Документ взят из кэша поисковой машины. Адрес оригинального документа : http://zebu.uoregon.edu/~uochep/talks/talks02/G010396.pdf Дата изменения: Thu Sep 26 00:26:30 2002 Дата индексирования: Tue Oct 2 10:34:36 2012 Кодировка:

LIGO-G020253-00-Z

________________________
Intersite Environmental Transients
E5, E6, and E7 investigations

lightningphotography.com

Robert Schofield (Oregon) G. Gonzalez (LSU), M. Landry (LHO), P. Sutton (PSU),

________________________


________________________
Power grid details
· LHO and LLO sectors connected by 6 AC-DC-AC ties. · Both sides reference frequency to GPS · When a clock tied to the generated frequency deviates from a GPS clock by 10s (Eastern Interconnection) or 2s (Western), time error correction is initated. · Target of 60.02 Hz or 58.98 Hz implemented until time error is within 6s (east) or 0.5s (west), then restored to 60.00 Hz. · Time error correction is often initiated and terminated by telephone contact with power plants (variable delay). · Operators may postpone correction for several reasons. · Range from several corrections per day to once per month. · Time error correction proceedure is in a state of flux Concerns · Glitches may leak through AC-DC-AC ties. · Glitches or increased coherence may be produced by human and natural events that affect both sides (auroras, work schedules, lightning). · Long term coherence due to GPS frequency reference

________________________


________________________
An algorithm for detecting the presence of temporally correlated bursts
· Coincident event rates for aligned and offset LHO and LLO time series are compared. · Skeleton: Masahiro Ito's glitchMon · Signals that exceed sigma thresholds increment coincidence tables · Coincidence tables are incremented into the future to allow for propagation delays (about 15 ms is used here). · Multiple thresholds (and coincidence tables) to allow search for low-amplitude intersite events · Time series are output for high threshold events

Motivation · Know your enemy · Such bursts could produce coherence that drops off as LHO and LLO time series are misaligned · Event tool was not yet ready

________________________


________________________
Coincidence algorithm example
LHO MAG1 LHO MAG2 LHO MAG3
Hanford table

|0|0|0|0|2|2|2|2|2|2|0|0|0|0|0|0|0|...|0|0|0|0|0|0|0|0|0|0|0|
< ~15 ms >

LLO MAG1 LLO MAG2 LLO MAG3
Liv. table Liv. offsecond table

|0|0|0|0|0|0|0|1|3|3|3|3|3|2|2|0|0|...|0|0|0|0|0|0|0|0|0|0|0| |0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|...|0|0|1|3|3|3|3|2|2|0|0|

If corresponding bins in LHO and LLO tables each contain greater than, for example, 2 events, an intersite event is recorded, giving, for this example, 1 or 2 intersite events (depending on dead time): |0|0|0|0|0|0|0|0|e|e|0|0|0|0|0|0|0|...|0|0|0|0|0|0|0|0|0|0|0| and 0 off-second (control) events: |0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|...|0|0|0|0|0|0|0|0|0|0|0|

________________________


________________________
Greater than chance coincidence for coil magnetometers but not voltage monitors
LLO and LHO Z-axis coil magnetometers:
L1:GDS-EY_TO2; H0:PEM-COIL_MAGZ

Threshold (sigma) 7.5 6.3 5.1 3.9 2.7 38

Events exceeding threshold

Events for offset time series 22 80 355 1518 19095

(on - off) 16 39 102 374 1670

sqrt(on+off) 6.3 14 28 58 200

119 457 1892 20765

LLO and LHO line voltage monitors:
L0:PEM-EX_V1; L0:PEM-LVEA_V2; L0:PEM-LVEA_V1; H0:PEM-MY_V1; H0:PEMLVEA2_V2; H0:PEM-LVEA2_V1

Threshold (sigma) 16 13.1 10.2 5.85 2.95 27 45 73

Events exceeding threshold

Events for offset time series 29 40 68 132 30350

(on - off) -2 5 5 1 -32

sqrt(on+off) 7.5 9.2 11.9 16 246

133 30318

210630 s of E5 data; only 1 event allowed each second; only 1 signal in excess of threshold demanded from each site.


________________________
Only chance coincidence for seismometers and MC control
210630 s of E5 data; only 1 event allowed each second; only 1 signal in excess of threshold demanded at each site.

LLO and LHO seismometers:
H0:PEM-LVEA_SEISX; H0:PEM-LVEA_SEISY; H0:PEM-LVEA_SEISZ; L0:PEMLVEA_SEISX; L0:PEM-LVEA_SEISY; L0:PEM-LVEA_SEISZ

Threshold (sigma) 6 5.1 4.2 3.3 15 21

Events exceeding threshold

Offset time series events 10 15 96 14687

(on - off) 5 6 4 186

sqrt(on+off) 4.6 6 14 172

100 14873

LLO and LHO mode cleaner control signals:
L0:PEL1:IOO-MC_F; H2:IOO-MC_F1

Threshold (sigma) 6 5.1 4.2 3.3 2.85 61 91

Events exceeding threshold

Offset time series events 59 80 146 3176 24099

(on - off) 2 11 22 -16 69

sqrt(on+off) 11 13 18 80 220

168 3160 24168

________________________


________________________
Intersite coil magnetometer event
Posible lightning strike

Output from 560-580 Hz band: timing good to about 50 ms
fourth order butterworth filter 400 L1:GDS-EY_TO2 H0:PEM-COIL_MAGZ

200

Relative units

0

-200

-400

0

200

400

600

800

1000

Milliseconds after 681019841

________________________


________________________
Largest tabulated i/r during E4 (129kA, 20km)
Lightning strike on fluxgate magnetometer but not other channels.
Output from 560-580 Hz band: timing good to about 50 ms.
fourth order butterworth filter 400.0
L0:PEM-EX_MAG1X L0:PEM-EX_V1 L1:IOO-MC_F H0:PEM-COIL_MAGZ

200.0

Relative Units

0.0

-200.0

-400.0 0.0

200.0

400.0

600.0

800.0

1000.0

Milliseconds after 673744623

________________________


________________________
Summary

· Stand alone code has been developed to search for correlated intersite events by comparing coincidences (15 ms window) for aligned and misaligned time series.

· Detected high rates of intersite bursts on coil magnetometers (probably lightning)

· Did not detect intersite bursts for votage monitors, seismometers, MC_F, fluxgate magnetometers

· No evidence yet of lightning strikes on other channels

________________________