Progress on FTK Studies FTK Physics Case Meeting 07/20/05 Erik Brubaker University of Chicago.

Progress on FTK Studies

FTK Physics Case Meeting

07/20/05

Erik Brubaker

University of Chicago

July 20, 2005 FTK Physics Case Meeting 2

Reminder: Purpose• Making the physics case for FTK means studying

signal and background trigger rates for several processes.

• ATLAS full simulation is very slow.

• Solution: use fully simulated events to parameterize the L1 jet and tau trigger response vs ATLFAST jet pT. Then all the various processes for physics studies can be studied using ATLFAST.

July 20, 2005 FTK Physics Case Meeting 3

Upgrade to 10.3.0—Why?• Bug causing energy to disappear in L1 fixed.

– Caused second band in response plot below.– Affected low-pt jets? Not clear…

– L1 ET > 1 TeV also a bug, not fixed yet, but minor effect.

• Rerun jobs anyway to loosen thresholds on jet, tau triggers.– Allow parameterization of

response w/o arbitrary cutoff.– Allow investigation of tau

trigger iso cuts.

July 20, 2005 FTK Physics Case Meeting 4

Upgrade to 10.3.0—How?• One of the first to use 10.3.0 for running many

jobs in parallel.• This is not necessarily a good idea… ATLAS

software releases X.Y.Z with Y=0 are production releases. With Y>0 are development releases.

• Besides patches found on the TWiki, I needed– IOVDbSvc-00-02-18-COOL-1-1-0

• Otherwise ~50 simultaneous jobs overwhelm the ATLAS database server.

– AtlfastAlgs-00-01-84• Otherwise random crashes occur in Atlfast.

July 20, 2005 FTK Physics Case Meeting 5

Upgrade to 10.3.0—Results• Second band gone

from jet response plot:

• Interesting tau triggers to look at.

J2 only J1-J8

?

July 20, 2005 FTK Physics Case Meeting 6

Response parameterization—Old• Problems with previous parameterization as

shown last time by Catalin.– Discontinuities: Pathological behavior such as

Landau0, etc.

– Odd behavior atlow pT, due toLandau.

July 20, 2005 FTK Physics Case Meeting 7

Response parameterization—New• Improved

functional form: tried & true double Gaussian.– No pathological

behaviors– Second Gaussian

can take care of high tails (?)

July 20, 2005 FTK Physics Case Meeting 8

Response parameterization—New• Trigger turn-on curves vs “truth” from

simulation.

Still used inthe same wayas last time.

Updated macrofile to be sentafter the mtg.

July 20, 2005 FTK Physics Case Meeting 9

Response parameterization—taus• Tau energy response

not as Gaussian as jets.– Not a surprise

• Need to decide how to treat iso cut!– Given iso cuts,

parameterize energy response.

– Or parameterize also iso response (harder).

– Need to study iso cuts for tau triggers?

July 20, 2005 FTK Physics Case Meeting 10

Double Higgs—Process• Resonant production of

lightest MSSM neutral scalar Higgs h0 through the heavier H0.

• Previous study (ATL-PHYS-97-104)– For MH=300 GeV and tan(=3, S/sqrt(B) of 5.7 can

be reached using 300 fb-1; this requires 4 jets with ET>40 GeV and ||<2.5.

– For lower integrated luminosity, more tricks are needed and jets down to 20 GeV are preferred.

– They concluded trigger rates probably prohibitive.

July 20, 2005 FTK Physics Case Meeting 11

Double Higgs—Signal MC• Initial intel was that double Higgs production not

implemented in Pythia.– Seems to be generally true, but this specific process is

available. Developing…

• Instructions available for modifying Pythia to include the relevant MSSM Higgs processes, if needed.– Claim is that it’s very complicated avoid if

possible!

July 20, 2005 FTK Physics Case Meeting 12

Double Higgs—Background MC• RomeBackgroundWiki: Multijet sample for

Rome production cancelled due to “insufficient resources”.

• Trying to find out details of what was planned, will produce this at UChicago Tier2.– ME for 3- or 4-jet production?– Or just Pythia 22 + parton shower with generator

cuts to ensure additional hard partons?– This will be a useful sample to have available!

• For now, look at dijet samples…

July 20, 2005 FTK Physics Case Meeting 13

Double Higgs—Background rates• Use J1-J8 dijet samples.• Require at least 4 ATLFAST jets inside ||<2.8.• Find the probability for the event to have four jets

passing a 40 GeV trigger.

Sample J1 J2 J3 J4

pT range 17-35 35-70 70-140 140-280

N Events 16527 19851 19807 892

N Passed 1e-9 3e-4 30.2 49.4

Fraction 7e-14 1.6e-8 1.5e-3 6e-2

July 20, 2005 FTK Physics Case Meeting 14

Future plans• Keep up with improvements in L1 trigger

simulation as needed (nothing now).

• Improve parameterizations as needed.– Regions where tails must be well described.– Tau param. w/ different iso cuts? As function of iso?– Event pass function.

• Get real signal, bkgd MC for double Higgs, continue study.

• Expand into other processes as appropriate.