Volume Trigger Proposal for the 2011 Season for horizontal low Energy events

Transcription

1 October 22, 2010 IceCube Filter Proposal Document Name: M. Danninger Physics WG: WIMPs Physics WG Leader: Carlos de los Summary Volume Trigger Proposal for the 2011 Season for horizontal low Energy events (supplement from IC79 test runs (VolumeTrigger CylinderTrigger)) This trigger proposal is devided in two separate parts. The physics motivation for the VolumeTrigger, as it was accepted by the TFT for the 2010 season, remains unchanged from last year and is attached at the end of the document. The first part analyses the last IC79 test run of the CylinderTrigger. This run is scrutinized in detail and all problems with the first DAQ implementation seem to be found and rates appear to be well understood. Additionally, it highlights the changes that were necessary in order to implement the trigger into the DAQ. In detail the transition from VolumeTrigger to CylinderTrigger. VolumeTrigger (VT) CylinderTrigger (CylT) The VT looks into configurable predefined volumes for a simple majority of HLC hits, e.g. in the IC79 case SMT4. The selected volume element is shaped like a cylinder. The cylinder plane includes one layer of strings around a center string. The cylinder height is 5 DOM-layers (2 up and down from the selected DOM). As this implementation of a predefined list of volumes is not possible to integrate into the DAQ, the algorithm was changed to simply calculate a cylindrical volume around an HLC-DOM with corresponding choice of radius and height. This implement should match the result of the initial VT with some very small deviation. Tests with the icetray implementations of VT and CylT 1 show a difference of less than 1 in triggers. (this is a retrigger module; work is currently in progress to also provide trigger-sim module.) 15 min Test run on the 8th. September Run was a 15 minutes CylT test run. This test run showed an unexpected high trigger rate. In order to identify the later discovered problems, the raw data was decoded and retriggered with the VT and CylT icetray modules. Global trigger rates and exclusive CylT rates are listed in table 1. The first problem was identified by looking at the exclusively triggered events after applying the CylT retrigger with Table 1: Summary of global and exclusive trigger rates. Trigger Cyl.Radius Cyl.Height Global Rate Exclusive Rate Configuration in (m) in (m) in (Hz) in (Hz) IC79 test run (Cyl.T) (incl. 2 problems) IC79 test run (Cyl.T) (incl. 1 problem) IC79 test run (Cyl.T) (fixed) IC86 scaled from IC79 (Cyl.T) (fixed) CylinderTrigger icetray implementation can be found under/sandbox/hias/filter 2011/ 1

2 the same settings as were used in the DAQ during the test run. In a matching implementation, these events would have been not triggered. Figure 1 (a) with (d) show two example events. It can be seen in the string occupancy plots (1 (a) and (c)) that the radial criteria within the DAQ-CylT has worked. All shown strings are neighboring strings. At the same time the DOM occupancy plots (1 (b) and (d)) suggest that the height criteria must have been faulty implemented, as none of the configured cylinder heights would have been large enough to enclose 4 HLC-DOMs. The problem in the DAQ implementation was identified and fixed. These exclusive examples also demonstrate that none of these events are expected to trigger the proposed CylT and should therefore have no influence on the IC86 global trigger rate. The second problem was identified much easier as it was a simple misconfiguration of the cylinder height. The height was configured to 9 DOM-layers in total (4 up and down from the selected DOM) rather than 5. In conclusion, the IC79 test run would have had an exclusive CylT rate of 309 Hz, if all would have been implemented (a) string occupancy event 1 (b) DOM occupancy event 1 (c) string occupancy event 2 (d) DOM occupancy event 2 Figure 1: 2 example events that are exclusive to the correctly working CylT, but were still triggered in test run. Event 1 in (a) and (b) and event 2 in (c) and (d) 2

3 correctly. This is within the expected range from IC77 MC prediction (details in table 1). Additional Investigations For a more complete study on the CylT, figure 2 illustrates the individual contributions from different in-ice triggers to the HLC-DOMLaunch distribution and the number of string distribution. The CylT is shown for two different settings (expressed in number of DOM-layers and SMT threshold, e.g. (5/4)). Additionally, the impact of the recently discovered HLC-problem within the DAQ is investigated within this test run. The problem was reported by Chang Hyon Ha and shows individual HLC-DOMs on strings. Figure 3 illustrates this problem within test run One can conclude that it is present, but affects the CylT rate less than the DeepCore SMT3 rate. (a) HLC DOMLaunches (b) nstrings Figure 2: Distributions of number of HLC-DOMLaunches (a) and number of hit strings (b) for all in-ice triggers and 2 different CylT settings.(expressed in number of DOM-layers and SMT threshold, e.g. (5/4)) Figure 3: Distribution of number of hit strings of CylT triggers for 3 different types of events. Events with HLC = 4, 5 and 6. Conclusion Another test run will be taken in the nearby future that includes all the fixtures to the CylT. This run is expected to have an exclusive CylT rate that matches the currently predicted value of 310 Hz. 3

