Environmental studies in Virgo

1
Environmental studies in Virgo

• Isidoro Ferrante
• For the Virgo collaboration

2
Talk summary

• This talk is a review of the status of
  environmental data analysis in Virgo.
• It will concentrate on seismic, acoustic and
  magnetic noise, and to their influence on dark
  fringe signal.
• To this effect, data taken during the C5
  commissioning run have been used
• 5 hours at the beginning of the run (GPS time
  786075317) are used for the study of low
  frequency contribution from seismic noise
• The interferometer was in science mode in
  recombined configuration.
• The first hour alone has been used for the
  studies about magnetic and acoustic noise.
• Others more recent data have been used whether
  possible.

3
The Virgo Enviromental monitoring system

• Slow monitoring
• 160 temperature probes inside and outside the
  vacuum chamber
• 11 humidity probes
• 11 pressure probes
• 1 weather station
• 1 lightning detector
• Fast monitoring
• 13 vertical accelerometers
• 5 triaxial low frequency accelerometers
• 1 velocimeter
• 9 magnetometers
• 6 microphones

4
Seismic noise
5
Seismic noise EpiSensors and accelerometers

• Episensors
• Triaxial force-balance accelerometers
• Frequency range from 0.2 Hz to 200 Hz
• Accelerometers
• vertical piezo accelerometers.
• Useful frequency range from 1 Hz to 4 kHz
• Episensors are much more sensitive in the low
  frequency region, (0.2-10 Hz) were seismic noise
  contributes to the mirrors motion
• In addition, vertical and horizontal
  accelerometers are placed on the top of the
  inverted pendulum, with a sensitivity band
  starting at about 10mHz
• One Guralp velocimeter is taking data since some
  month.

6
Seismic acceleration spectra

• Episensor spectra order of magnitude is the
  same, but the shape is different.

0.11 Hz Sea microseism
14 Hz Traffic
410 Hz Human activity
10200 Hz Electronic equipment
7
Coherence among seismometers

• The very low frequency region is likely due to
  sea microseism
• The region around 2 Hz is due to car traffic in
  the nearby highway

8
Correlation of seismic activity with dark fringe
noise

• Coherence between seismic acceleration and
  interferometer signal is significative in the
  region from 0.2 to about 1 Hz.

• At lower frequency, the episensor noise masks the
  coherence
• However, there is also a significant correlation
  also between the Central building accelerometers
  and the dark fringe, in the region around 50 and
  100 Hz, in very narrow lines.
• This region will be better studied using acoustic
  data.

9
Seismic contribution to dark fringe

• Using coherences, the seismic contribution to
  dark fringe noise at low frequency can be
  calculated.

• Horizontal movements, taking into account all
  three mirrors, dominate up to 0.5 Hz
• Vertical movements dominate from 0.5 to about 1.2
  Hz.
• The inertial control will be extended also to
  vertical direction to further reduce the seismic
  noise.

10
Acoustic noise
11
Acoustic noise

• Acoustic noise is correlated with seismic
  vibrations in the acoustic band.
• Microphones placed in the North and West end
  buildings show no correlation with the dark
  fringe signal.
• However, correlation can be seen between the
  microphone placed in Central building and the
  dark fringe, in the same frequency region were
  the coherence with episensors is high.
• The correlation is bigger with the microphones
  placed in the laser injection lab.
• Several lines, due to air conditioning and other
  electronics have been successfully identified.

12
Coherence and noise contribution
Acoustic noise contribution estimated using the
laser lab microphone
13
Acoustic tests

• To understand how noise enters the
  interferometer, a dedicated test has been
  performed during the C5 run placing a loudspeaker
  inside the laser lab driven with white noise or a
  frequency sweep.
• The white noise test did not give conclusive
  results however, pushing up the volume level we
  were able to de-lock the interferometer.
• The frequency sweep was much more interesting,
  since showed non-linearities in the noise
  propagation mechanism

14
Frequency sweep - up to 2kHz
Spectrograms during frequency sweep. In the dark
fringe one can see harmonics which are not
present in the microphone.
Zoom
15
Other signals - zoom to 250 Hz
Top angular errors Middle transmitted power
from Mode Cleaner and Reference Cavity Bottom z
error common mode
Non-linearities seems to affect mainly
transmitted power, as expected
16
Test result

• The most likely mechanism which could couple
  acoustic noise to the dark fringe is through
  laser angular jitter.
• Structure vibrations in the laser system produce
  an angular deviation of the beam axis.
• This modulates the intensity and the phase of the
  power at the mode cleaner output through a
  quadratic term.
• This mechanism could explain the harmonics
  observed in the dark fringe
• Power stabilisation, operative only after run C5,
  should reduce greatly this effect.
• We hope to solve this problem with the new
  injection bench.

17
The weather station

• A weather station has been installed on top of
  the Virgo central building.
• It provides atmospheric temperature, humidity,
  pressure and rain rate
• It measures also wind velocity and direction.

The effect of wind on mirrors motion has been
studied taking two different periods of no wind
and high wind in April.
18
Wind as seen from noise sensors

• The wind produces a wide-band (DC-100Hz) seismic
  and acoustic noise
• It is measured by seismometers and microphones
  in all buildings
• Low frequency velocimeter (Guralp) installed in
  the Central building measuring down to 10mHz
• The largest noise increase (below 1 Hz) occurs in
  the MC building

Central building
Mode Cleaner
North End
black no wind violet strong wind
West End
19
Wind effects on mirrors movements
IP motion

• Some noise reinjection
• around 200mHz

Mirror motion by LC
20
Magnetic fields
21
Magnetic Fields

• Three MFS-06 magnetometers from Metronix are
  installed in each building, in an orthogonal
  configuration
• Working range from 0.00025Hz to 10kHz
• Sensor noise
• 110-2 nT/ÖHz _at_0.01 Hz
• 110-4 nT/ÖHz _at_1 Hz
• 510-7 nT/ÖHz _at_1000 Hz

Magnetic spectra show mainly the presence of 50
Hz harmonics up to very high frequencies.
22
Correlation with interferometer signal

• Correlation with dark fringe signal is very high
  up to about 20 Hz

This is due to the magnetic field produced by the
marionetta coils! We can estimate the magnetic
field attenuation due to the vacuum tube.
23
From outside to inside

• To study the effect of external magnetic fields
  on the dark fringe signal a big coil has been
  placed near the wall of the west end tower

The field expected in absence of modification
from esternal structure is a line of 100 nT at 61
Hz. The test is still in progress.
24
The lightning detector

• Boltek LD-250
• record lightning strokes with about 1s precision.
• Provides azimuth and distance of lightning, up to
  about 450 km

25
Lightinings on magnetometers

• Near lightning can also be seen on magnetometers.
• No evidence, so far, of any effect on dark fringe.

26
Conclusion

• The main environmental disturbances which can
  interfere in Virgo operation are under control.
• Work is in progress to understand how the noise
  enters the interferometer, and to reduce the
  effects already known.
• Thanks to the works of Rosario De Rosa, Francesco
  Fidecaro, Irene Fiori, Lara Giordano, Federico
  Paoletti and many others Im forgetting right
  now....