SudarshanK, TravisS
Travis and Darkhan took Pcal calibration measurements at the end station last week (alog 22489). Using the gps time information from that alog we have calculated the new pcal calibration factors. The summary is that the calibration at Y-End is pretty close to what we measure last time (Aug 27) however, the calibration at X-end, atleast for RxPD, has changed significantly due to clipping of one of the pcal beam. The clipping issue at X-end is discussed in detail on another recent alog by Rick (alog 22529).
Attached is the calibration factors from each endstation compared to the most recent calibration. This report includes the parameters used and relevant intermediate numbers as well. A summarized report of the final calibration numbers is uploaded to the DCC document (T1500252). The calculation of TxPD factor and its uncertainty on X-end will have to be done little differently in light of the clipping issue. For now, I just reported what is calculated by the old code but I will work on implementing a more accurate representation.
J. Kissel In order to get the latest updates to GWINC inspiral range calculation authored by John Miller and Salvo Vitale @MIT, I've updated the local checkout of gwinc on the workstations in /ligo/svncommon/IscSVN/iscmodeling/trunk/gwinc/ corner of the Isc SVN repo: /ligo/svncommon/IscSVN/iscmodeling/trunk/gwinc$ svn up A redshift_to_dist.m A dist_to_redshift.m U precompIFO.m U int73.m U gwinc.m U IFOModel.m A calculate_horizon.m A int73_FminsearchVersion.m U SourceModel.m U int73_GWINCV3.m Updated to revision 2719.
VAC: Continue leak checking at mid Y AUX cart is still running at end X
TITLE: 10/19 [DAY Shift]: 15:00-23:00 UTC (08:00-16:00 PDT), all times posted in UTC STATE Of H1: Observing @ ~ 76 MPc. OUTGOING OPERATOR: Jeff QUICK SUMMARY: From the cameras the lights are off in the LVEA, PSL enclosure, end X, end Y and mid X. I can not tell if they are off at mid Y. Riding out end of earthquake. Microseism has come down. Winds are less than 5 mph.
Activity Log: All Times in UTC (PT) 00:00 (00:00) Take over from Ed 10:53 (03:53) ETM-Y saturation 11:08 (04:08) ETM-Y saturation 12:37 (05:37) Reset timing error on ETM-X 15:00 (08:00) Turn over to Patrick End of Shift Summary: Title: 10/19/2015, Owl Shift 07:00 – 15:00 (00:00 – 08:00) All times in UTC (PT) Support: None needed Incoming Operator: Patrick Shift Summary: Quiet shift – IFO locked at LOW_NOISE for the past 10 hours. 80Mpc range. Intent Bit set to Observing all shift. Winds calm all night. Seismic activity low. There was a 5.8 mag EQ in the Philippines, R-Wave due at LHO ~14:43 (07:43). No apparent problems with the EQ at LHO.
A quiet shift for the first 4 hours. The IFO has been locked at NOMINAL_LOW_NOISE, 22.5W, 80Mpc for the past 6 plus hours. The Intent Bit is set to Observing. Wind and seismic activity are low. There has been one ETM-Y saturation during the shift.
During ER8, there was a calibration artifact around 508 Hz - a non-stationary peak with a width of about 5 Hz. The peak went away on Sep 14 16 UTC probably due to an update of the calibration filters which was documented in this alog. When re-calibrated data is produced, it's worth having a look at some of this ER8 time to check that the peak is removed. I made a comparison spectrum a bit before and after the change of the filters (plot 1). The wide peak is removed and the violin modes that it covers (ETMY modes, maybe some others) appears. I did the same thing for a longer span of time, comparing Sep 11 and Oct 17 (plot 2). The artifact manifests itself also as an incoherence between GDS-CALIB_STRAIN and OMC-DCPD_SUM (plot 3). The only other frequency where these channels aren't coherent is at the DARM_CTRL calibration line at 37.3 Hz. I've also made a spectrogram (plot 4) of the artifact. It has blobs of power every several seconds. The data now looks more even (plot 5), though it's more noisy because the calibration lines are lower.
I agree that it was due to bad digital filters in CAL-CS. See my recent investigation on this issue at alog 22738.
Title: 10/19/2015, Owl Shift 07:00 – 15:00 (00:00 – 08:00) All times in UTC (PT) State of H1: At 00:00 (00:00) Locked at NOMINAL_LOW_NOISE, 22.5W, 81Mpc. Outgoing Operator: Ed Quick Summary: Environmental conditions are good. IFO in Observing mode. All appears to be normal.
TITLE: Oct 18 EVE Shift 23:00-07:00UTC (04:00-00:00 PDT), all times posted in UTC
STATE Of H1: Observing
SUPPORT: Jenne, Keita, Kiwamu
LOCK DURATION:
INCOMING OPERATOR: Jeff B
ACTIVITY LOG:
03:05UTC Strange glitch in DARM from 100Hz down. DIdn’t seem to correlate with ETMY or 45MHz RFAM.
