Determining SusyHiggs Parameters for a Physics Rich Scenario - PowerPoint PPT Presentation

Title: Determining SusyHiggs Parameters for a Physics Rich Scenario


1
Determining Susy/Higgs Parameters for a Physics
Rich Scenario
P. Grannis LCWS Jeju Korea August 2002

We study the precision obtainable for the SM2
(SPS1) Susy scenario and a light Higgs, based on
the Snowmass SM2 Run Scenario.
update
of M. Battaglia et al.

hep-ph/0201177
2
Assumptions
SM Higgs mass of 120 GeV (or Susy Higgs h0 in
nearly decoupling limit)
Use mSUGRA benchmark Snowmass Group E2, SM2
( Allanach et al.,
hep-ph/0202233 'SPS1a'),
( Battaglia et al.
hep-ph/0106204 B ) m0 100 GeV m1/2 250
GeV tan b 10 A0 0 sgn(m) This has
relatively low mass sparticles, but the large
tanb means that there are dominant t decays that
make life difficult.

We assume 1000 fb-1 1 ab-1 luminosity
acquisition (equivalent at 500 GeV )
Year 1 2 3 4 5 6 7
(Lequivdt) 10 40 100 150 200 250 250
(fb-1)
2/22
3
SM2 sparticle masses and BRs
particle M(GeV) Final state (BR()) eR(mR) 143 c
10 e (m) 100 eL(mL) 202 c10 e(m) 45 c1
ne (nm) 34 c20e(m) 20 t1 135 c10 t
100 t2 206 c10 t 49 c1 nt 32
c20 t 19 ne (nm) 186 c10 ne (nm) 85
c1 e (m) 11 c20 ne (nm) 4 nt 185 c10
nt 86 c1 t 10 c20 nt
4 c10 96 stable c20 175 t1 t 83
eR e 8 mR m 8 c30 343 c1 W
59 c20 Z 21 c10 Z 12 c20 h 1
c10 h 2 c40 364 c1W 52 nn 17
t2t 3 c10 Z 2 c20 Z 2 c1
175 t1nt 97 c10 qq 2 c10 l n 1 c2
364 c20 W 29 c1 Z 24 l nl 18 c1
h 15 nl l 8 c10 W 6





















































3/22
4
Run Plan for SM2 Susy sparticle masses
Substantial initial 500 GeV run (for end point
mass determinations). Scans at some thresholds
to improve masses. Special e-e- run and a run
above 500 GeV.
Beams Energy Polztn Ldt (Ldt)equiv
comments ee- 500 L/R 335 335 sit
at top energy for end point measurements ee-
MZ L/R 10 45 calibrate with Zs
ee- 270 L/R 100
185 scan thresholds c10c20 (L pol.) t1t1
(R pol.) ee- 285 R 50 85 scan
mRmR- threshold ee- 350 L/R
40 60 scan tt thresh scan eReL
thresh (L R pol.) scan c1c1-
thresh. (L pol.) ee- 410 L 60 75
scan t2t2 thrsh (L pol) scan mLmL thrsh (L
pol) ee- 580 L/R 90 120
sit above c1c2- thresh. for c2 end pt.
mass e-e- 285 RR 10 95 scan with
e-e- for eR mass


















S(Ldt)equiv 1000 fb-1
4/22
5
Initial (end point) mass determinations
The traditional end point method


For A ? B C
E 1/2 (1b ) (1 - mA2/mB2) b (s/4mA2
-1)
dN dEC
(A B are sparticles C is observed SM
particle). Measuring 2 end points gives both A
and B masses. Statistics, backgrounds,
resolutions smear the edges.
½
E- E

• Making an box distribution mass measurement
  requires
• A given final state ( e- polarization) should be
  fed by only 1 dominant reaction
• Two body decay with C a stable observable SM
  particle.
• Neither of these conditions are generally true
  for benchmark SM2 with large BRs into ts and nt

