. The rows would be similar to the ones in Table II with the addition of the signal events gamma and pi0. But it would have an additional column which is the number of evens in the sample. This will help the reader immediately see what the largest subtractions are. By the way, these numbers may be in the text, but the non-expert (me) cannot find them.

p. 9 fig 12 caption second line "...off of..."?

p. 10 left column second paragraph

States that invariant mass cut has no effect because of the kinematic constraint. However, there is a selection cut in Fig 3 with a range that might affect the ratio.

p. 10 Eq. 17, comments/ questions

Immediately following states that the large statistical uncertainty comes from the uncertainty in the full width. I assume this means that the error on the Gamma=36+/-5 was added in quadrature with the statistical error. This is not correct, as most of the uncertainty in the '5' is systematic [The reference Aguilar has 5722 events, or 1.3% statistics]. I also appears that the 5 is added both to the statistical and systematic uncertainties, resulting in 16% errors. I'm not sure what the optimum procedure is, but I'm confused about the present analysis.

Also note: Given the accuracy of the measurements all decimals should be dropped and numbers rounded.

p. 11 Table IV.

What are the errors in the lower and upper values? It seems to me that only the central values are used-- which is what I would do. But then the errors are simply confusing because one can ask if one should include them in the limits. I suggest omitting the +/- errors. One could mention in the caption that the lower and upper values were determined with a precision of about +/-0.12.