Top Banner
Review of Ice Models • What is an “ice model”? • PTD vs. photonics • What models are out there? • Which one(s) should/n’t we use? Kurt Woschnagg, UCB AMANDA Collaboration Meeting Madison, June 29, 2003
16

Review of Ice Models

Jan 12, 2016

Download

Documents

bozica

16-layer. stdkurt. asens. Review of Ice Models. bulk. sudkurt. Kurt Woschnagg, UCB AMANDA Collaboration Meeting Madison, June 29, 2003. mam. f125. kgm. photonics. What is an “ice model”? PTD vs. photonics What models are out there? Which one(s) should/n’t we use?. - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Review of Ice Models

Review of Ice Models

• What is an “ice model”?• PTD vs. photonics• What models are out there?• Which one(s) should/n’t we use?

Kurt Woschnagg, UCBAMANDA Collaboration Meeting

Madison, June 29, 2003

Page 2: Review of Ice Models

Ice Model Basics

• Need to describe photon(A,t) around light sources (muon tracks, cascades) for simulations of AMANDA events

• Tables of photon(A,t) are calculated using Monte Carlo propagation of photons through scattering and absorbing medium (ice)

• Dependence on distance, OM orientation, ice properties, etc

• Prefer to have simplifications (symmetries, averages, approximations) but must pay price for these

Page 3: Review of Ice Models

Light propagation codes: two approaches

PTD• Photons propagated through

ice with homogeneous prop.• Uses average scattering

• No intrinsic layering: each OM sees homogeneous ice, different OMs may see different ice

• Fewer tables• Faster• Approximations

Photonics• Photons propagated through

ice with varying properties• All wavelength dependencies

included• Layering of ice itself: each OM

sees real ice layers

• More tables• Slower• Detailed

Page 4: Review of Ice Models

photonicsBulk PTD Layered PTD

PTD vs. photonics: layering

average ice type 1 type 2 type 3 “real” ice

3

3

3

1

1

2

2

2

2

2

2

Page 5: Review of Ice Models

PTD vs. photonics: wavelength dependence

PTD

1. Create scattering tables with colorless photons of infinite lifetime propagated through average scattering

2. Apply absorption on tables assuming Cherenkov spectrum

Photonics

1. Generate photons from Cherenkov spectrum2. Scattering depends on wavelength3. Absorption depends on wavelength and is applied

during propagation as weight

Page 6: Review of Ice Models

Too many photons surviving at larger distances in PTD.

Caused by applying absorption to

scattering tables assuming

Cherenkov spectrum at all

distances.

PTD vs. photonics: light yield

Page 7: Review of Ice Models

Bulk Ice

• Assume homogeneous optical properties• Average scattering and absorption (F) over wavelength• Sometimes referred to by it’s F value (100): f125, f096

+ simple model, fast

– does not have layering

PTD

Page 8: Review of Ice Models

F value for dust absorption

Definition (from 20th century):

adust = 0.01 F (/337)-1.16 used in PTD

Compare to

adust = CMdust -1.08 used in

photonics

Page 9: Review of Ice Models

The “kurt” Model(S. Hundertmark)

• Use measured ice properties available at the time* (532 nm, DC data)• Introduce vertical ice profile – layers (16-layer model)• Average scattering over wavelength by scaling: 0.9·e(532)• Take absorption at 420 nm and apply scaling: 0.85·a(420)• Assume linear correlation between absorption (dust) and scattering• Flavors:

►standard kurt (stdkurt) – uses Goobar glass measurements►Sudhoff kurt (sudkurt) – uses Sudhoff glass measurements►asens – uses angsens modification of angular OM acceptance

+ based on measured ice properties with layers– measured ice does not work well with PTD – averaging incorrect– abs/scat correlation incorrect (not known at the time)

PTD

* May 2000

Page 10: Review of Ice Models

The Kurt-Gary Model (KGM)(Gary, Kurt)

• Use same measurements as for “kurt” plus additional wavelengths = best measured properties as of Fall 2001

• Recalculate averages over wavelength: 0.9 → 0.7-0.8 for absorption/scattering length scaling

• Reoptimize depth layers

+ based on measured ice properties with layers+ more realistic averaging+ dirtier ice consistent with data/MC disagreement– measured ice does not work well with PTD – abs/scat correlation still incorrect

PTD

Page 11: Review of Ice Models

Effective Wavelength Spectrum

Cherenkov spectrum + glass + gel + PMT

But…1/2 only at source,so not really correct

Page 12: Review of Ice Models

“kurt”-KGM comparison

Page 13: Review of Ice Models

The Muon Absorption Model (MAM)(Gary, Albrecht, Paolo)

• Look at time residuals for downgoing muons → too many late hits in MC

• Increase absorption to make time residuals match between data & MC

• Use same depth layering as in KGM• Modified absorption independent of input model

+ Downgoing muons “fit”

+ Better MC/data agreement

– Tuning

… Does not use measured absorption

PTD

Page 14: Review of Ice Models

Absorption in Old Photonics

Old photonics

KGM

Data

Better dust/be(532)correlation

Incorrect correlation between dust concentration and scattering at 532 nm lead to too little absorption in 1st generation (= old)photonics tables

Page 15: Review of Ice Models

Old vs. New Photonics

~20% less light for new ice model*

(*see previous talk for details)

From Ped’s talk

Page 16: Review of Ice Models

Summary & Recommendations• The commonly used ice models (“kurt”, KGM, MAM, bulk, asens)

are all based on PTD tables• There’s nothing wrong with that – but know what you’re using• Due to (justified) simplifications in the treatment of wavelength

dependence and layering in PTD-based models certain features in the data are not reproduced (e.g. cogz structure*), especially when actual ice properties are used

► Make effort to transition to photonics!

► Still ~OK to use MAM/KGM/bulk for certain cases (i.e. when the MC does describe the specific class of data) while we wait for photonics

► Do not use “kurt” model

► Do not use 1st generation photonics tables

*See Marek’s study introducing thin layers of infinite absorption as a fix

Next time: hole ice