Jets in the standard model and beyond the standard model
Gavin Salam
CERN, Princeton & LPTHE/CNRS (Paris)
PANIC, MIT, 28 July 2011 Parton fragmentation
Gluon emission: dE dθ quark αs ≫ 1 Z E θ
At low scales:
αs → 1
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 2 / 25 Parton fragmentation
Gluon emission: dE dθ quark αs ≫ 1 Z E θ θ At low scales: gluon αs → 1
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 2 / 25 Parton fragmentation
Gluon emission: dE dθ quark αs ≫ 1 Z E θ
At low scales:
αs → 1
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 2 / 25 Parton fragmentation
Gluon emission: dE dθ quark αs ≫ 1 Z E θ
At low scales:
αs → 1 hadronisation non−perturbative
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 2 / 25 Parton fragmentation
Gluon emission: π+ dE dθ quark αs ≫ 1 Z E θ KL π0 At low scales: K+ − αs → 1
hadronisation π non−perturbative
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 2 / 25 Parton fragmentation
Gluon emission: π+ dE dθ quark αs ≫ 1 Z E θ KL π0 At low scales: K+ − αs → 1
hadronisation π non−perturbative
This is a jet
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 2 / 25
Jets as projections
π π p φ K
LO partons NLO partons parton shower hadron level
Jet Def n Jet Def n Jet Def n Jet Def n jet 1 jet 2 jet 1 jet 2 jet 1 jet 2 jet 1 jet 2
Projection to jets provides “common” view of different event levels But projection is not unique: we must define what we mean by a jet
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 4 / 25 This talk
1. How are jets defined? 2. What pp physics is being done with jets? 3. What’s the forefront of research into jet techniques? 4. How about jets in heavy-ion collisions?
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 5 / 25 1. How are jets defined?
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 6 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance 1 diB = 2 pti
1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a single particle
3. If it’s a diB call i a jet and remove it from list 4. Update all distances, go to step 1
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.0001 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.000197997 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.000424056 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.000487076 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.000573632 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.00063427 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.000646091 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is diB = 0.000523303 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.00467798 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is diB = 0.003136 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is dij = 0.0130612 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance dmin is diB = 0.013971 p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV 1 t diB = 2 40 pti
30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] The anti-kt jet algorithm
Define “distance” between every pair of particles: [Cacciari, GPS & Soyez ’08] 2 1 ∆Rij 2 2 2 dij = ∆R = (yi − yj ) + (φi − φ ) max(p2 , p2 ) R2 ij j ti tj Define a single-particle distance p /GeV The1 algorithm involvest two parameters: diB = 2 40 p1.ti R, the angular reach for the jets
2. A pt threshold for the final jets to be considered relevant 30 1. Find the smallest of dij and diB
2. If it’s a dij , merge i and j into a 20 single particle
3. If it’s a diB call i a jet and remove it from list 10 4. Update all distances, go to step 1 0 0 1 2 3 4 y
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 7 / 25 [1. Defining jets] Jet contours – visualised
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 8 / 25 2. Standard use of jets in pp collisions?
As a stand-in for a final-state parton, where all you care about is the parton’s energy and direction.
