• John
  • Felde
  • University of Maryland
  • USA

Latest Posts

  • USA

  • James
  • Doherty
  • Open University
  • United Kingdom

Latest Posts

  • Andrea
  • Signori
  • Nikhef
  • Netherlands

Latest Posts

  • CERN
  • Geneva
  • Switzerland

Latest Posts

  • Aidan
  • Randle-Conde
  • Université Libre de Bruxelles
  • Belgium

Latest Posts

  • Vancouver, BC
  • Canada

Latest Posts

  • Laura
  • Gladstone
  • MIT
  • USA

Latest Posts

  • Steven
  • Goldfarb
  • University of Michigan

Latest Posts

  • Fermilab
  • Batavia, IL
  • USA

Latest Posts

  • Seth
  • Zenz
  • Imperial College London
  • UK

Latest Posts

  • Nhan
  • Tran
  • Fermilab
  • USA

Latest Posts

  • Alex
  • Millar
  • University of Melbourne
  • Australia

Latest Posts

  • Ken
  • Bloom
  • USA

Latest Posts

Brian Dorney | USLHC | USA

View Blog | Read Bio

Anatomy of a Jet in CMS

We talk often about Jets here at US LHC.  We talk about ways to identify them, their structure, and we even mention some crazy phenomenon involving them.  But one thing we don’t always talk about is what a jet looks like.  And this is what I would like to show today, in gory detail.  So this post is about pictures, lots and lots of pictures.

We can’t see jets with our eyes.  The particles that make up a jet are just to small.  But using a device as large as the CMS detector, we can take a “snap-shot” of a jet created in a proton-proton collision at the LHC.  So it behooves us to start with a brief illustration of CMS.

(Clicking on any of the images below allows you to blow them up in another window, just in case you need a bigger picture)

The CMS Detector

It’s a gigantic cylinder, twenty-one meters long and fifteen meters in diameter!  For comparison, the average height of all American women & men over 20 years old is 1.62 & 1.76 meters (5′ 4″, 5′ 9″), respectively.  But being gigantic isn’t the only thing CMS has in common with an ogre, CMS also has layers; and each layer is a different sub-detector responsible for identifying a different class of particles.

But Here’s computer generated image of CMS, with a cut-away section showing some of these layers:


Cut away view of CMS, Hadronic & Electromagnetic Calorimeters not shown.  The camera guide in the bottom right shows CMS’s coordinate axis, (x,y,z) for (red,green,blue).


The blue disks and red rectangles on the outside of the detector are part of CMS’s muon chambers (the sub-detector responsible for picking up muons).

The inner cylinder represents CMS’s silicon tracker, this sub-detector is a rather complex instrument.  Its made of silicon strips, and is essentially a giant CCD camera (with over ten million pixels).  The silicon tracker is responsible for reconstructing the trajectories of charged particles as they pass through CMS; this is done by basically playing a giant game of connect the dots.  A close-up of the silicon tracker is shown here:


Close-up of the silicon tracker


The green and yellow portions are the silicon tracker.  The grey/silver part is what’s called the silicon pixel detector.  It is less then an inch away from where the proton-proton collisions occur in CMS, and thus the closest detection element.

While not shown in the image above, CMS’s calorimeters and superconducting magnet are located between the silicon tracker and the muon system.  They can be seen in this interactive applet on particle detection, of which I’ve taken a screenshot of and shown below:



The calorimeters are responsible for measuring the energy and momentum of charged & neutral particles (the tracker plans a role in this as well).  They are critical to jet identification & reconstruction…without them we would not be able to do any jet physics in CMS.

Basically what happens is calorimeters are designed so that a particle loses all of its energy as it travels through the calorimeter.  From the energy deposited, and the location of where the deposit occurs we can determine the direction and momentum of charged particles (again, the tracker also plays a role in this).

CMS has two types of calorimeters: an electromagnetic calorimeter (ECAL) for detecting electrons and photons; and a hadronic calorimeter (HCAL) for detecting heavy particles that can pass through the ECAL.  The HCAL is also the only place in CMS where we can detect neutral particles with non-zero mass (such as a neutron).

I should mention that neutrino’s escape detection, and we have to infer their presence by looking for “missing energy.”

But now that I’ve introduced you to CMS, let’s get down to business and talk about Jets.


