AlgoritmoParticleFlow < ILC

ILC
ILC.AlgoritmoParticleFlowr1.4 - 23 Dec 2008 - 11:50 - Main.iglesiastopic end
---++ 4. *Algoritmo Particle Flow en la zona hacia delante*


---+++ 4.a. Algoritmo Particle Flow

---++++ 4.a.i. Introduccion

El algoritmo Particle Flow empezo a desarrollarse en la era LEP, y tambien ha sido utilizado en Tevatron y en el LHC, tanto en CMS como en ATLAS (ver mi web de ATLAS:  http://ific.uv.es/~iglesias), y para el futuro acelerador lineal ILC, los detectores _ILD_ y _SiD_ adoptaran un diseño basado en el Particle Flow, porque este concepto promete alcanzar la resolucion en energia de los jets que se quiere conseguir.

Pero para este fin, Particle Flow requerie que se desarrolle una nueva tecnologia de detectores ademas de unos nuevos algoritmos de reconstruccion mas sofisticados.
 
Paricle Flow tiene ademas unas altas exigencias tanto a nivel de calorimetria como del detector de trazas:
   * requisitos de la calorimetria: alta granularidad y cobertura a todo angulo
   * requisitos del detector de trazas: buena resolucion el momento de las trazas, incluso a muy bajos angulos. 

---++++ 4.a.ii.Concepto de Particle Flow
As you know, in a typical jet, we have:
   * 64%  of jet energy in charged hadrons (mainly &pi;±)
   * 25 %  in photons (mainly from &pi;&rarr;&gamma;&gamma; decays) 
   * 11 %  in neutral hadrons (mainly KL and neutrons) 
About 2/3 of the jet energy are carried by charged particles (protons, kaons...). However jet algorithm makes no use of tracking information

Particle Flow algorithm make an optimal use of the detector information combining the measurement of the energy deposition in calorimeter cells  with the reconstructed track in the inner detector to improve jet energy resolution and ETMiss.

For ILC energies, the tracker is more precise than the calorimeters. As you can see in this picture, tracking is better than:
   * ECAL for Epart < 120 _GeV_
   * EHAD for Epart < 370 _GeV_
So for charged particles, their energy resolution will be sustituted by the track momentum resolution in order to have a better resolution in jet ET. 
 <br />
    <img src="%ATTACHURLPATH%/Energy_resolution_ILC.jpg" alt="Energy_resolution_ILC.jpg"  width="367" height="254"  />




---++++ 4.a.iii. Metodo de calorimetria tradiccional vs Paradigma de Particle Flow

*Traditional calorimetric approach:*
   * Measure all components of jet energy in ECAL/HCAL !
   * ~70 % of energy measured in HCAL: <img src="%ATTACHURLPATH%/HCAL_energy_resolution.jpg" alt="HCAL_energy_resolution.jpg"  width="244" height="30"  />
   * Intrinsically poorHCAL resolution limits jet energy resolution

 <br />
    <img src="%ATTACHURLPATH%/Particle_Flow_Paradigm.jpg" alt="Particle_Flow_Paradigm.jpg"  width="600" height="182"  />


*Particle Flow Paradigm:*

The idea is, for a detector ideal in separating the different particles:
   * The charged tracks will be measured from the tracker
   * The photons from the Ecal: <img src="%ATTACHURLPATH%/ECAL_energy_resolution.jpg" alt="ECAL_energy_resolution.jpg"  width="244" height="28"  />
   * And only the neutral hadrons will be measured in the Hcal (and possibly in Ecal)
So, only 10 % of jet energy from HCAL &#61664; much improved resolution


---++++ 4.a.iv. Reconstruccion con Particle Flow

But for a *real detector* two effects play a role:
   * 1- Particles travelling at small angles to the beam axis will not be detected.
   * 2- It is not possible to perfectly associate all energy deposits with the correct particles:
      * a- The existence of an *effective threshold* on
         * Charged particles due to the high magnetic field needed for background, precision and separation 
         * photons due to cell threshold and physical background
      * b- The probability of *confusion*:
         * efficiency of track reconstruction
         * vertex misidentification
         * wrong associations between tracks and calorimeter cells

In this way, for the recontruction of a particle flow calorimeter we must to *avoid double counting* of energy from same particles, as well as we have to *separate energy deposits from different particles*, because, as you can see in the picture
 <br />
    <img src="%ATTACHURLPATH%/Part_Flow_Reconstruction.jpg" alt="Part_Flow_Reconstruction.jpg"  width="700" height="120"  />

So, the energy resolution of the jet will be determinates by the level of mistakes and confusion not only by the intrinsic calorimeter performance of ECAL and hadronic calorimeter.


---++++ 4.a.v. La calorimetria con Particle Flow

In practice, particle flow performance depends on the ability to associated energy deposits 
without double counting or merging energy deposits. This places stringent requirements on 
the granularity of ECAL and HCAL and, consequently, Particle Flow  performance is one of the 
main factors driving the overall detector design. 

The jet energy resolution obtained is the combination  of the *detector* and PFA *software performance*.

---++++ 4.a.vi.Como funciona el Particle Flow?
The crucial step of the particle flow algorithm is the correct assignment of calorimeter hits to the charged particles and the efficient discrimination of close-by showers produced by charged and neutral particles.

The main steps in the Particle Flow algorithm are:
   * Track Reconstruction
   * Extrapolate track to calorimeters
   * Assign MIP to stubs to tracks
   * Clustering in calorimeters
   * Particle ID for Charged particle
   * Removed charged particle hits in calorimeter
   * Clustering of neutral hits
   * Particle ID for neutral

 <br />
    <img src="%ATTACHURLPATH%/howPFworks.jpg" alt="howPFworks.jpg"  width="357" height="400"  />



-- Main.iglesias - 22 Dec 2008
to top
End of topic
Skip to action links | Back to top
Edit | Attach image or document | Printable version | Raw text | More topic actions
Revisions: | r1.4 | > | r1.3 | > | r1.2 | Total page history | Backlinks
You are here: ILC > AlgoritmoParticleFlow

to top