Hadronic Calorimeter Granularity

[ChrisVer]

Well that's a fast question, I mainly seek for a confirmation... Is it a fact that the reason behind having lower granularity for the HCAL in ATLAS is that the Hadronic Showers are in general wider and deeper than the Electromagnetic ones? So the HCAL doesn't need high granularity (dealing with wider) but rather radially extended (dealing with deeper) geometry?

 

[mfb]

It is a matter of cost and computing power. If both were unlimited...
Jets are wider than individual electrons/photons, and just a few analyses look into the jet substructure. Also, the energy resolution for jets is much worse (energy lost to neutrinos, neutral pions that form smaller electromagnetic showers with a different calorimeter response, ...).

 

[Vanadium 50]

Also, the energy resolution for jets is much worse

But that's also cost. ATLAS could have made their calorimeter compensating (or at least more compensating), or they could have gone to very fine granularity so that particle flow works better. Both of these would be very expensive. If the budget is fixed, that choice would have meant something else would have to get worse, and the experiment didn't want to do that.

 

[mfb]

You can increase the precision with a better granularity, but you'll never reach the precision of electromagnetic calorimeters for electromagnetic showers.

 

[Vanadium 50]

but you'll never reach the precision of electromagnetic calorimeters for electromagnetic showers.

You'd be surprised. The ATLAS hadron calorimeter has a resolution of 19%/\sqrt{E} and for the ZEUS EM calorimeter it's 18%/\sqrt{E}. Not equal, but pretty close. The main reason hadron calorimeters have poorer resolution isn't the calorimeter - it's the hadrons. Hadron showers are all different, and that difference makes the resolution worse.

 

[mfb]

The main reason hadron calorimeters have poorer resolution isn't the calorimeter - it's the hadrons. Hadron showers are all different, and that difference makes the resolution worse.

That's what I said.

Also, there are 15 years of development between ATLAS and ZEUS.

 

[Vanadium 50]

That's what I said.

Are you sure? It seemed to me you were discussing the calorimeter, not the particles it measured

Also, there are 15 years of development between ATLAS and ZEUS.

Despite the 15 years, those two detectors use very similar technologies. One could have built the ATLAS calorimeter in 1990. That said, I did pick an EM calorimeter favorable to my argument: ZEUS is compensating, and ATLAS is not.

 

[mfb]

Are you sure? It seemed to me you were discussing the calorimeter, not the particles it measured

Things I wrote so far:

• Jets are wider than individual electrons/photons
• Also, the energy resolution for jets is much worse (energy lost to neutrinos, neutral pions that form smaller electromagnetic showers with a different calorimeter response, ...).
• but you'll never reach the precision of electromagnetic calorimeters for electromagnetic showers.

Yes I am sure I discussed the measured particles. And I never discussed the calorimeters - the third bullet point mentions them because the electromagnetic showers just happen to be in the part called electromagnetic calorimeter.

One could have built the ATLAS calorimeter in 1990.

Was money the only difference? ATLAS electromagnetic calorimeter has (design) of about $\sigma/E = 10\%/\sqrt E \oplus 0.6\%$, significantly better than 18%.
Also, where does the number for the hadronic calorimeter come from? The TDR expects 60%/sqrt(E) sampling term.

I wonder how the 40 MHz (low granularity, I know) readout would have worked.

ZEUS is compensating, and ATLAS is not

I know.

 

[ChrisVer]

good to mention those numbers...
I read that the stochastic term was indeed 10%, but the constant term was around 0.17%.
Then, in the next paragraph I read that the constant term is somewhere between 0.5% and 0.7% and so within the acceptance of the detector.
https://twiki.cern.ch/twiki/pub/Atlas/AtlasTechnicalPaper/Published_version_jinst8_08_s08003.pdf
p.154-155

 

[Vanadium 50]

Also, where does the number for the hadronic calorimeter come from?

Electron test beam. That tells you for sure what the calorimeter is capable of as an instrument.

Was money the only difference?

That was part of it, sure, but I think constructability was at least as important. The "natural" HCal for ATLAS would have been LAr, like the endcap. The barrel, however, is huge, and so would need an even larger cryostat. If it could be built at all it would have to be built at CERN, and CERN had a lot on their plate. The TileCal, however, could be built in pieces around the world, and only the final assembly would need to be done at CERN.

 

[mfb]

Electron test beam. That tells you for sure what the calorimeter is capable of as an instrument.

Ah, so 18%/sqrt(E) are for the hadronic calorimeter with electromagnetic showers. That sounds more realistic.