However, it is not necessary to have a box
distribution for determining mass any known
distribution will do. But if there are not sharp
edges, the precision is lower. (Recall that the
top quark mass was measured to within 4 in
semileptonic decays with a broad mass
distribution (using templates) with only about 40
events and S/B 2.
5/22
6
The reaction overlap problem
Among all-leptonic ( missing energy) decays of
sparticle pairs, tt is the dominant final state.
It is fed by 9 different sparticle pair
reactions ! (and moreover the taus are not
stable, so the end points of the observed final
state (1 prong p , r ) are washed out.
c1 c1
c10 c20
c1 c1
c10 c20
ee- (left)? tt- 152K evnts
ee- (right)? tt- 52K evnts
6/22
7
A new look at end points in SM2
Examine all final states involving 2, 4 or 6
leptons plus missing energy (with no hadrons in
final state). These should be low background
from SM sources, and relatively free of
cross-talk due to misidentification of leptons Do
the spreadsheet for the contributing reactions to
each channel more completely than before. Keep
the sub-reactions distinct e.g. ? c10 e
has different end points from ? c20 e and
must be treated separately. Assume no SM
backgrounds Begin to look at mass determinations
for cases without box distributions. Coupled
channel analyses fitting several distributions
with several unknown masses will be needed There
are many cross-checks get a mass from a
dominant channel, but can check it in subdominant
channels. channel specific final state
(e.g. eett) reaction specific 2 body
process (e.g. ee- ? c10 c20 )
7/22
8
So, how to get initial sparticle masses ?
start with the easier cases
smuonR ee-(right pol)? ? mm-
missing energy Both ? c10 m , so use
either m as observable. Determine both and
c10 masses from end points. Susy background is
5 In 335 fb-1, find dM( ) 0.077 GeV

dM(c10) 0.11 GeV
smuL smuL
chi10 chi20
smuR smuR
eeR- ? mm- E 30.7K evnts
smuonL ee-(left pol)? ? (c10 m)
(nm c1-) ? (c10 m) (nmntt- c10) ? mt-
missing energy ( cc) ? c10 m (45) , with m
as observable. Susy bknd is 5 In 335 fb-1,
find dM( ) 0.70 GeV
(dM(c10) 1.9 GeV )
smuL smuL
eeL- ? mt E 3.9K evnts
Mass precisions scaled from Colorado group
Snowmass01 analyses.
8/22
9
selectrons L R
4 distinct coupled reactions analyze them
together
ee-(left pol) ? / /
/
? ee- missing energy Both and
? c10 e . Colorado group has analyzed the
coupled channels using double differences between
e and e- for L and R polarization. Determine
, and c10 masses from end points.
Background is 5 (left Pol), 0 (right Pol) In
335 fb-1, find dM( ) 0.19 GeV dM( )
0.27 GeV dM(c10)
0.13 GeV