Partons (≃jets) are useful objects, e.g. to
a) find out how many partons you have in the proton to start with
b) look for new physics that couples to partons (typically with large cross sections, allowing probes of highest scales)
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 9 / 25 [2. Jets in pp physics] Jets from scattering of partons Tevatron results Largest source of jets is simply 1010 K D=0.7 QCD scattering of incoming par- T CDF data ( L = 1.0 fb-1 ) 7 tons 10 Systematic uncertainties NLO: JETRAD CTEQ6.1M
jet 4 corrected to hadron level µ µ JET µ 10 R = F = max pT / 2 = 0 PDF uncertainties [nb/(GeV/c)] 10 JET T
p JET 6 dp 10-2 |y |<0.1 (× 10 ) JET
-5 JET 10 0.1<|y |<0.7 (× 103) / dy
σ JET p 2 -8 0.7<|y |<1.1
d 10
1.1<|yJET|<1.6 (× 10-3) 10-11 jet 1.6<|yJET|<2.1 (× 10-6) 10-14 0 100 200 300 400 500 600 700 JET pT [GeV/c] Jet cross section: data and theory agree over many orders of magnitude ⇔ probe of underlying interaction
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 10 / 25 [2. Jets in pp physics] Jets from scattering of partons CMS results Largest source of jets is simply QCD scattering of incoming par- tons jet
p
p
jet
Jet cross section: data and theory agree over many orders of magnitude ⇔ probe of underlying interaction
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 10 / 25 [2. Jets in pp physics] Jets from scattering of partons ATLAS results Largest source of jets is simply QCD scattering of incoming par- tons jet
p
p
jet
Jet cross section: data and theory agree over many orders of magnitude ⇔ probe of underlying interaction and also of proton structure
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 10 / 25 [2. Jets in pp physics] Dijet resonance searches
Because of strong coupling of partons (and abundance in initial state), jets serve to probe the highest energy scales reachable at hadron colliders. jet q
q X p q p
q
jet
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 11 / 25 [2. Jets in pp physics] Dijet resonance searches
Because of strong coupling of partons (and abundance in initial state), jets serve to probe the highest energy scales reachable at hadron colliders.
q
q X p q p
q
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 11 / 25 [2. Jets in pp physics] Dijet resonance searches
Because of strong coupling of partons (and abundance in initial state), jets serve to probe the highest energy scales reachable at hadron colliders.
q
q X p q p
q
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 11 / 25 [2. Jets in pp physics] Dijet resonance searches
Because of strong coupling of partons (and abundance in initial state), jets serve to probe the highest energy scales reachable at hadron colliders.
q
q X p q p
q
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 11 / 25 3. New Jet Techniques
When a jet contains more than one parton
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 12 / 25 [3. Jets research] Boosted massive particles, e.g.: EW bosons
Normal analyses: two quarks from X → qq¯ reconstructed as two jets
X at rest X jet 1
jet 2
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 13 / 25 [3. Jets research] Boosted massive particles, e.g.: EW bosons
Normal analyses: two quarks from High-pt regime: EW object X X → qq¯ reconstructed as two jets is boosted, decay is collimated, qq¯ both in same jet
X at rest z single X jet 1 boosted X jet (1−z) jet 2 Happens for pt 2m/R pt 320 GeV for m = mW , R =0.5
As LHC starts to explore far above EW scale, such configurations become ubiquitous
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 13 / 25 [3. Jets research] 3 key ingredients
◮ Resolving the underlying 1 → 2 splitting and using its characteristic kinematics to help reject background [Leading-order Structure]
◮ Exploiting different colour structures of signal and background and resulting different energy flows [Higher-order structure]
