physics (“status” and “plans”)

P. SphicasPhysics (status and plans) Sep 25, 2008

Physics Coordination Meeting1

Physics (“status” and “plans”)

Outline Introduction/reminder of original plan And now? New meeting structure (very rough) plan of work for the next six

months

P.SphicasCMS Collaboration Meeting

Sep 25, 2008

Reminder (of what we were planning

up until Friday Sep 19, 2008)

P. SphicasPhysics (status and plans)

Making a physics plan Main characteristics of fall 2008:

Dominated by detector startup. Need to increase involvement and attention (from “physics”) in detector commissioning matters (and from the detector community in physics).

Physics startup. Focus must be on electrons, muons, jets, missing Et, p-flow, tau… not on Higgs to two-photons

Communication: need to make links between the detector and physics groups stronger, more effective and faster

The central theme has to be (obviously) “detector performance and early physics”

Explicit Goals: we have to get out some early results: First event pictures; First publicity plots Some very early papers on 10 TeV Measure trigger rates, exercise data flow, analysis teams…

The key: vertical integration & clear physics goals

Sep 25, 2008Physics Coordination Meeting

3


Vertical Integration and Physics Goals We thus need to tackle two things:

The formation of effective, efficient groups working on “detector performance and physics object commissioning”

We need to have the “vertical integration meetings” as the norm not as “special events”

The formation of a number of analyses with clear physics goals (i.e. measurements)

We need to have the “CSA08 analyses” as the norm not as “proof-of-principle/debugging” exercises

Practical steps: Need to reform (complete overhaul in fact) of the current

meeting structure Identify explicitly the early publications of CMS (along with the

publicity plots) and launch the work on these – directly


4


How to continue The direction we are moving:

Factorize the work leading to a physics result from CMS into two pieces:

Joint DPG-Physics meetings starting in end September (after CMS week).

The output of these are the physics objects of CMS. Electrons, jets, muons, etc.

Given these objects, explicit analyses (e.g. W cross section measurement) under the responsibility of the corresponding Physics Analysis group (EWK in this case)

PAGs will (have already…) launch(ed) explicit activity on each “paper”


5

And now?


From the press release…


7

Incident in LHC sector 34. Geneva, 20 September 2008. During commissioning (without beam) of the final LHC sector

(sector 34) at high current for operation at 5 TeV, an incident occurred at mid-day on Friday 19 September resulting in a large helium leak into the tunnel. Preliminary investigations indicate that the most likely cause of the problem was a faulty electrical connection between two magnets, which probably melted at high current leading to mechanical failure. CERN 's strict safety regulations ensured that at no time was there any risk to people.

A full investigation is underway, but it is already clear that the sector will have to be warmed up for repairs to take place. This implies a minimum of two months down time for LHC operation. For the same fault, not uncommon in a normally conducting machine, the repair time would be a matter of days.

Further details will be made available as soon as they are known.


New situation By now we know that the delay of “at least two

months” has been translated to “no beam in 2008” Obviously, the loss is not just the Fall of 2008 Physics-wise, in addition to losing “Fall08”, we also lose all the

analysis work that we had planned for Jan-May We need to redefine our program of work with an effective

delay of 6-9 months Our entire planning was pegged on the assumption that

collisions were imminent We would get a few days at 900 GeV And we would be accumulating ~5-10 pb-1 at 10 TeV this fall

Thus the plans for Winter 2009 And the early papers And the halt to Monte-Carlo based approvals


8


What next (I)

Sep 25, 2008

It’s clearly too early at this point to know precisely what the machine will do

Presumably, they will be asking us (the experiments) at some point what we prefer: X pb-1 at 10 TeV “in the pocket” or Y pb-1 at 14 TeV (presumably X>Y)

So we need to prepare to answer these questions Presumably, our (CMS) position is that we want data, as much

data as possible (even if it is at 9.8 TeV) Imagine (just imagine) how much more we would know in

four weeks from now if we had seen 900 GeV collisions this past weekend



What next (II)

Sep 25, 2008

In physics, we had two activities and one grand project right before Friday

The 900 GeV analyses – the samples (min bias) were ready The 10 TeV analyses – in the form of the “early papers” The new meeting structure (with the joint DPG-PhysicsGroup

meetings as the highlight) We can and should continue the 900 GeV exercise

There is no reason to change it in any way The 10 TeV data: we should turn to these samples as

soon as possible To repeat the analyses we have already done at 14 TeV To gain experience with the new software and settings

And we must go on and implement what is fundamentally a really good way forward: the joint DPG-POG meetings


(very rough) plan for the next six months


Highlights Continue the 900 GeV analyses – to completion

Some nice results shown in the QCD meeting this week Start analyses with the (soon arriving) 10 TeV samples;

assume this is what we will run with Recall: they have 3.8 T, 10 TeV and new tracker format

(backwards incompatible) Continue work on early publications

we should have full drafts ahead of time What about new analysis approvals?

