Basics & applications of effective areas - DESY
Transcript of Basics & applications of effective areas - DESY
Basics & applications of effective areasMartin Tluczykont, Zeuthen, November 21, 2006
● effective areas– basic definition: most simple experiment– application to IceSim (ANIS)– application to nusim– effective volumes
● other applications
Disclaimer
“The problem with Hegel is that he liked to explain. That's why nobody understands him.”
JeanPaul Sorg (private communication)
Encouragement
“Actually it's easy.”
Martin Tluczykont (helpful comments)
BASICS
Definition of effective areaNumber of detected events:
or: area of an ideal detector
i.e. area of 100% efficient detector
N d=T∬ dE dE ,Aeff E ,
=
N d
Trivial formula : number of genearated Monte Carlo events : generation area : number of detected events : detector area : simulated flux
&
Aeff=N d
N gen
Agen
Ad
N d
Agen
N gen
gen
gen=N gen
Agen Tgen=
N d
Aeff T
EXAMPLE:WATER DROP EXPERIMENT
Most simple case: water drops fall in cup● Simulate drops on area● Detect drops on area ● No dependencies (energy, ...)● all drops on are detected
This is an ideal detector:
N d=gen Ad=N gen
Agen
Ad
Aeff=N d
N gen
Agen=Ad
N gen Agen
N d Ad
Ad
More complex case: water drops but...● The physics of water drops
– created in clouds, falling through atmosphere: interaction
– can evaporate: only a fraction drops reach the cup– drop reflection: not all drops detected (energy dependent effect)
● The simulation of water drops– drops close to the detector, interaction enforced (CPU power)
true drop number:
– energy dependent drop simulation:
N gen=N gen' /Pev
N genE =N gen fE
Pev
N gen'
More complex case: water drops but...
● Effective area calculation
Aeff E =N d E
N genE Agen
Aeff E =N d E
N gen' f
E
Agen Pev
N genE =N gen
'
Pev
fE
APPLICATIONS
Event MC in nonthermal astrophysics● here: nusim & IceSim = ANIS
●
● typically
● large # simulated events &
dNdE
=N 0 E− , E1EE2
=1
AgenAd
Effective areas in astrophysics
● Reconstruction & analysis dependent– Algorithm (First guess, likelihood...)– Trigger / Filtering level– Analysis cuts
● Zenith angle (declination) dependent● Energy dependent
Energy & Declination bins● Energy bin i: boundaries
fraction of events simulated in energy bin i:
● Declination bin j:fraction of events simulation in decl.bin j:
[e1 , e2]i
f Ei=∫e1
e2dE E−
∫E1
E2dE E−
=ln e2 /e1iln E2 /E1
f dec j=2[cos1−cos2] j
[1 ,2] j
Application to IceSim (ANIS)
● neutrinos parallel to axis of cylindrical volume● neutrino interaction inside Volume is enforced:
●
●
N geni , j=
N gen'
Pinter
f Ei f dec j
Pinter
N gen'
Application to IceSim (ANIS)● Simulation of:
● constant simulation area:● interaction probability: ● simulated solid angle:● effective area in energy bin i, declination bin j:
dNdE
=N 0 E− , E1EE2
Agen=r2=N gen
T gen
Aeffi , j=∑events
N di , j
N geni , j
Agen Pinteri , j
Pinteri , j
Application to IceSim (ANIS)
Aeffi , j=∑events
ln E2 /E1
ln e2 /e1i
2[cos1−cos2] j
Agen Pinteri , j
effective area in energy bin i, declination bin j:
Application to IceSim (ANIS)
Application: nusim+amasim● Simulation of:
● variable simulation area:● interaction weight: ● simulated solid angle:● effective area bins in energy (i) and declination (j):
dNdE
=N 0 E− , E1EE2
Agen j =r14 1010[cm2]
Aeffi , j=∑events
N di , j
N geni , j
Agen j W inter
i , j
W interi , j
Application: nusim+amasim
effective area bins in energy (i) and declination (j):
Aeffi , j=∑events
ln E2 /E1
ln e2 /e1i
2[cos1−cos2] j
r14 1010 W interi , j
Application: nusim+amasim
Some IceRec framework details
● nusimintweight, R14 available in analysistree & flatntuple: RDMCUserlines
● IceSim (I3NeutrinoGenerator = ANIS) – Information stored in
I3Map<string, double> I3MCWeightsDict
– now available in flatntuple: MCWeights– not yet in analysistree (maybe soon)
Effective Volumes● same principle:
● ANIS case: is defined for all flavors
● effective areas for diffuse fluxes: no declination dependency
V eff=N d
N gen
V gen
Pinter
One last question about effective areas
Why to the kuckuk do I need this effective area?
● to predict the number of events detected for a given model spectrum
● to impress your fellow students with color plots
N =∫ dE Aeff E ,model E ,
Other applications of (other) Weights...● predicted pointsource event rates from a model:
● predicted atmospheric event rates:
N events=∑ Pinter W norm
Fmodel E ,
FgenE
N atmos=∑ Pinter W norm W atmo
Summary● Recipe for effective area calculation● example using nusim & IceSim/ANIS● Now possible within IceRec framework (flatntuple)● First preliminary result for IC23 using
I3NeutrinoGenerator