03:26 Lockloss
03:40 DRMI locked in split mode (pitch)
03:41 Increased AOM diffracted power from 4.5% to 7.5%
03:50 Stopped at ENGAGE_ASC_PART3 to watch the StripTool show. After it converged I decided to move on but Guardian was stalled there. I switched to manual mode, selected the next stage and when it started moving I put it back into Auto and selected NOMINAL_LOW_NOISE. It seems to be fine now and we’re almost all the way back up.
04:03 ISS Second loop failed to engage. Wiki bailed me out. I set IMC_LOCK Guardian to Manual mode and selected ISS_ON. (Then back to Auto). ISC_LOCK Guardian continued.
04:05 According to the ISS MEDM screen it doesn’t appear that the Second loop actually engaged??
04:50 Livingston called to tell me they are having trouble with their PSL. Matt is heading to the site to investigate. Brian O’Reilly said it may have to wait until tomorrow.
04:52 after a couple of phone calls, Jenne helped me clear the ASC and PR3 Diffs. Kiwamu talked me through manually engaging the second loop.
04:52 Back to Observing...80Mpc
SHIFT SUMMARY: IFO was locked and doing fine until 03:26UTC. Perhaps it was my impatience at ENGAGE_ASC_PART3 that caused my woes upon relocking. uSe is trending downward, still, at about .2 microns and Eq graph is nominal. Winds have been calm and there was a small bit of glitching due to saturations.
SUS E_T_M_Y saturating (Oct 19 02:35:03 UTC)
SUS E_T_M_Y saturating (Oct 19 02:35:05 UTC)
SUS E_T_M_Y saturating (Oct 19 03:19:32 UTC)
SUS E_T_M_Y saturating (Oct 19 03:26:06 UTC)
SUS B_S saturating (Oct 19 03:26:06 UTC)
DRMI Unlocked (Oct 19 03:26:06 UTC)
SUS E_T_M_Y saturating (Oct 19 05:57:58 UTC)
SUS E_T_M_Y saturating (Oct 19 06:35:05 UTC)
C. Cahillane I have completed the preliminary uncertainty analysis of the LLO detector for O1. Joe Betzwieser was kind enough to provide the measurements and modify the directory infrastructure to match LHO, making my job easier. Thanks Joe! Currently, the budget only includes the Sept 14 LLO Sensing measurements and the Sept 3 LLO Actuation measurements. I know that additional measurements have been taken in the meantime, and it should be simple to include those as soon as I know where they are/what they contain. Some changes were necessary: The LLO L1 Actuation stage had a relatively short frequency vector, which is fine since its influence falls as 1/f^6. The shortness of the frequency vector made the systematic fit unusual (Plot 10). I also had to change the interpolation method from 'spline' to 'linear' for the LLO Actuation stages because interp1's extrapolation wasn't working properly on both ends of my frequency vector. (My freq vector is more dense than the measurement frequency vectors.) Plots 1-4 show the Nominal O1 model, which includes systematic errors added quadractically with statistical uncertainty. Plots 5-8 show the Systematic-corrected model, which includes only statistical uncertainty. Plot 9 is the comparison between the Nominal and Systematic models and their uncertainty bars. Plots 10-12 show the LLO L1, L2, and L3 Actuation stages and their systematic fits. Plot 13 shows the LLO Sensing Measurement compared to the Sensing Model at the time. Plot 14 shows the Sensing systematic fit Next step is to improve the kappa uncertainties! They are still hard-coded to be three percent and three degrees. I think a full calibration group discussion is necessary to make sure everyone agrees on how to propagate this particular uncertainty/error.
09:14UTC 80Mpc
@ 03:26:06UTC. Hard to say it was earthquake related. The eq band had only risen to barely .1micron but there was an ETMY and a BS saturation immediately before.
03:05UTC Strange glitch in DARM from 100Hz down. DIdn’t seem to correlate with ETMY or 45MHz RFAM. This happened between the 0 and 5 minute mark in DMT Omega plot.
Omegascan: https://ldas-jobs.ligo-wa.caltech.edu/~jordan.palamos/wdq/H1_1129259114.5/
Looks to me like it belongs to class number 4 from detchar's known glitch class document https://dcc.ligo.org/DocDB/0119/G1500642/017/aligo-glitch-classes-Oct2015.pdf
MID-SHIFT SUMMARY: Making this summary a few minutes early. Just got a report of a 5.3 event near Taiwan that’s supposed to roll a .47µm/s R- wave through here in about 20 minutes. So far I’ve only seen one glitch and that seemed to be coincidental with LLO coming out of lock. Wind has been calm and µSeism was starting to trend downward slowly. I may be starting to see early signs of .03-.1Hz band z-axis starting to increase. Otherwise, everything has been just peachy!