Jets in CMS

I use jets extensively in my own research and its sometimes hard to get a handle on what a jet really is.  I like to think of it like a shotgun blast of particles slamming into the detector.  Jets arise from the hadronization of colored particles, and because of this they are made up of many particles.  Jets can be made of leptons, hadrons and even bosons (specifically the photon).  These particles are usually collimated in a given direction, and you can kinda draw a cone around them (like a shotgun blast!).

For this reason in CMS we like to think of jets as cones, like in this image:


A jet cone in CMS


This is a jet cone created in a single proton-proton collision recorded by CMS detector in 2010.  In this image I’ve turned the silicon tracker’s graphics off, along with everything else that happened in this event (it can get really messy anyway).  A zoom in of this jet cone can be seen here:


Zoom in of the jet cone


Now this jet cone may look small in comparison to the entire CMS Detector; but don’t be fooled, I choose a very energetic jet for this post.  This jet’s component of momentum in the xy-plane (green & red axis above) is 115 GeV/c.  Most jets created in proton-proton collisions have xy-momentum components of less then 30 GeV/c.  In fact, if you plot the number of jets detected against their xy-momentum components, you get a distribution that looks similar to an decaying exponential.  So, 115 GeV/c jet is rather energetic.

But what makes up this jet!?  Simple answer A LOT!  This one jet shown above is made up of over 20 different particles, all of which are conveniently hidden at the moment.  So let’s go about fleshing this jet out, piece by piece.


Jet Anatomy

So what did CMS see as this jet hit the detector?  Let’s start with the silicon tracker:


A jet & its tracks in CMS


So here I’ve turned off the view of the muon chambers, and just shown the jet cone and the tracks (the green lines) reconstructed in the silicon tracker.

A few things to note, the green lines appear to be coming from the same point in space (for the most part).  This point is called the primary vertex, its the point at which the proton-proton collision actually occurred.  Another interesting feature is that these tracks go outside of the jet’s cone!  What’s up with that!?

Well the answer is two-fold.  First, these tracks where made by charged particles; and thus their trajectories are bent in the presebce of CMS’s magnetic field.  In other words, we forced the tracks to go outside the jet cone by having a superconducting magnet in our detector (this allows us to make more precise momentum measurements).  Second, treating jets as cones is just a model (which works well).  It often happens that these tracks are indeed all inside the cone, but I purposefully chose a jet with tracks that had large curvature for this demonstration.

In fact, jets are created by using different algorithms; and not all algorithms use a cone geometry!  There are many different algorithms that you can use, they all have subtle differences…but that is really a story for another day.  I just want to show  you what makes up a jet, and what it looks like to the detector.

So this is all the information that the silicon tracker gave us about the jet.  It’s time to ask the calorimeters what they saw, starting with ECAL:

A Jet, its tracks, and its energy deposits in CMS’s electromagnetic calorimeter


These yellow-ish squares represent the energy deposited in ECAL.  How far these squares protrude from that wire-frame represents how much energy has been deposited (for those of you who are keeping track, the scale is 10 GeV per meter).  I’ll show some images illustrating this protrusion later on; right now I want to talk about the relation between the tracks, the jet cone and these ECAL energy deposits.

It looks like the jet cone is centered on the bulk of the energy deposits in ECAL.  We actually intended this to happen because of something called clustering.  We group pieces of the calorimeters (both ECAL & HCAL) into clusters, and then clusters into superclusters.  And it is these clusters/superclusters that we go to when beginning to construct a jet.

It also looks like some of the tracks in our silicon tracker match up with these ECAL energy deposits.  But, there are clearly some energy deposits that don’t match up to any track!  What has happened here!?

The answer, is photons and other neutral particles!

The silicon tracker is incapable of detecting particles without an electric charge (like the photon).  But the ECAL was designed specifically to capture electrons & photons (which is why it is called the electromagnetic calorimeter).  Adding “photon candidates” to the picture gives us this result:


A Jet, its tracks, its “photon candidates”, and its energy deposits in CMS’s electromagnetic calorimeter


These dotted purple-ish lines are the trajectories of the “photon candidates” in this jet.  I say candidates because they might not actually be photons.  To be an actual photon the candidate must past very stringent quality requirements which only a real photon will satisfy.  I haven’t enforced any quality requirements here, so all bets are off (remember I’m just trying to show what made up this jet!).

