Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.atnf.csiro.au/research/pulsar/orange10/pdf/GeorgeHobbsOrange2010.pdf Дата изменения: Wed Sep 29 12:00:59 2010 Дата индексирования: Mon Feb 14 00:01:56 2011 Кодировка: Поисковые слова: п п п

Developing a pulsar time scale
George Hobbs CSIRO Astronomy and Space Science george.hobbs@csiro.au


Overview
Collaborators: W. Coles (UCSD), Chen Ding) (NTSC), R. Manchester (CSIRO) + PPTA team · What will irregularities in a terrestrial time standard look like in our data? · Developing a pulsar time scale · Initial results

CSIRO. Gravitational wave detection


Timing residuals = unmodelled physical effects

Spin-down irregularities

No angular signature

CSIRO. Gravitational wave detection


Time standards
· Most pulsar observations referred to TT(TAI) · Post-corrected time standard TT(BIPM2010) can be used · Requirement to fit for the pulsar's period and first derivative => a quadratic must be removed from the expected clock errors

CSIRO. Gravitational wave detection


Terrestrial time standard irregularities

Monopolar signature

CSIRO. Gravitational wave detection


Basic idea
· Irregularities in terrestrial time standards will show up as residuals that are the same for different pulsars · Can find this correlated signal to: - identify any errors in the terrestrial time standards - correct for any such errors

CSIRO. Gravitational wave detection


What happens if irregularities exist in an Earth-based time-scale? TT(TAI)-TT(BIPM2010)

Different data spans Timing model fits Different sampling Varying error bars Unexplained timing noise Note: can never recover a linear (or quadratic) drift

CSIRO. Gravitational wave detection


What happens if irregularities exist in an Earth-based time-scale? TT(TAI)-TT(BIPM2010)

Different data spans Timing model fits Different sampling Varying error bars Unexplained timing noise Note: can never recover a linear (or quadratic) drift

CSIRO. Gravitational wave detection


What happens if irregularities exist in an Earth-based time-scale? TT(TAI)-TT(BIPM2010)

Different data spans Timing model fits Different sampling Varying error bars Unexplained timing noise Note: can never recover a linear (or quadratic) drift

CSIRO. Gravitational wave detection


What happens if irregularities exist in an Earth-based time-scale? TT(TAI)-TT(BIPM2010)

Different data spans Timing model fits Different sampling Varying error bars Unexplained timing noise Note: can never recover a linear (or quadratic) drift

CSIRO. Gravitational wave detection


What happens if irregularities exist in an Earth-based time-scale? TT(TAI)-TT(BIPM2010)

Different data spans Timing model fits Different sampling Varying error bars Unexplained timing noise Note: can never recover a linear (or quadratic) drift

CSIRO. Gravitational wave detection


Technique
· Define clock function to be simple Fourier expansion:

f (t) =



Ak cos( k 0 t )+ Bk sin( k 0 t )

(note: can use other functional forms if needed) · Carry out a standard least-squares fit of pulsar timing model parameters + f(t) as usual, except: - simultaneously fit to multiple pulsars - use measurement of the covariance in the residuals for a given pulsar as part of the least-squares-fit fit (to deal with timing noise)

P

est

= (M C M) M C R
T T

-1

-1

-1

Timing residuals

Covariance matrix of the residuals
CSIRO. Gravitational wave detection

Pulsar timing model


Testing: can we recover TAI-TT(BIPM2010) x 10?
· Simulate 10x expected TAI-TT(BIPM2010) in real pulsar data

5s

CSIRO. Gravitational wave detection


Testing: can we recover TAI-TT(BIPM2010) x 10?
· Simulate 10x expected TAI-TT(BIPM2010) in real pulsar data

5s

CSIRO. Gravitational wave detection


Testing: can we recover TAI-TT(BIPM2010) x 10?
· Simulate 10x expected TAI-TT(BIPM2010) in real pulsar data

5s

CSIRO. Gravitational wave detection


Testing: can we recover TAI-TT(BIPM2010) x 10?
· Simulate 10x expected TAI-TT(BIPM2010) in real pulsar data

5s

CSIRO. Gravitational wave detection


Final result (no simulations) EPT-TT(TAI) and TT(BIPM2010)-TT(TAI)

1s

CSIRO. Gravitational wave detection


Final result (no simulations) EPT-TT(TAI) and TT(BIPM2010)-TT(TAI)

1s

CSIRO. Gravitational wave detection


EPT-BIPM(2010) ­ time transfer?

CSIRO. Gravitational wave detection


EPT-BIPM(2010) ­ time transfer?

PKS->TID>UTC(NIST)->TT

PKS->GPS->TT(TAI)

CSIRO. Gravitational wave detection


Future improvements
· · · · Adding more data sets Adding more recent data Producing y and z stability plots Compare results from different observatories to distinguish between time transfer errors from TT(BIPM2010) errors · Correct the pulsar timing residuals using EPT

CSIRO. Gravitational wave detection


Summary
· Can recover recent deviations between TT(BIPM2010) and TT(TAI) using pulsar observations · Have significant deviation from TT(BIPM2010) prior to the year 1999 · Can not (currently) distinguish between errors in TT(BIPM2010) and errors in the time transfer from the Parkes observatory · New data sets should significantly improve the results · New pulsar discoveries and improved observing techniques are significantly improving the precision with which pulsars can be timed. · Pulsars may be able to provide confirmation/addition to Earthbased timestandards on timescales of years and decades. · Contact: george.hobbs@csiro.au
CSIRO. Gravitational wave detection