Ice Cube optimal end configuration for increased HE

Ice. Cube - optimal end configuration for increased HE response. May 2, 2008 Collaboration meeting, Madison

Final configuration of Ice. Cube optical strings • Ice. Cube 86 options: – IC 80 + low energy core – IC 80 HE configuration – IC 80 HE + low energy core By the way it is really great that we are in a position to discuss these options.

High energies: Diffuse flux Ice. Cube, 5 , 3 yr Result of AMANDA limit: potential signal E-2 at higher energies

Diffuse E-2 µ-spectrum peaks at 1 Pe. V (after atm. Background rejection) • • • Neutrino event energy spectrum after energy cut for a 3 year diffuse analysis. Signal events peak at ~1 Pe. V Optimize final detector configuration for higher energy range, to maximize sensitivity of Ice. Cube.

Muons in IC 80 Eµ=10 Te. V ≈ 90 hits Eµ=6 Pe. V ≈ 1000 hits Light diameter : 600 m

Cascade event 1 Pe. V shower: • Based on flasher events: Horizontal diameter about 1 km!!! ~ 800 m upper part ~1200 m lower part -Substantially bigger than original simulations. Energy = 375 Te. V

The first 40 -string event (data taken March 10, 2008) Flash 46 -57 “Qi” Flash 46 -27 “Tulip”

High energy benefits • Muons, diffuse E-2 flux – AMANDA upper limit: at 10 -7 – Diffuse flux, 80 string maximum sensitivity requires energy cut at around E>300 Te. V to reject atmospheric neutrino background. – Signal region: 1 Pe. V • Estimate (very rough) increase of effective volume at – at ~1 Pe. V (diffuse muon flux) by factor > 1. 3 (? ) – at ~1 Ee. V (GZK neutrinos) by factor > 1. 5 (? ) • Also high energy GRB, like any other high energy analysis will benefit

Q&A • Will Monte. Carlo, event reconstruction and background rejection more complicated? – A: We have already a good sense for how to do that. We may also learn useful things from the symmetry break. This is not very unusual. • Are there risks in drilling: – Drilling is understood for the cluster approach, which is highly favored from drilling perspective. Outer ring of 1 km is seen very challenging and would require investments in cost and time. • Other: ….

High Energy Ice. Cube Optimization Best case scenario: 12 last strings placed on the rings with radii of 0. 8, 0. 9, or 1. 0 km Hard constraint: cannot cross the skiway Strings 1 -7, 14, 22, 31, 79, 80 are displaced Other constraints now known from LEC are ignored in parts of this talk

High Energy Ice. Cube Optimization Taking into account drilling and deployment constraints: drilling camp locations in the last 3 seasons and possible 1 -2 segment of hose extension. • original x 1 spacing • x 1. 5 spacing • x 2. 0 spacing • x 2. 5 spacing • consider removing 4 DOMs in the dust layer (DOMs 35 -38)

Analysis of the numu simulation noocut 45 o N >64 40 mx<20 ch o mx<20 21 mx<10 mx<5

Effective area (atm. nu + E-2 flux) 1. 0 km Trigger 30 20 10 5 2 degree level AR AR original NDL Event rate ratio to the original geometry 1. 5 x 2. 0 x 2. 5 x 5% 10% 15% 0. 8 km 25% 0. 9 km 38% 1. 0 km 50% Geometries: original, NDL, 1. 5 x, 2. 0 x, 2. 5 x, 0. 8 km, 0. 9 km, 1. 0 km

Combined performance with LE core by Darren Grant

SBM cut progression Both “angular resolution” and SBM were trained on a half of MC set, and then the other half of the MC set was studied using the developed cuts.

