STAR Flow, Chirality and Vorticity PWG

[Sergei Voloshin <voloshin AT wayne.edu>]

Hi all,

I would agree with Fuqiang's estimate. (In terms of v^2) the nonlfow in central collision is likely around 30%. Look at the right plot in

With eta-subevents the nonflow contribution can be probably reduced, but I doubt it can be suppressed by a large factor.

A couple other comments

- I disagree that "quenching and thermalization" would reduce non-flow. As I mentioned I would probably expect a slight increase, but again this has to be studied. Without any real confirmation of either effect, we should probably preat it as unmodified. 

- Most real flow in the most peripheral ( and for sure any _real_ flow in pp) would play a role of nonflow in AA, as the "event planes" individual NN collisions are uncorrelated with the participant plane in AA.

regards,

Sergei 

On Thu, Jan 25, 2024 at 7:05 PM Wang, Fuqiang via Star-fcv-l <star-fcv-l AT lists.bnl.gov> wrote:

Hi Sooraj, 

Sorry I meant to reply-all but accidentally hit reply. I'm adding back fcv. 

Here is our publication on flow from Au+Au 200 GeV:

Phys. Rev. C 72 (2005) 14904
e-Print Archives (nucl-ex/0409033)

Fig.29 upper panel, and Table II:

Take 70-80% as all nonflow 7%, Nch=38.

Most central Nch=691.

Nonflow in central = 7%^2*38/961 = 2E-4.

v2^2 in central = 2.4%^2 = 6E-4.

That's 30% nonflow!

If half of 70-80% v2^2 is nonflow, that's still 15%!

Best regards,

Fuqiang

---

From: Sooraj Radhakrishnan <skradhakrishnan AT lbl.gov>
Sent: Thursday, January 25, 2024 6:52 PM
To: Wang, Fuqiang <fqwang AT purdue.edu>
Subject: Re: [Star-fcv-l] today's PWG discussion on nonflow in U/Au paper

 

---- External Email: Use caution with attachments, links, or sharing data ----

Hi Fuqiang,

   I was referring to the numbers from Jiangyong’s slide. Using peripheral collisions data and 1/N_ch scaling .. 

thanks,

Sooraj

Sooraj Radhakrishnan

Research Scientist,

Department of Physics

Kent State University

Kent, OH 44243

Physicist Postdoctoral Affiliate

Nuclear Science Division
Lawrence Berkeley National Lab
MS70R0319, One Cyclotron Road

Berkeley, CA 94720

Ph: 510-495-2473

Email: skradhakrishnan AT lbl.gov

On Thu, Jan 25, 2024 at 3:45 PM Wang, Fuqiang <fqwang AT purdue.edu> wrote:

Hi Sooraj,

Where did you get 1.5-4% nonflow for Hijing by 1/Nch scaling? I showed yesterday that it was ~10% in Hijing.

Best regards,

Fuqiang

---

From: Sooraj Radhakrishnan <skradhakrishnan AT lbl.gov>
Sent: Thursday, January 25, 2024 6:22 PM
To: Wang, Fuqiang <fqwang AT purdue.edu>
Cc: STAR Flow, Chirality and Vorticity PWG <star-fcv-l AT lists.bnl.gov>
Subject: Re: [Star-fcv-l] today's PWG discussion on nonflow in U/Au paper

 

---- External Email: Use caution with attachments, links, or sharing data ----

Hi Fuqiang,

   Where do I see the 30% non-flow on v_2 from AMPT in the paper? 

I was just listening to the discussion from yesterday's meeting. What I see is there are two main things to consider when using HIJING to evaluate non-flow 

1) In the discussion yesterday it was pointed out that the non-flow correlations in HIJING could be overestimated. The reason given was the estimated non-flow from using peripheral collisions and 1/N_ch scaling is much smaller than that from HIJING (HIJING also shows 1/N_ch scaling as you show in your slides). Using peripheral collisions and 1/N_ch scaling brings the non-flow estimate down to 1.5 - 4% 

2) This was what we were discussing above and also in the meeting yesterday. There are modifications to jet and mini-jet correlations during QGP evolution and also potentially to correlations from resonance decays during hadronic rescatterings later in the real data. These could likely be bringing a decorrelating/diluting effect to the non-flow correlations. If this is the case, non-flow is likely ~1 - 2%. But these effects could be evaluated in a model dependent way, could turn off hadronic transport or hydro and see the impact. 

If we want an evaluation of non-flow, a strategy could be to use peripheral events with 1/N_ch scaling as the starting point and use model dependent evaluation to see how much further the effects are reduced/changed. I think at least the first part could be done in a straightforward way and from discussions yesterday what I gather is this is not a large impact/relative value. This could then be treated as an one-sided uncertainty in the comparisons?

thanks,

Sooraj

On Thu, Jan 25, 2024 at 2:14 PM Wang, Fuqiang <fqwang AT purdue.edu> wrote:

Hi Sooraj,

Yes, those studies were for other purposes, and the nonflow is just one element in those studies. I was merely referring to the fact that AMPT central has 30% nonflow and Hijing shows 12% of the measured v2^2. 