Given that we have a few months now ahead, we need to determine which results we want to update and which results we want to make public (afresh)

Need to tend to needs of younger collaborators working on theses and limited-term positions


12


New results

Sep 25, 2008

New approvals: Higgs: our currently approved results date back to the P-TDR.

We should update them for the winter conferences (with Split 2008 we finish with current confs)

Aim at a full collection of new results on WW, ZZ, … Top: our current results come from the special TOP2008

approvals. They are incomplete and do not represent the current status quo.

We should update this set of results SUSY: we need to put out some key results – e.g. updates on

the all-hadronic search, using the new data-driven ideas. Ditto for the early search with leptons.

Loose ends from every group (e.g. analyses that were stopped short in order to turn to the data, e.g. in Exotica, some late-comers in QCD, etc)

Goes without saying: we will apply the established (high) standards. All analyses are to be done as if we had data (the “real thing”)



Feeding the analyses (I)

Sep 25, 2008

We need to plan a fastsim production Assume 10 TeV, 50pb-1 like before. As soon as fastsim is ready

(say in ~1 month). We’ll probably go with CMSSW_2_1 Assume that fullsim samples (200 Mevt) are expendable

Whenever genuine new knowledge of the detector is included (e.g. CaloTaskForce, pixels, etc) we just push the button for another 200 Mevt

Question has come up: should we plan another challenge, e.g. CSA09?

The answer is probably “no”. It will be much more productive – and will help us prepare for data-taking: if we can run computing operations continuously

So we can afford to run the 200 Mevt for the “new knowledge”

For now: include signal samples in the 200 Mevt production



Feeding the analyses (II)

Sep 25, 2008

Consolidate our state: there are three other parallel activities related to “tools” and the “how-to”

We now have a few more months than before We should REALLY deploy the PAT throughout all groups

We have to co-organize the “feedback loops” that are currently taking too long:

Releasevalidationbugfindbufixre-release Special samples and turn-around

We also need to check on the “analysis turn-around” The large fastsim samples can be used for much of this

Should also consider (possible) exercises with the 200 Mevt fullsim as well

We need to complete the primary dataset exercise


New meeting structure


“Vertical Integration” Like all else, the discussion becomes one of meetings

Joint meetings on detector performance and physics (objects) Our meeting structure should be one centered on the

detector, its performance and the commissioning of physics objects

Change the current POG meetings into joint DPG-POG meetings, focused on all the “performance” issues

Supplement with ONE big plenary for the collaboration We cannot add meetings; We have to hold them in the

afternoon (GVA) Detectors will clearly need a forum to discuss the startup and

all that will occur in early data days. So target is the physics side. After some thought: concluded that we have to overhaul the complete meeting structure and adapt it to data-taking era


17

P. SphicasPhysics (status and plans) Sep 25, 2008


New meeting structure Broad-brush description:

All meetings bi-weekly; only afternoons, 14:30-18:30 (to allow for two meeting slots)

Mon afternoon: joint DPG-Physics and Trigger-DPG-PH The agenda of each joint meeting will be agreed upon between

POG and DPG conveners; Meetings run by POG conveners. Tue afternoon: all physics meetings Wed afternoon: plenary; “Detector Performance and Physics” Thu and Fri: free for parallel meetings. E.g. for subgroups,

working groups. Smaller groups: they can meet ad hoc (any time of the day or night)

Note “plenary investment” is minimal: one afternoon for DPG-physics and one afternoon for physics

Concentrating all physics meetings on a single day (Tue) makes scheduling of other activities easy(ier). Ditto for detector performance meetings. Plenary is major synchronization point.


Architecture/layout


19


Monday


20


Tuesday


21


Wednesday


22

But we start with biweekly plenary for now…


Summary

Sep 25, 2008

We have to consolidate our gains from the work of the last few months and stay on track:

We will launch the new joint meetings and the new meeting structure beginning next week

We should turn to the new MC samples asap Continue 900 GeV analyses – to completion Start 10 TeV analyses; assume this is what we will run with Work on early publications: want full drafts ahead of time

New goals approvals (with caution): Higgs, Top, SUSY. In all other cases, we should discuss things

before we launch the analyses (for approval) Need to work on: PAT (this is a must); Feedback loops;

Analysis Turnaround; Primary datasets And need to tell the machine to play it safe


physics (“status” and “plans”)

Documents

physics startup

physics coordination

physics goalswe

physics objects of cms

physics groups stronger

physics object commissioningwe

joint dpgphysics meetings

detector performance