Again all these photon candidates appear to be coming from the primary vertex, most of them are within the jet’s cone (not all); and almost every photon candidate is linked to an energy deposit in the ECAL.  These photon candidates are also linked to ECAL energy deposits that are not linked to tracks identified by the silicon tracker.

But we still have some ECAL deposits that are clearly not linked to either tracks identified by the silicon tracker or photon candidates!  We need to bring up the rest of the neutral particles:


A Jet, its tracks, its “photon candidates”, its neutral hadrons, and its energy deposits in CMS’s electromagnetic calorimeter


The dotted blue lines are the “neutral hadron candidates” within this jet.  Similar observations as before can be made.  But since neither a photon or a neutral hadron leaves a track in the silicon tracker, how do I distinguish between them?   This is were HCAL comes into play.


A Jet, its tracks, its “photon candidates”, its neutral hadrons, and its energy deposits in CMS’s hadronic calorimeter


Now I’ve turned off the energy deposits in ECAL and turned on the deposits in HCAL (teal rectangles).  I’ve also drawn a crude circle around one of the jet’s photon candidates.  It clearly has no HCAL energy deposit around it, but all the dotted blue lines are linked to HCAL deposits.  Now we can clearly see the difference between neutral hadrons and photon candidates; one has energy deposited in HCAL, the other doesn’t.

It might be interesting to note that some photon candidates appear to have HCAL deposits.  There are two reasons for this: 1) the photon candidate isn’t a real photon (it would fail the quality requirements I mentioned above), or 2) there is a nearby hadronic particle that is actually responsible for the HCAL energy deposit.

Now Let’s add the charged hadrons into the picture as well:


A Jet, its tracks, its “photon candidates”, its neutral hadrons, its charged hadrons, and its energy deposits in CMS’s hadronic calorimeter


These bright blue lines represent the trajectories of charged hadrons.  Notice that they are also coincident with the tracks in the silicon tracker (as they should be!).  These charged hadrons also link to energy deposits in HCAL.  In addition, they also have a chance of depositing energy in ECAL as well:


A Jet and its components in CMS


I’ve turned the ECAL deposits back on in the image above.

Notice the ECAL and HCAL deposits are stacked on top of each other (with ECAL appearing first).  We like to do this because this gives us the full idea of the direction of energy deposition in the CMS detector.  Let’s turn our view around so we can see the differences in this jet’s energy deposition:


A jet’s calorimeter deposits in CMS (also note the curvature of the jet’s tracks!!)


Rotated view of a jet’s calorimeter deposits in CMS


Final view of a jet’s calorimeter deposits in CMS


So from these views we can see the amount of energy deposited around the jet’s cone.

Again, height above the wire frame corresponds to the amount of total energy deposited in a region.  In some cases this energy was deposited in both the ECAL (yellow-ish rectangles) and the HCAL (teal rectangles).  The height of these different rectangles corresponds to the amount of energy deposited in their respective calorimeters.

I’ve also now colored all of the jet’s constituents blue, this now is the complete jet, its a spray of particles that goes along a specific direction (shown by the black shaded cone).

For those of you wondering, this is a specific type of jet called an anti-kT particle flow jet.  The algorithm, anti-kT particle flow, used to “reconstruct” this jet made use of the energy deposited in the calorimeters, the tracks in the silicon tracker, and the primary vertex (for determining tracks of neutral particles).

Some algorithms make use of only the calorimeters and the primary vertex (these are called calo-jets).  But discussing the different jet algorithms is a story for another day.

Remember how I said that the scale of those calorimeter deposits was 10 GeV per meter.  Let’s put that into perspective now:


The complete jet recorded by CMS, note the intense curvature of some of the jet’s tracks!


Remember the diameter of CMS is 15 meters, so from the primary vertex to the edge of the red muon system (near my coordinate axis guide at the bottom right) is 7.5 meters.  Hopefully, this gives you an idea of the amount of energy deposited in each of the calorimeter clusters.

So this is what a jet looks like!  All in all this jet had 29 different particles that were used in its construction.

So when we talk about Jets here at US LHC (and the rest of Quantum Diaries) I hope you will have a much better idea of what a jet really is.


Until next time,


For some further reading on Jets, I suggest taking a look at these older posts:


Tags: , ,