Yes, the gamma/v2 is different from nonflow though related. 

Regarding the question if we need to take care of nonflow in all flow measurements, my take would be:

- that would be the best, with nonflow subtraction and estimated systematic uncertainty associated with that subtraction; 

- if we do not do a subtraction, indicate the magnitude of nonflow contamination, e.g., by full event and subevent or different eta gaps; 

- if we rely on comparisons to hydro models (or any other flow models) to extract the physics, e.g., eta/s or U deformation, we must correct for nonflow with adequate syst uncertainty. 

Best regards,

Fuqiang

On Jan 25, 2024, at 4:21 PM, Sooraj Radhakrishnan <skradhakrishnan AT lbl.gov> wrote:



---- External Email: Use caution with attachments, links, or sharing data ----

Hi Fuqiang,

   Thanks for the reference to the AMPT studies. I see the extraction of the non-flow estimate from AMPT on the delta gamma/v2 ratio, but I could not find a comparison between non flow estimates from HIJING and AMPT. In the paper, what I understand is that was not the focus and AMPT and HIJING are used for extracting different contributions. 

But I see from one of the references cited (https://arxiv.org/pdf/1912.00299.pdf), that in Au+Au, HIJING significantly over-estimates delta gamma/v2 than data and AMPT under-estimates (Fig.5). And in the discussions, one of the potential reasons mentioned is rescatterings resulting in weaker correlations in AMPT. But I guess this is not a fair comparison as v2 in HIJING means very different than in data or AMPT. I was thinking if a more direct comparison of the non-flow contribution between AMPT and HIJING could be made, like from the delta eta - delta phi correlations 

Yes, it is not very useful to have a discussion on which model implements correct non-flow. But my point was that since HIJING is missing a crucial part of heavy-ion collision evolution - one that can potentially alter the initial pair correlations - it would be good to check its impact. If the medium evolution has a significant impact on non-flow correlations, that would also change the conclusions on the impact of non-flow. If not, its a significant non-flow correction or estimate that we have to include in all flow measurements?

thanks,

Sooraj

On Thu, Jan 25, 2024 at 10:50 AM Wang, Fuqiang via Star-fcv-l <star-fcv-l AT lists.bnl.gov> wrote:

Hi Sooraj,

I’m not sure hijing is an overestimate of nonflow. I tend to think the opposite. 

AMPT has larger nonflow. I think AMPT is an overestimate of nonflow. We had some AMPT results on this in Phys. Rev. C 105, 024913 (2022), arXiv:2106.15595 [nucl-ex]. I think Jamie Nagle had an earlier publication reaching the same conclusion (I’ll have to dig out the reference).

In any case, we are not debating whether hijing or AMPT or other models can quantitatively describe nonflow in real data or not. The important point is that there’s nonflow contamination and the current full event and subevent estimate of 1-2% syst uncertainty is far from sufficient. One cannot simply ignore it and yet extract physics quantity based on data-hydro comparison. A more earnest assessment is needed. 

Best regards,

Fuqiang

On Jan 25, 2024, at 11:08 AM, Sooraj Radhakrishnan via Star-fcv-l <star-fcv-l AT lists.bnl.gov> wrote:



---- External Email: Use caution with attachments, links, or sharing data ----

Hi All,

   Sorry I could not join the meeting yesterday during the discussions. So if this was discussed, apologies 

HIJING could be overestimating non-flow as it does not have medium evolution/thermalization. For non-flow, the medium evolution would be contributing a decorrelating and diluting effect. 

Since AMPT uses HIJING as input, it would be good to see comparison of AMPT and HIJING. Looking at two particle delta eta - delta phi correlations should tell what the impact is 

Best,

Sooraj

On Thu, Jan 25, 2024 at 5:33 AM Jiangyong Jia via Star-fcv-l <star-fcv-l AT lists.bnl.gov> wrote:

Hi, Sergei,

1) We have provided an upper limit estimate based on 1/Nch scaling assuming most peripheral bin is all non-flow for v2 (1/Nch scaling is seen in HIJING and also roughly seen in Isobar, though latter show jet quenching effects).  For fixed pT, u and Q, <UQ*>are the same, so it is what is done in both studies.

We just say the effect is only a few percents in the ratio of UU/AuAu, since only the DIFFERENCE matters.

2) Particles after jets quenching is correlated with geometry, i.e. v2 \propto e2 as we know.

3) When jet are quenched and become medium particles. They will also flow and become correlated with participant plane. 