◮ Protecting jet-mass resolution from the mess of underlying event and pileup [Non-perturbative structure]
There’s no way I can do justice to this highly active field in just a few slides
See parallel sessions and recent Boost 2011 conference
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 14 / 25 [3. Jets research] [Undoing clustering] Ingredient 1: undo jet clustering & cut on pieces
First proposed for W’s by Seymour ’93 Refined by Butter- worth, Cox & Forshaw ’02 Refined more + showed how to use it to find H → bb¯ at LHC, Butterworth, Davison, Rubin & GPS ’08 Later in ’08: extended to top quarks by ATLAS; Thaler & Wang; Kaplan, Rehermann, Schwartz & Tweedie. [3. Jets research] [Undoing clustering] Ingredient 1: undo jet clustering & cut on pieces
First proposed for W’s by Seymour ’93 Refined by Butter- worth, Cox & Forshaw ’02 Refined more + showed how to use it to find H → bb¯ at LHC, Butterworth, Davison, Rubin & GPS ’08 Later in ’08: extended to top quarks by ATLAS; Thaler & Wang; Kaplan, Rehermann, Schwartz & Tweedie. [3. Jets research] [Undoing clustering] Ingredient 1: undo jet clustering & cut on pieces
First proposed for W’s by Seymour ’93 Refined by Butter- worth, Cox & Forshaw ’02 Refined more + showed how to use it to find H → bb¯ at LHC, Butterworth, Davison, Rubin & GPS ’08 Later in ’08: extended to top quarks by ATLAS; Thaler & Wang; Kaplan, Rehermann, Schwartz & Tweedie. Beginning of a wave of intense activity — “fat” jets in SM, SUSY (standard & R-parity violating), Z ′ Extra Dimensions, etc. — motivating ∼ 6 dedicated workshops Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 16 / 25 [3. Jets research] [Undoing clustering] ATLAS and CMS have both shown this working
and CDF have a related jet-mass based top study
ATLAS require lepton, missing ET and 1 jet R = 1.2 with “mass-drop and filter” and plot mass of resulting jet — clear W peak
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 17 / 25 [3. Jets research] [Undoing clustering] ATLAS and CMS have both shown this working
and CDF have a related jet-mass based top study
ATLAS require lepton, missing ET CMS use their top tagger (+ prun- and 1 jet R = 1.2 with “mass-drop ing + mass drop variables) — see and filter” and plot mass of resulting W and top peaks and place con- jet — clear W peak straints on t¯t resonances!
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 17 / 25 [3. Jets research] [Undoing clustering] ATLAS and CMS have both shown this working
and CDF have a related jet-mass based top study
ATLAS require lepton, missing ET CMS use their top tagger (+ prun- and 1 jet R = 1.2 with “mass-drop ing + mass drop variables) — see and filter” and plot mass of resulting W and top peaks and place con- jet — clear W peak straints on t¯t resonances!
With the observation high-pt collimated W’s and tops, we truly start to exploit LHC’s ability to probe EW theory well above the weak scale.
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 17 / 25 [3. Jets research] [Energy flow] 2. Exploiting energy-flow beyond LO structure
Background (e.g. g → gg) and signal (e.g. W → qq¯) often have different colour structure → different radi- Jankowiak, Hook and Wacker ’11 ation patterns.
◮ Pull (non-boosted context) Gallicchio & Schwartz ’10 ◮ N-subjettiness Jihun Kim ’10; Thaler & Van Tilburg ’10 ◮ “Buried Higgs” light singlets Falkowski et al ’10; Chen et al ’10 ◮ Boosted decision trees Cui & Schwartz ’10 ◮ Dipolarity, applied to HEPTopTagger Jankowiak, Hook and Wacker ’11 ◮ Jet deconstruction Soper & Spannowsky ’11 ◮ Template method beyond LO Almeida et al ’11 ◮ . . .
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 18 / 25 [3. Jets research] [Energy flow] 2. Exploiting energy-flow beyond LO structure
Background (e.g. g → gg) and signal (e.g. W → qq¯) often have different colour structure → different radi- Jankowiak, Hook and Wacker ’11 ation patterns.