eR
eR
selL selL-
chi10 chi20
chi10 chi20
selR selL-
selL selL-
selL selR-
selR selL-
selR selR-
selR selR-
eeR- ? ee- E 210K evnts
eeL- ? ee- E 62.7K evnts
selL selR-
9/22
10
neutralino1 LSP
Several reactions have dominant decays to c10
from combination of just the ee and mm final
states (dominated by selectron pair and smuon
pair), we estimate dM(c10) 0.08 GeV Adding
the channels et, mt, eett, eeee, all of which
have a dominant reaction with a primary decay to
c10 I guess that the precision would be dM(c10)
0.06 GeV
10/22
11
The harder t channels
ee- (left)? tt- 152K evnts
c1 c1
c10 c20
ee-(left pol)? c1 c1- ? ( nt) ( nt) ?
(t nt c10) (t- nt c10) ?
tt- missing energy 64 gt These ts tend
to be back to back and ee-(left pol)? c10 c20 ?
c10 ( t) ? c10 (t- c10 t)
? tt- missing energy 19
gt These ts tend to be collinear ee-(left
pol)?stau1 stau1 ? (t c10) (t- c10)
? tt-
missing energy 8 gt ts back to back
Can assume c10 mass is well measured, but here
c1, c20, masses are all to be determined
in this e eL- ? tt channel, as well as with
eeR- ? tt, eeL- ? mmtt,
eeL- ? mttt, eeL- ? tttt
11/22
12
tt channel comments
Opening angle distribution of the 1 prongs from t
can partially distinguish between the c1 c1- and
c10 c20 reactions. Making a cut (Qopen lt
p/2) increases the fraction of c10 c20 by a
factor of 2 while retaining 73 of c10 c20
c1 c1-
c10 c20
Qopen
Qopen
One can fit the observed 1-prong energy
distribution to a template to get a particular
mass.
c2
all reactions in tt final state
Allowing just M(stau1) to vary, get M134.88
0.22 GeV. (M134.89 input)
M(stau1)
1 prong energy
12/22
13
tt channel comments
Can do better than use 1-prong energy e.g.
larger of the two 1-prong energies
Or with the good calorimeter, see the p0 and can
use the r (p0 p ) energy for the dominant case
of t ? r nt These more sharply peaked
distributions offer better mass determination.
NEEDS a proper study, but I am guessing that the
c1 , c20 and masses can be found to 1
GeV, good enough to fix the energy for scans.
13/22
14
stau2
does not dominate any channel besides the
6t final state for which there are only 262
evnts (L pol) or 93 evnts (R pol) (before t
BRs). 6 of tt E (L pol) 152K events total 6
of tt E (R pol) 52K events 2 of eett E (L
pol) 25K events 3 of mmtt E (L pol) 8.6K
events 6 of mmtt E (R pol) 1.5K events 8 of
tttt E (L pol) 35K events 20 of tttt E (R pol)
4.8K events Thus we would use the selectronL/R
and smuonL/R masses and the measured stau1 mass
to estimate the stau2 mass (model dependent) for
a subsequent energy scan. Nevertheless, since
stau2 contributes to many reactions, there is a
least a good cross-check of the mass estimate!
14/22
15
Higher mass gauginos
The c30 is special as it has decays c30 ? c10 Z
(12) and c30 ? c20 Z (21) with Z ? ee/mm The
cross section at 500 GeV for eeR- ? c10 c30 is
16 fb. Taking into account the Z BRs , we
estimate that using the Z as an end point
particle (we scale from a Colorado group
measurement of c2 ? c1 Z ) dM(c30) 8.5 GeV
(statistics are limited but bknd negligible)
c40 The c10 c40 threshold is 460 GeV, but the
event rates are too small to allow a
measurement. c2 Threshold for ee- ? c1
c2- is 539 GeV. Do special run at 580 GeV,
trading luminosity for energy. Decays c2 ?
c1 Z (Z ? ee/mm) give 55 events, allowing
dM(c2) 4 GeV
15/22
16
sneutrinos
eeL- ? ? (c1 e- ) (c1- e) ?
ee-tt- E is 15 of eett final state (25K
total events major contributors are selectron
pairs and c20 c20 pairs. eeL- ? ?
(c1 e- ) (c20 ne) ? emmt E is 39 of emmt final
state (628 total events). The rest are from
selectron L. eeL- ? ? (c1 e- ) (c20
ne) ? ettt E is 39 of ettt final state (6.5K
total events). The rest are from selectron
L. Can these be dug out? If one knows the
selectron and c10 masses precisely, one should be
able to estimate the snue mass to a few GeV?
and These never come close to dominating
any final state seem very tough to get end
point masses for these !
NEEDS A STUDY!
16/22
17
Threshold scans for sparticle masses
Martyn Blair (hep-ph/9910416) studied the mass
precision available from scans near two-body
thresholds (Tesla point RR1). For p-wave
threshold (gaugino pairs), s b1, while for
s-wave (sfermion pairs), s b3.
Martyn-Blair used 10 points perhaps not
optimal. Strategy should depend on events,
d(sBR)/sBR, backgrounds and b-dependence.
Mizukoshi et al. (hep-ph/0107216)
studied ne,nm,nt thresholds (low sBR and large t
decays) and found that 2 points on the rise and
one well above threshold was better. Blair at
Snowmass found that 2-point scans could be
optimal for dm and G (Benchmark SPS1a) can get
dG/G 30 for typical sparticles). Cahn
(Snowmass) did analytic study of mass precision
from scans vs N pts, spaced at DE and
found With L total scan luminosity and su
XS at upper end of scan. Good agreement with
MC results. Little improvement for Ngt3,
particularly for p-wave.