What we mean is experimentally you can not precisely define non-flow in central AA collisions (ALICE for Xe+Xe/Pb+Pb did not do non-flow subtraction).

4) "surface emission" is not flow.  By that we mean correlation between fragments within unmodified jets is non-flow. But of cause the leading particle from surface emission will have a non-zero v2 and is correlated with e2. 

5) hot spots" effect (mentioned in relation to 2-particle CF) is non-flow (again I could have misunderstood it) -- this is incorrect.

We agree correlation arising from hot spot at initial condition is flow.

Jiangyong

On 1/24/24 11:21 PM, Sergei Voloshin via Star-fcv-l wrote:

Hi all,

There was a rather interesting discussion at today's meeting (thanks to Fuqiang and Jianyong for preparing the presentations!), though I had to leave after 2.5 hours...  I have a few comments (not in any particular order) below.

- The discussion could be much more productive, if somebody would chair the session. It could be better if during the presentation only short questions  were allowed, and longer discussions delayed for later. Especially when it is clear that people agree to disagree (but not only).

- I agree with Fuqiang that a clear estimate of nonflow should be made, e.g.for reported v_2, but for other quantities as well.

- For the nonlfow estimate HIJING can be used, though the data might be preferable. From my point of view, using pp data would be the best if used <uQ*>, AA-pp method. One can use peripheral collisions, as tried by Jianyong, including even the scaling presented (though again much better would be <UQ*> approach). Even better would be to compare different estimates.

- The statements were made that it is not really possible to define nonflow. Very strange - what do we report then? Let us define what we want to measure, e.g. the correlation with the participant plane (another possibility would be the correlation with the reaction plane), then what is extra in e.g. measurements with two-particle correlations would be non-flow. 

-  Again, if we report v_2, let us define what we mean by that and report a realistic estimate of "non-flow" in that measurement. The values around a few percent for (ultra)central collisions look very unrealistic, to say the least.

- The paper mentions "standard" and subevent methods, but neither is defined (unless I missed that).

- The effect of nonflow for AuAu and UU collisions is likely rather similar, and the effect on the ratios should be calculated taking this into account. (This is probably what Jianyong meant by "difference" in nonflow)

- Statement that "surface emission" is not flow, is incorrect (unless I misunderstood what was said) at least for the definition of flow I used above,  but it indeed depends on the definition of flow.

- The statement that quenching reduces nonflow is (likely) incorrect, in fact it could increase it. The result of quenching is that instead of one high(er) pt particle we have several low pt. But those still have azimuthal angle strongly correlated with "jet" - then the contribution to nonflow would only increase. 

 

- "hot spots" effect (mentioned in relation to 2-particle CF) is non-flow (again I could have misunderstood it) -- this is incorrect. More exactly it again depends on the definition of flow used. If it is correlations relative to the reaction plane, the statement is correct, if it is relative to the participant plane, the hot spots effect is already included in geometry/flow plane fluctuations.

Best regards,

Sergei

-

--

======================================
 Sergei A. Voloshin    

 Distinguished Professor, 

 Department of Physics and Astronomy,

 Wayne State University,
 666 W. Hancock,  Detroit,  Michigan,  48201
 e-mail: sergei.voloshin AT wayne.edu
======================================

_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fcv-l

_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fcv-l

--

Sooraj Radhakrishnan

Research Scientist,

Department of Physics

Kent State University

Kent, OH 44243

Physicist Postdoctoral Affiliate

Nuclear Science Division
Lawrence Berkeley National Lab
MS70R0319, One Cyclotron Road

Berkeley, CA 94720

Ph: 510-495-2473

Email: skradhakrishnan AT lbl.gov

_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fcv-l

_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fcv-l

--

Sooraj Radhakrishnan

Research Scientist,

Department of Physics

Kent State University

Kent, OH 44243

Physicist Postdoctoral Affiliate

Nuclear Science Division
Lawrence Berkeley National Lab
MS70R0319, One Cyclotron Road

Berkeley, CA 94720

Ph: 510-495-2473

Email: skradhakrishnan AT lbl.gov

--

Sooraj Radhakrishnan

Research Scientist,

Department of Physics

Kent State University

Kent, OH 44243

Physicist Postdoctoral Affiliate

Nuclear Science Division
Lawrence Berkeley National Lab
MS70R0319, One Cyclotron Road

Berkeley, CA 94720

Ph: 510-495-2473

Email: skradhakrishnan AT lbl.gov

_______________________________________________
Star-fcv-l mailing list
Star-fcv-l AT lists.bnl.gov
https://lists.bnl.gov/mailman/listinfo/star-fcv-l

--

======================================
 Sergei A. Voloshin    

 Distinguished Professor, 

 Department of Physics and Astronomy,

 Wayne State University,
 666 W. Hancock,  Detroit,  Michigan,  48201
 e-mail: sergei.voloshin AT wayne.edu
======================================