Particle Flow Calorimetry

I took a break today from thesis writing to attend the weekly computing seminar, since it had calorimetry in the title, which should supposedly be one of my skills. The slides aren’t posted yet, but whenever they are you can find them here. Mark Thomson was the speaker, who also has a few papers out on the same subject on the web, this one seems to have the most overlap with the talk.

The idea goes a bit like this: If you want to test new physics at a collider, you need to measure collision products very carefully. Specifically, testing electroweak sector models (all this higgs hubub) requires identifying and differentiating Z and W bosons real well. Single production Z and W’s both decay in to quark-antiquark pairs. We can’t see quarks in detectors directly, but we observe sprays of particles called jets which have basically the same energy and direction, etc. So they want to measure jets real well at the proposed and under-development International Linear Collider (ILC), and the main technique to do this that they are trying to develop is called Particle Flow Calorimetry (PFC).

A Generic Detector

A Generic Detector

Pedantic point: The muon in the above figure shouldn’t be arcing after leaving the solenoid, and definitely not after leaving the iron yoke.

We measure particles by the way they interact with our detector, and different particles interact in different ways. Two of the most important signatures of particles are tracks and calorimeter hits. Tracking chambers only measure charged particles, try to barely interact with them, and watch where they go in a magnetic field. Calorimeters do the opposite: they try to stop everything completely, and add up all the energy they absorbed. We can measure charged particles like electrons, protons, and charged pions very well with trackers, even better than with calorimeters. Photons have no charge, so we don’t see them in trackers, but interact with the calorimeter in a way that’s pretty well understood. Hadrons shower in the cal in kind of nasty, uncertain ways, sometimes leaking out a bit, sometimes showering early like electrons. Neutrons are a big source of uncertainty, because they leave no tracks, and only act weakly, showering late. Neutral pions are really annoying because sometimes they look like a neutron, and sometimes they convert into two photons. Jets are diverse, uncertain mixtures of all these things. The ZEUS calorimeter (which is probably on a ship on its way back to America for burial right now) is pretty fantastic, because it is what’s called compensating. This means the original designers fine tuned the absorber materials in it so that all the stable particles have a very similar response. Why people stopped designing detectors this way is a mystery to me.

One common way to improve jet measurements is to combine tracks and CAL deposits into single objects, usually called Energy Flow Objects. This helps, but since you don’t always know how much of a jet is neutral or charged there’s still a bit of uncertainty introduced. According to Thomson has been shown to be far less effective than PFC at ILC. The basic idea of PFC (seems to me to be) that you use a really spatially-fine-grained calorimeter, and when you have a track leading to the cal, you measure what you can with the tracker alone, and whatever else with the cal. This means adding the track energy(ies) and everything except the deposit(s) that belongs to the track(s). Moreover, you try to tag every particle in your calorimeter to undertand your shower profile uncertainities. Sounds easy, but the problem comes when deposits overlap, or the particle showers in a funky way, or two tracks point close to each other, or when you have electronics producing noise and giving fake signals. The people working on this algorithm seem to think they have all the possible scenarios accounted for, but I’m a bit skeptical. ILC people are designing a detector, not running one, so they’re running over simulations to get their answers, and real world collider data never looks like the simulations for the first few years. Then again, this is for an electron-positron collider everyone seems to think the background there will be negligible. Perhaps I’m just pessimistic from two years of painstakingly weeding calorimeter sparks, beam-gas and cosmic rays out of my data sample.

For your viewing pleasure, a simulated jet in the simulated TESLA detector that is one of several proposed designs for a detector that will built in 10 years ( depending on changes in GDPs and future elections in several nations). Different colors label different particles as identified by the PandoraPF algorithm. a partial view of the tracking chamber is on the left, and cal hits are the dots on the right. I stole this from Thomson’s paper here.

Simulated event in TESLA

Simulated event in TESLA


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s