Jonah Kanner Mon, Apr 6, 2015 at 11:09 PM To: "Klimenko, Sergey" Cc: "francesco.salemi@ligo.org" , Ik Siong Heng , Eric Chassande-Mottin , vedovato , Marco Drago , Jeff Kissel , joseph.betzwieser@ligo.org, Craig Cahillane , Madeline Claire Wade , Xavier Siemens Hi Jeff, Thank you for considering the issue of high frequency calibration. I had some e-mail exchange with Sergey and the cWB team. There were some studies done to assess calibration needs for the 2-5 kHz band. To maintain the ability to localize sources on the sky, the team found 10% amplitude / 10 degrees in phase was needed. But, for detection purposes, they found 20%/20 deg was sufficient. This may be the reason there was some confusion about the requirement in this band. The high frequency search will *not* be run in low-latency in O1, so it will not trigger EM follow-up, and so I think sky localization is not a major driver of requirements for this search. My suggestion is that for O1, 20% in amplitude / 20 degrees in phase is a great goal for the uncertainty in the 2-5 kHz band. By O2, our goals may be more ambitious, and we would ask for 10%/10 deg uncertainty for O2 and beyond. Of course, I am happy to talk about this in more detail with you any time. Thank you for your help! Best, jonah Burst group (Jan 31st 2013) Dear Jonah, Thank you for your reply. We are indeed looking for updated feedback on those numbers. You help would be much appreciated. Take care, Xavi On Thu, Jan 31, 2013 at 1:07 PM, Jonah Kanner wrote: > Hi Xavi, > > I realize I'm a bit late on this, but this e-mail was recently brought to my > attention. Are you still looking for feedback on the burst group > requirements? If needed, I could take a look at this. > > I was likely involved in creating the 5 degrees, 50 microsecond spec (those > numbers sound familiar!). Probably, if I took a second look, I would come > to a similar conclusion. > > Best, > jonah > > -------- Original Message -------- > Subject: [DAC] calibration uncertainty requirements for the advanced > detector era > Date: Tue, 4 Dec 2012 16:17:33 -0600 > From: Xavier Siemens > Reply-To: LSC Virgo Data Analysis Council > To: LSC Virgo Data Analysis Council > CC: LSCcalibration > > Dear Search Group Chairs, > > Two years ago we asked for calibration requirements for advanced > instruments. The response we got was the following: > > Amplitude Phase Timing Latency > Burst 10% 5 deg. 50 us ~10s > CW 10% XXX 30 us XXX > CBC few % 5-10 deg. 16us@1kHz XXX > Stochastic XXX XXX XXX XXX > > In earnest preparation for aLIGO, I was hoping you could revise these > numbers and fill in the Xs where possible. > > Thanks, > > Xavi CW group Hi Xavi, In the CW group we think it makes sense to distinguish between calibration requirements used in setting upper limits and making detections vs those we want in the event of a detection, where we are trying to make the most of our GW observations and to correlate them with electromagnetic observations. So we list requirements below under different scenarios. We also list separately our calibration latency requirements for first-look analyses of new data and what we want for published analyses. Finally, we note that we might be wanting even better precision in the event of a high-SNR detection. cheers, Keith, Graham & Andrzej ============================================================================ First-look analyses and no detections (driven by targeted searches to ~1500 Hz) Amplitude Phase Timing Latency CW 10% 25 deg 50 us 1 month Published searches and no detections (driven by targeted searches to ~1500 Hz) Amplitude Phase Timing Latency CW 10% 25 deg 50 us 6 months In the event of detection of a signal (driven by all-sky searches to ~2000 Hz) Amplitude Phase Timing Latency CW 10% 18 deg 25 us 6 months Note: If we were to detect a CW source in a semi-coherent all-sky search and successfully zoom in using a coherent search over a long observation time, we could imagine achieving SNR values of 50-100 or even higher, In that event calibration precision would become very important. Given the large uncertainties in predicted gravitational wave strengths of sources (including ~(factor of 2) uncertainty from the neutron star moment of inertia), we are less concerned about amplitude uncertainty than about phase uncertainty, where temporal correlation with radio, X-ray and gamma-ray (and even optical) signals could yield insights into NS astrophysics and enable tests of general relativity. So it is likely that in the happy event of a CW detection with very high SNR, we would be requesting still better precision than listed above ============================================================================ On Tue, Dec 4, 2012 at 5:17 PM, Xavier Siemens wrote: Dear Search Group Chairs, Two years ago we asked for calibration requirements for advanced instruments. The response we got was the following: Amplitude Phase Timing Latency Burst 10% 5 deg. 50 us ~10s CW 10% XXX 30 us XXX CBC few % 5-10 deg. 16us@1kHz XXX Stochastic XXX XXX XXX XXX In earnest preparation for aLIGO, I was hoping you could revise these numbers and fill in the Xs where possible. Thanks, Xavi