◮ Pull (non-boosted context) Gallicchio & Schwartz ’10 ◮ N-subjettiness Jihun Kim ’10; Thaler & Van Tilburg ’10 ◮ “Buried Higgs” light singlets Falkowski et al ’10; Chen et al ’10 ◮ Boosted decision trees Cui & Schwartz ’10 ◮ Dipolarity, applied to HEPTopTagger Jankowiak, Hook and Wacker ’11 ◮ Jet deconstruction Soper & Spannowsky ’11 ◮ Template method beyond LO Almeida et al ’11 ≃ ◮ . . . Jet-deconstruction all-order analytic matrix-element method
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 18 / 25 [3. Jets research] [Cleaning jets] 3. Noise removal from jets
Key idea: ◮ Look at jet on smaller angular scale ◮ Discard its softer parts
◮ Filtering Butterworth et al ’08 ◮ Pruning Ellis, Vermillion and Walsh ’09 ◮ Trimming Krohn, Thaler & Wang ’09 [With earlier methods by Seymour ’93 and Kodolova et al ’07]
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 19 / 25 [3. Jets research] [Cleaning jets] 3. Noise removal from jets
Key idea: ◮ Look at jet on smaller angular scale ◮ Discard its softer parts
◮ Filtering Butterworth et al ’08 ◮ Pruning Ellis, Vermillion and Walsh ’09 ◮ Trimming Krohn, Thaler & Wang ’09 [With earlier methods by Seymour ’93 and Kodolova et al ’07]
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 19 / 25 [3. Jets research] [Cleaning jets] 3. Noise removal from jets
Key idea: ◮ Look at jet on smaller angular scale ◮ Discard its softer parts
◮ Filtering Butterworth et al ’08 ◮ Pruning Ellis, Vermillion and Walsh ’09 ◮ Trimming Krohn, Thaler & Wang ’09 [With earlier methods by Seymour ’93 and Kodolova et al ’07]
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 19 / 25 [3. Jets research] [Cleaning jets] 3. Noise removal from jets
[Boost 2010 writeup]
◮ Filtering Butterworth et al ’08 ◮ Pruning Ellis, Vermillion and Walsh ’09 ◮ Trimming Krohn, Thaler & Wang ’09 [With earlier methods by Seymour ’93 and Kodolova et al ’07]
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 19 / 25 [3. Jets research] [Cleaning jets] Noise & different kinds of event
Plain pythia event
The problem of noise and contamination is common to low-lumi pp running, high-lumi LHC pp running, and heavy-ion running
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 20 / 25 [3. Jets research] [Cleaning jets] Noise & different kinds of event
Plain pythia event + today’s pileup
The problem of noise and contamination is common to low-lumi pp running, high-lumi LHC pp running, and heavy-ion running
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 20 / 25 [3. Jets research] [Cleaning jets] Noise & different kinds of event
Plain pythia event + 10× today’s pileup
The problem of noise and contamination is common to low-lumi pp running, high-lumi LHC pp running, and heavy-ion running
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 20 / 25 [3. Jets research] [Cleaning jets] Noise & different kinds of event
Plain pythia event + heavy-ion type bkgd
The problem of noise and contamination is common to low-lumi pp running, high-lumi LHC pp running, and heavy-ion running
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 20 / 25 4. Full jet reconstruction in heavy-ion collisions
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 21 / 25 ET 1 − ET 2 Quantitative measure: AJ = ET 1 + ET 2 [with ∆φ as a cross-check]
AJ
∆φ
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 22 / 25 p − p Average quantitative measure from CMS: t1 t2 pt1
Order of magnitude of medium effects: 10%
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 23 / 25 For quantitative HI studies at 10% level, details of methods and their systematics become key: ATLAS (& ALICE, STAR, PHENIX): subtract off estimated background from jet, without noise suppression → ✓ zero bias ✗ large fluctuations (∼ 16 GeV; enhances asymmetries)
CMS: subtract more than estimated background, suppress “neg- ative towers”, i.e. powerful noise suppression → ✓ lower fluctuations [like filtering, pruning trimming] ✗ but extra biases that cancel partially in a jet-structure dependent way (CMS sees 5% difference between quarks and gluons) Tradeoff discussed e.g. in Cacciari, Rojo, GPS & Soyez ’10 and Cacciari, GPS & Soyez ’11
As LHC pp program → high lumi (high noise) and HI → precision, both stand to gain from deeper understanding of jet finding [Conclusions] Conclusions
We are at a turning point in jet finding
LHC is probing the highest scales ever through jets LHC is starting to see W’s, Z’s, tops as single jets — as we reach above the electroweak scale they “become” light Fully reconstructed jets are a concrete and powerful reality in heavy-ion collisions and are starting to become quantitative probes Theoretical understanding continues to accompany this progress
The time is here for advanced jet methods to be part of the search for the Higgs, BSM physics and probe the hot dense medium of HI collisions
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 25 / 25 EXTRAS
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 26 / 25 [Extras] Quenching v. fluctuations
High pt event
QUENCHED JET THAT HAS LOST ENERGY [Extras] Quenching v. fluctuations
High pt event Moderate pt event
QUENCHED JET NO QUENCHING, THAT HAS LOST ENERGY [Extras] Quenching v. fluctuations
High pt event Moderate pt event
At LHC, each jet accompanied by O (100 GeV) of noise.