4 October 28, 2009 IceCube Filter Proposal Document Name: M. Danninger Physics WG: WIMPs Physics WG Leader: Carlos de los Volume Trigger Proposal for the 2010 Season for horizontal low Energy events Introduction This trigger is designed to capture low energy horizontal events that would not trigger either the nominal multiplicity trigger or a string trigger. Such events are expected to be important for low energy analyses such as solar wimp. This document motivates the usage of a Volume Trigger for the 2010 season and describes the algorithm, provides a status of the software, and gives rate estimates for a range of configurations. Figure 4: Improvement in the effective Volume for solar WIMP searches in % for the proposed VT configuration for soft (dashed) and hard (solid) WIMP models. In the upcoming 2010 season, DeepCore, IceCube s low energy extension, will be fully operational for the first time. Despite the therefore guaranteed improvement in the low energy range for solar WIMP analyses, the IceCube array still instruments a much larger detector volume. The Volume Trigger (VT) proposes a low multiplicity threshold trigger for IceCube strings only, in order to fully utilize the combined Ice- Cube + DeepCore neutrino detector array for low energy solar WIMP searches. The VT will improve the effective Volume at trigger level for low mass WIMP models in the order of 200 %. (see Fig.4) There is an ongoing effort called Track Engine (TE) to improve IceCube s low-e trigger efficiency in general, by using all hits instead of only Hard Local Coincident (HLC) hits. The TE study, shown as at the Berlin collaboration meeting 2009, indicates an even more significant improvement in trigger efficiency, independent from the number of generated HLC hits. However, its implementation is still under way so the WIMP group proposes for 2010, the use of the already existing, studied and tested VT. Furthermore, the implementation of the VT into the 2010 InIceTriggers would significantly decrease the impact of a currently discussed rise in the standard IceCube SMT threshold on low-e analyses. VT was developed by the Penn State group in Algorithm The trigger is intended to run in the InIceTrigger. The algorithm is split in two parts. First, a low-threshold time coincidence requirement is made, using the same logic as the SimpleMajorityTrigger (SMT). For 4

5 Table 2: Exclusive and inclusive single corsika rates for all investigated VT settings. Settings as proposed by the WIMP group are indicated in bold-face. Volume Trigger Inclusive Rate Exclusive Rate avg. Number Height Threshold in (Hz) in (Hz) Hits events that satisfy the timing condition, a positional analysis is applied. Here, each hit is used as the center of a pre-defined volume element and a look-up table provides a list of the other DOMs composing the element. If an element is found to contain enough hits, the trigger is satisfied. The trigger has several configurable parameters: a time window, a trigger threshold, and a volume height. The volume element is simply a flat disk containing of 7 strings, 1 central and 6 nearest neighbors. The height of the disk is configurable in step sizes of DOM spacings by the volume height parameter. The Volume Trigger module is in the DomTools project. The most current release is found under: releases/v /private/domtools/i3volumetrigger.cxx Studies and Results The here described study was done in order to match the 2009/2010 deployment plan of 77 IceCube and DeepCore strings. Therefore, the proposed 71 IceCube strings were used to evaluate trigger rates and signal sensitivities. For the evaluation of the single corsika trigger rates, 1000 files of dataset 1509 were used. The configurable time window parameter was set for all studies to 1µs. The VT SMT threshold was varied between 4, 5 and 6. The volume height of the volume element was investigated for 3, 5 and 7 layers of DOMs. Table 2 shows for the discussed parameters the determined inclusive as well as exclusive trigger rate and the average number of HLC-hits of the exclusively triggered corsika events. Figure 5 and 6 show the effective Volume for WIMP models for the different trigger settings, as well as a comparison to the current trigger settings. For additional information on the VT, see the 2008 VT-proposal Conclusions The WIMP working group proposes the implementation of the VT for the 2010 season with the settings Vol 5/Trig 4. This additional trigger will significantly increase the trigger efficiency of solar WIMP signals, as seen in Figs 5 and 6. The expected exclusive single Corsika trigger rate of 267 Hz is not insignificant, but could be possibly handled by the improved DAQ system (indicated by TFT). In the case that the DAQ is overloaded with the additional VT trigger rate, the alternatively proposed VT setting is Vol 5/Trig 5, with a very low exclusive rate of approximately 30 Hz ( 2% of global rate). The backup VT-setting has a much lower improvement in signal efficiency at trigger level compared to the first choice of only 30 to 50 % for the lower masses. A continuative filter proposal for the VT triggered events, using the 2010 low-up-filter, is in preparation, in order to remove the additional exclusive Corsika rate in the online filter chain. 5

6 Figure 5: Effective Volume at trigger level for the most interesting VT settings for the hard WIMP(Neutralino) models. VT lines are shown in color and include the total efficiency of SMT8 + StringTrigger + VT. For comparison, the SMT8 + StringTrigger (solid-black) and SMT8-only (dashedblack) lines are shown. Figure 6: Effective Volume at trigger level for the most interesting VT settings for the soft WIMP(Neutralino) models. VT lines are shown in color and include the total efficiency of SMT8 + StringTrigger + VT. For comparison, the SMT8 + StringTrigger (solid-black) and SMT8-only (dashedblack) lines are shown. 6