0.36 vN
DE v18 L su
(1 )
0.38 vN
(1 )
DE N-1/4 v2.6 L su
dm
dm
(p-wave)
(s-wave)
17/22
18
Threshold scans
One needs to allocate scans carefully there is
a trade off between luminosity at 500 GeV (all
end points and searches) and use of lower energy
(at reduced luminosity). Do only those scans
that give the most restrictive information on
Susy model parameters. (In SM2, get some scans
for free as as thresholds overlap.)
Feng Peskin (hep-ph/0105100) study showed that
e- e- operation (both beams R polarized) at the
eReR threshold (b1) could give substantially
better dm(eR) than the e e- scan (b3), even
after inclusion of beamsstrahlung. We adopt
this idea in our run plan.



With DEbm beamstrahlung Dm(eR) 0.1 GeV

In establishing the mass precisions from scans,
we have scaled the dms from existing studies by
the ratio of assumed vs(500 GeV)Lt . (Probably
naïve to ignore details of backgrounds at
different benchmarks, and the effect of uncertain
sBRs.) (Used only dominant reaction/polarization,
so is conservative)

• Note that for scans, we need not identify
  particular exclusive decays -- the total visible
  cross section may be used. But beware
  overlapping thresholds!

18/22
19
Sparticle mass precision
sparticle dMEP dMTH dMCOMB
(end pt) (scan)
(combined) eR 0.19 0.02
0.02 GeV eL 0.27
0.30 0.20 mR 0.08 0.13
0.07 mL 0.70 0.76
0.51 t1 1 2 0.64
0.64 t2 --
1.1 1.1 ne 1
-- 1 nm 7 ??
-- 7 ?? nt --
-- -- c10 0.07
-- 0.07 c20 1 2
0.12 0.12 c30 8.5
-- 8.5 c40 --
-- -- c1
1 - 2 0.18 0.18 c2
4 -- 4

For run plan indicated for SM2














19/22
20
mSUGRA parameter determination
The ultimate aim of the Susy program at the LC is
to determine the character of the Susy breaking
(GMSB, mSUGRA, AMSB cMSB, NMSSM, etc.), and
illuminate the physics at the unification scale.
This will require measurements of the sparticle
masses, cross-sections and branching ratios,
mixing angles and CP violating observables. A
start on this has been made G. Blair, et al. PRD
D63, 017703 (01) S.Y. Choi
et al., hep-ph/0108117, G. Kane, hep-ph/0008190.
Here we ask the more restricted question
Assuming we live in mSUGRA (as for benchmark
SM2), what are the Susy parameter errors ?
Mass resolutions quoted for our Run Plan give

• dm0 mainly from eR, mR masses
• dm1/2 mainly from c1 , c2 masses
• dA0 mainly from t1, t2 masses
• dtanb mainly from c1 , c10 masses
• Conservative, since additional info from t, H/A,
  sL/R will give added constraints on mSUGRA
  parameters

Parameter SM2 m0 (GeV) 1000.08 m1/2
(GeV) 2500.20 A0 (GeV) 013 tanb
100.47






20/22
21
Higgs, top quark parameter errors
Scale the errors fromTESLA TDR Snowmass Orange
Book
Relative errors on Higgs parameters (in
) parameter error parameter
error MHiggs 0.03 GTot 7 s(ZH)
3 lZZH 1 s(WW) 3 lWWH
1 BR(bb) 2 lbbH 2 BR(cc) 8
lccH 4 BR(tt) 5 lttH
2 BR(gg) 5 lttH 30
Errors on top quark parameters Mtop 150
MeV (0.09) Gtop 70 MeV (7)
Systematics limited
21/22
22
Conclusinos

• Even for the physics rich scenarios of Susy
  benchmarks SM2 and low Higgs mass, the Linear
  Collider can do an good job on precision
  measurements in a reasonable time.
• Runs at the highest energy should dominate the
  run plan -- to optimize searches for new
  phenomena, and to get sparticle masses from
  kinematic end points.
• The details of the run plan depend critically
  on the exact Susy model -- there is large
  variation as models or model parameters vary.
  It will be a challenge to understand the data
  from LHC and LC well enough to sort out sparticle
  masses/cross sections and predict the appropriate
  threshold energies.
• For Susy, it remains very likely that higher
  energy will be needed to complete the mass
  determination and fix the Susy breaking mechanism.

23
(No Transcript)