Remarks • A complete neutrino search for IC-80 HE extension geometries is presented • 10 -30 % increase in effective area at Pe. V energies is possible compared to default configuration • 10 -50 % increase in effective area at GZK energies is possible compared to default configuration • Background rejection is possible at the same levels as in the original geometry (verified at ~10% purity level)

Ice. Top acceptance with outriggers by Tom Gaisser Bakhtiyar Ruzybaev • Increased acceptance > 1. 5 for E > Pe. V • Acceptance for coincident events increases quadratically, > 2. 2 • Probably 3. 0 at E>100 Pe. V • Improved statistics ~ Ee. V • Other possibilities –One tank per station for “guard ring” 1 2 3

Example event Original geometry: 50 o deviation from mc truth

Example event 1. 5 x geometry : 30 o deviation from mc truth

Example event 2. 0 x geometry : 15 o deviation from mc truth

Example event 2. 5 x geometry : 9 o deviation from mc truth

Example event 0. 8 km geometry : 79 o deviation from mc truth

Example event 0. 9 km geometry : 21 o deviation from mc truth

Example event 1. 0 km geometry : 45 o deviation from mc truth

Bigger picture - Staged Ice. Cube Enhancements How to get from here to there? LHC Optical: 80 Ice. Cube + 13 Ice. Cube -Plus (astro-ph/0310152) holes at 1 km radius (2. 5 km deep) Radio/Acoustic: determine GZK event rates with 6 + 12 radio detectors at the surface or D. Besson et al. at depth astro-ph/0512604 calibration with Ice. Cube!

Summary • Options: – Easy drill configuration: Enhance rates by ~25% above 1 Pe. V – Difficult drill configuration: Enhance rates by ~50% above 1 Pe. V • Area bigger than baseline at almost all energies when combined with low energy core. • Potential for coincidences in radio and acoustic detection interfacing. • Bigger picture: Ice. Cube is an enormous multiplier for upgrade potentials. Comparably small investments allow huge detection potentials.

Deployment considerations by Jeff Cherwinka • Last successful season demonstrates that the remaining 46 strings of Ice. Cube + LE core can be comfortably deployed in 3 seasons • LE core holes can be drilled with the default setup of the drilling hardware • HE ring holes will require changes to power, signal, and surface hoses. Increase drilling time due to hole move and lower flow rate. • At the large spacing of 311 m between holes, 5 holes can be in the normal drill range circle. Holes within the normal drill circle should be drilled with the nominal rate of 2 days per hole. • For holes outside the normal drill range circle we should budget 4 days per hole. This is adding one additional day for hole to hole move and one additional day for drilling at slower rate • Two 300 m extensions have been ordered this year • Final Layout of all moved holes needs to be made before Dec 2008 so adequate planning can be done 2008/9 Baseline has no moved holes. 2 A is +180 m 2009/10 would need to drill 0 to 3 moved holes. 51 A is +120 m 2010/11 would drill all moved holes. 1 A is +300 m

Background study To reject background an SVM approach used in my IC-22 analysis was tried. However, after training SVM with a RBF kernel and a variety of g and c parameter values background rejection power on the testing dataset turned out to be extremely poor. At 50% signal efficiency a purity of 2. 4% only could be achieved. At any signal efficiency the purity never exceeded 3. 3%. After some investigation the reason appeared to be under-training. A single background corsika event weights as much as 20% of signal at trigger level. It was possible to find a better machine learning method that would work reasonably well (well, better than SVM) using existing simulation statistics. In the following it is called SBM: Surface Boundary Method. It relies on the property of most cut parameters that to achieve the same quality the cut value can be relaxed as we go to larger Nch or Nstr or move away from the horizon (into more vertical muon directions). http: //icecube. wisc. edu/~dima/work/WISC/sbm/

Juliet UHE studies by Aya Ishihara An improvement of up to 73 -80% was possible to achieve in the number of UHE events Some of this might be due to somewhat changed/improved(? ) method, as the multiplicity 80 UHE threshold condition was evaluated only using the inner strings. Also, the outside strings of the default geometry were kept, rather than moved. However, this is not expected to change the results too much. muon effective area 400 m … 125 m Default Design 310 m