noise
QUENCHED JET NO QUENCHING, THAT HAS LOST ENERGY [Extras] Quenching v. fluctuations
High pt event Moderate pt event
upwards noise At LHC, each jet fluctuation accompanied by O (100 GeV) of noise.
noise
QUENCHED JET NO QUENCHING, THAT HAS LOST ENERGY NOISE−INDUCED ASYMMETRY [Extras] Quenching v. fluctuations
High pt event Moderate pt event
upwards noise At LHC, each jet fluctuation accompanied by O (100 GeV) of noise. Large, O (50 GeV), fluctuations rare.
But moderate-pt noise events much more common than high-pt QUENCHED JET NO QUENCHING, THAT HAS LOST ENERGY NOISE−INDUCED ASYMMETRY
Asym. may mix quenching & flucts; quantify latter to learn about former [Extras] Results from Hydjet embedding (ATLAS cuts)
5 5 5 5 40-75% 20-40% 10-20% 0-10% 4 4 4 4 J 3 3 3 3
2 2 2 2 1/N dN/dA 1 1 1 1
0 0 0 0 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
AJ AJ AJ AJ 10 10 10 10 √ min LHC, s=2.76 TeV Pythia 6.4 Hydjet, pt =70 GeV anti-kt, R=0.4 Hydjet 1 1 1 1 "pp" ref ∆φ
10-1 10-1 10-1 10-1 1/N dN/d 10-2 10-2 10-2 10-2
2 2.5 3 2 2.5 3 2 2.5 3 2 2.5 3 ∆φ ∆φ ∆φ ∆φ ◮ Effect of fluctuations appears significant Cacciari, GPS & Soyez ’11 ◮ It is crucial to include sufficiently low-pt Pythia events For the fluctuations present in Hydjet, original ATLAS choice of generation cut, 70 GeV, fails to reveal the true impact of fluctuations
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 28 / 25 [Extras] Crosscheck that HYDJET fluctuations are OK
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 29 / 25 [Extras] Crosscheck that our detector simulation is not crazy
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 30 / 25 [Extras] AJ v. pt from ATLAS
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 31 / 25 [Extras] ATLAS resolution from simulations
What is absolute resolution? Take Lowest bin: 26% × 60 GeV ≃ 16 GeV This is same ballpark as in our HYDJET simulations.
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 32 / 25 [Extras] Jet-structure reconstruction dependence
1.1
> CMS Simulation
T quark and gluon PYTHIA, |η| < 2 jets have different internal structure
1 that induces 5% difference in recon- / Gen Jet p
T structed energy caused by noise suppression? Gluons, Iterative Cone, Calo, R = 0.5 Gluons, Anti-k , PF, R = 0.3 0.9 T how modified is Quarks, Iterative Cone, Calo, R = 0.5 quenched jet struc-
< Reco Jet p Quarks, Anti-k , PF, R = 0.3 T ture? 40 60 80 100 120 140 some data say not much Gen p (GeV/c) T
Gavin Salam (CERN/Princeton/Paris) Jets in SM and beyond PANIC, 28 July 2011 33 / 25