Detector Characterization in GEO 600 by Alicia M

Зарегистрируйтесь, чтобы просмотреть полный документ!

Detector Characterization in GEO 600 by Alicia M. Sintes-Olives on behalf of the GEO-DC team Universitat de les Illes Balears http: //www. aei. mpg. de/~sintes/GEO_DC/

Detector Characterization in GEO 600 Who we are • Kick-off by January 2001. Since then regular telephoneconferences (~2 weeks) and several workshops have been organized (to which external observers have been invited). • Today more than 40 scientists are on the GEODC mailing list from AEI, Hannover, Cardiff, Glasgow, Birmingham and UIB. • GEODC brings together both experimentalists and theorists. This is very important since detailed knowledge of the detector and also expertise on data analysis are required. • Members are involved in the LSC Upper Limit Working Groups. • Information on the group activities: http: //www. aei. mpg. de/~sintes/GEO_DC/ A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Detector Characterization in GEO 600 Purpose • Understand characterize the GEO-detector: – identify noise sources and misbehaviors, so that experimentalists can improve the detector and bring it to the design sensitivity – study noise propagation • Give support to the different astrophysical GW data analysis searches, e. g. , – providing vetoes for data analysis A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Detector Characterization in GEO 600 Scope Members of GEODC group are involved in: • Data acquisition, data transfer and storage. • Development of software tools to facilitate data access and analysis, including online monitors. • Development of new algorithms for data analysis. • Analysis of data from the GEO detector and auxiliary channels. • Data quality studies. Development of veto strategies. A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Detector Characterization in GEO 600 Data access GEO is continuously taking data which is transferred to Hannover and converted into frame-format files, including also trend information. GEO has performed several engineering runs, some of them in coincidence with LIGO and TAMA: E 7: 28 Dec 2001 - 14 Jan 2002 (duty cycle 75%) S 1: 23 Aug 2002 - 8 Sep 2002 (>1 TB of data, duty cycle 97%) Data have been transferred to AEI and Cardiff either by network transfer or tapes. Currently data can be accessed from: Hannover (only about 3 most recent days), Cardiff and AEI where there is full S 1 and E 7 data, all trend data, and h(t). Frame-servers are running at these locations all time. A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Detector Characterization in GEO 600 Software tools I To easy data access, data analysis and in order to have a more uniform software environment, several tools have been developed and are continuously updated: • GEO-Tools [M. Hewitson]: it has plenty of utilities to deal with GEO data (data access, filter capabilities, and provided plenty of mex files as an interface to Matlab). • Dataviewer [K. Kötter]: to display GEO data on a browser that is available on any of the servers. It can display time domain data, trend information, and (amplitude) power spectral (density) using different frequency resolution. It has zooming capabilities and it can use different axis scales. It has some functionality: windows, calibration… • Triana [developed at Cardiff]: to access data and data analysis products stored in My. SQL databases [D. Churches], and to perform A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002 data analysis.

Detector Characterization in GEO 600 GEO++ is a software centered around C++ classes. [R. Balasubramanian, S. Babak, D. Churches, S. Sathyaprakash, S. Mohanty, K. Kötter, S. Heng, A. Vecchio, R. Ingley, …] GEO++ provides an environment where a large quantity of data can be filtered through data analysis or detector characterization pipelines in real time. Functionality: – Frame File interface for GEO/LIGO data – Ability to distribute data over networked workstations using MPI – Record results in the form of database records, frame files, … – Power spectrum and power spectral density estimators – Time domain filters (IIR, FIR) – Resampling GWDAW, kyoto (Japan), December 17 -19, 2002 A. M. Sintes

Detector Characterization in GEO 600 GODCS: GEO Online Detector Characterization System • GODCS is based on GEO++. • It facilitates online data analysis, handling a large data rate. • It offers a flexible scheme to develop and integrate monitors into a pipeline and run several monitors at a time. • It has the capacity to chain together monitors, each monitor analyzing data modified by the previous monitors. • A flexible scheme to exchange numeric information between monitors developed by different people. Very useful when we want to develop monitors which act only when another monitor triggers an event. Monitors available: Saturation. Mon, Psd. Mon, Power. Tracker. Mon, Lock. Status. Mon, Inspiral. Mon, Glitch. In. Power. Line. Mon, PQMon, Frame. Channel. Summary. Mon, Excess. Power. Mon, TFCluster. Mon… A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Detector Characterization in GEO 600 DCR Detector Characterization Robot [S. Mohanty, S. Mukherjee] The aim is to construct a reliable and efficient automated system able to keep track of all change points in all relevant channels. See: Mohanty’s presentation: Controlling burst detection confidence in uncharecterized noise. Mukherjee’s presentation: Robust detection of noise floor drifts in interferometric data. A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Detector Characterization in GEO 600 Data analysis has been performed on E 7, S 1 data and, in particular, on 3 hours of playground data, under GODCS and Matlab + GEO-Tools Investigations: • Analysis of GODCS results (running online during S 1). • Lock losses and their likely causes (laser spikes, sudden misalignment, van passing near a building, earthquake in Italy mag. 6. 1…) • Excess power and glitches in different channels and if they couple into h(t), in particular, laser spikes. • Correlations between seismic channels and h(t). • Calibration [ see M. Hewitson’s talk]. • Automated Line detection. • Sensitivity to binary inspirals. • Analysis of inspiral injected signals using TFCluster. Mon and matched filtering. A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Detector Characterization in GEO 600 Data analysis cont. Many of the work has been carried out in the context of burst searches [chaired by K. Strain and S. Mukherjee] in S 1 data, with the aim of identify well understood veto channels. Results: • LSC_MID_EP-Q-HP should be very useful [see K. Kötter’s talk: PQMon: A powerful veto for burst events]. • Most of the large frequency and amplitude spikes in the laser system were not coupling (linearly) into h(t). • Several environmental monitors were not seen to give convincing vetoes for many fast glitches. • LSC_MID_EP-P under investigation. In the context of pulsar searches [led by M. A. Papa] there have been also investigations on power, phase and phase derivative distribution of ‘short’ FT. Results are posted on the GEO-DC lab-book. A. M. Sintes GWDAW, kyoto (Japan), December 17 -19, 2002

Скачать презентацию Detector Characterization in GEO 600 by Alicia M

6d5b1867020e1b741fabd55c110a6bc3.ppt

Количество слайдов: 11