TITLE: Oct 18 EVE Shift 23:00-07:00UTC (04:00-00:00 PDT), all times posted in UTC
STATE Of H1: Observing
OUTGOING OPERATOR: Patrick
QUICK SUMMARY:IFO in Observing mode ~76Mpc. All lights in LVEA, PSL and M/E stations are off. Wind is ≤10mph. EQ sei plot is back to nominal after quake that caused most recent lockloss. Microseism plot Z-axis is elevated .25microns/s. CW injections are running. Calibration lines are on.
TITLE: 10/18 [DAY Shift]: 15:00-23:00 UTC (08:00-16:00 PDT), all times posted in UTC STATE Of H1: Locked. Observing @ ~76 MPc. SUPPORT: None SHIFT SUMMARY: Recovered from earthquake in Tonga. Seismic in 0.03 - 0.1 Hz recovered from earthquake but then trended up to .1 um/s at end X and end Y. Seismic in 0.1 - 0.3 Hz appears to be settling between .1 and .8 um/s. Can hear what may be the wind on the side of the building in the control room. Came up to about 20 mph and is now decreasing. INCOMING OPERATOR: Ed ACTIVITY LOG: 16:16 - 16:22 UTC Stepped out of control room 19:17 - 19:30 UTC Stepped out of control room 21:33 UTC Vern taking family to X arm overpass
Andy, Josh We've developed a tool to keep an eye out for problems in the SUS coil driver electronics like with PRM M3 (alog). When a coil driver goes bad, it can be hard to work out which one is the culprit because the feedback makes many degrees of freedom glitch. But the bad coil will have glitches that are not in the drive signal, hence the coherence between the analog readback and the drive signal will drop. This tool plots the spectra of the NOISEMON and MASTER_OUT channels, and their coherence, for the four quadrants of many different SUS channels. From these, it should be possible to check if there are any unexpected lines in the coil driver output, and whether it's not correctly following the drive signal. It's also possible to check the health of the noisemons themselves. The result is attached for Oct 17 8:00 UTC. We'll be looking through this and following up on any problems. The code doesn't yet deal nicely with channels that have no drive signal. Those will just be left blank. The code is Python and requires gwpy; it is available on github as noismon_check.
Moved ETMX pitch from 2.4 to 2.7 to lock the X arm on green. Found IR by changing ALS DIFF from -900 to -3126. Guardian found it in later attempts. Had trouble staying locked on DRMI. Went from LOCK_DRMI_1F to LOCK_PRMI. Had trouble locking on PRMI. I think I started to move the BS and PRM. Locked on PRMI. Moved BS and PRM. Went to LOCK_DRMI. Stopped at DC readout and increased the PSL ISS diffracted power from ~6% to ~8.6% by changing the REFSIGNAL from -2.04 to -2.02. (Was getting diffracted power is low messages) Ran the a2l script. Back to observing mode. 0.03 - 0.1 Hz seismic band has still not completely recovered from the earthquake. Is around .02 - .1 um/s. 0.1 - 0.3 Hz seismic band has trended up to around .1 - .3 um/s. Range is around 76 MPc.
The attached plots show the inspiral range integrand, and cumulative integral, for a stretch of recent H1 strain data. This is just the standard integration of the strain noise power, weighted as (frequency)^(-7/3). I was also interested in the impact the 35-40 Hz calibration lines have on the range calculation, so the plots include a cumulative integral curve for which the calibration lines have been artificially removed from the strain spectrum (the strain noise in the 35-38 Hz band was replaced with the average strain noise at nearby frequencies). These curves (magenta) show that the calibration lines reduce the range calculation just a bit -- by just less than 1 Mpc.
The inspiral range for the spectrum used is 75 Mpc. 90% of the total is accumulated by 150 Hz; the second plot thus shows the same data from 0-150 Hz. At the lower frequency end, 10% of the total range comes from the band 16-26 Hz.
Hi Peter, Andy pointed me to this post, indicating that this result shows we might want to filter from lower frequencies in the PyCBC offline CBC search. However, when we run our own scripts to generate the same result we don't see nearly as much range coming from the 20-30Hz band. Instead, we see only ~1% of the inspiral range coming from this band. Initially Andy had a script that agreed with your result, however I've convinced him that there was a bug in that script. I think that it might be possible that the same bug is also present in yours. I've attached a python script and a PSD from that time that should generate a relative range plot. I hope that it is clear enough to check if your scripts are doing the same thing.
Yes, indeed my script was making the error you allude to -- thanks for the correction. The integrand curves in my plots are correct, but the cumulative integral curves are not -- see Alex's plots for those. The corrected statements for Oct 6 - now somewhat obsolete due to reductions in the ~300 Hz periscope mount peaks - are that 90% of the range comes from the band 47 Hz - 560 Hz. About 1% of the range comes from frequencies below 29 Hz.