Reference: ING Newsl., No. 9, page 26-27.
Article mirrored at: La Palma
server | Cambridge server
Other available formats: PDFPlanetPol: A High Sensitivity
Polarimetre for the Direct Detection and Characterisation of Scattered Light
from Extra-solar Planets
J. Hough1,* P. Lucas1, J.
Bailey2, E. Hirst1, M. Tamura3, D. Harrison1
1: University of Hertfordshire; 2: Macquarie University, Australia; 3: NAO, Japan.
After commissioning on the University of Hawaii
88-inch telescope, PlanetPol has been used successfully on the WHT in April
and October 2004. The instrument, funded by PPARC, was designed and built
at the University of Hertfordshire.
PlanetPol is a stellar polarimetre designed to measure fractional polarisations
of 10–6 or less. With this sensitivity PlanetPol should be capable
of detecting the polarisation signature of so-called hot-Jupiters. These
are extra-solar planets (EXP) whose size is approximately that of Jupiter
but with orbits that are 0.1 AU or less (orbital periods of a few days). The
linear polarisation should vary with phase angle from zero at full phase to
a maximum whose amplitude and position depends on the nature of the scattering
particles in the planetary atmosphere. Measuring the polarisation signature
not only gives a direct detection of the EXP, in contrast to the more usual
indirect detections by which most EXPs are discovered, but can provide information
about the planet’s albedo and radius, and on the nature of the scatterers.
Further, from the position angle of polarisation the inclination of the planet’s
orbit (i) can be determined thereby enabling the planet’s mass to
be determined. In contrast, techniques such as the RV method only measure
Figure 1. Left: Picture shows PlanetPol
on the WHT, with left to right: Edwin Hirst, Phil Lucas, Jim Hough, Dave Harrison
and Jeremy Bailey. [ JPEG | TIFF
]. Right: PlanetPol instrument. [ JPEG | TIFF ].
Polarimetry is a technique that is capable of very high sensitivity as
it is a differential technique that in principle is not affected by the Earth’s
atmosphere, and hence is limited only by photon noise. However, fractional
polarisations of a few parts in a million are lower than most astronomical
polarimetres can achieve, although comparable sensitivities have been obtained
before, albeit under somewhat idealised conditions. Kemp et al. (1987,
Nature, 326, 270) measured the integrated light from the
sun and gave an upper limit for the fractional linear polarisation of 2×10–7.
However, Kemp et al. used a polarimetre that directly viewed the sun, rather
than using an intermediate telescope, and hence avoided the potential problem
of telescope polarisation.
PlanetPol has a classical design and takes advantage of some of the techniques
pioneered by Kemp. It was designed for use on a range of telescopes, mounted
at the unfolded Cassegrain so as to minimise telescope polarisation.
Figure 2. Schematic of PlanetPol. [ JPEG | TIFF ].
All high sensitivity polarisation measurements to date have made use of
photoelastic modulators (PEM) in which a slab of non-birefringent material
is stressed using a piezo at the resonant frequency of the slab, f0,
thereby reducing the power needed to sustain a standing wave in the PEM.
Such devices are ideal as polarisation modulators as they operate at frequencies
of tens of kHz, well above seeing or scintillation fluctuations produced
by turbulence in the Earth’s atmosphere and they do not involve any rotating
parts and so do not produce any periodic motion of the image on the detector,
nor any periodic light modulation produced by dust on the modulator. The
PEMs in PlanetPol are type I/FS20 made from fused silica, with a PEM90 Controller,
all manufactured by Hinds Instruments.
A 3-wedge Wollaston is used as the analyser, giving better image quality
than the more usual 2-wedge device. Following the analyser are wheels with
colour filters and neutral density filters. Two-element Fabry lenses image
the primary mirror onto single element detectors, sufficient for a stellar
polarimetre, and these also eliminate any problems with flat-fielding. The
very high modulation rates of the PEMs (20 kHz for PlanetPol) are, in any
case, incompatible with the readout rates for CCDs, although the solar ZIMPOL
Polarimetre, see http://www.noao.edu/noao/staff/keller/
uses the charge shifting and storage capabilities of CCDs to act as a synchronous
demodulator, thus largely overcoming the readout limitations.
PlanetPol uses Avalanche Photodiodes (APD), which have higher quantum
efficiency than photocathodes and less noise than the external amplifier
of a photodiode, providing the best signal to noise for the photon rates
achieved with PlanetPol. The APDs were specially designed by Hamamatsu (type
C4777-SPL-S2383-70K), operating with a gain of 100, a spectral response
covering 400–1000 nm, a frequency response of 0–70kHz (3dB) and employ a
2-stage thermoelectric cooler (TEC), giving a detector temperature of –20°C.
They have a nominal size of 1 mm with an active area of 0.70 mm. Their NEP
PlanetPol has 2 channels, the star channel, on the telescope axis, and
a sky channel, offset by 95 mm. As only one PEM is used in each channel,
the instrument has to be rotated through 45 degrees to measure the second
Stokes parameter for linear polarisation. The analyser, together with the
filters, and detector assemblies, can be rotated through 90 degrees so as
to change the phase of the modulated signal by 180 degrees, and hence eliminate
any offsets in the signal detection train.
Each of 4 signal channels uses a Stanford Research SR830DSP lock-in amplifier
to extract the linear polarisation signal modulated at 40kHz (twice the modulation
frequency of the PEM). The DC signal from the detectors is fed to a 16-bit
ADC. An ARCOM Industrial PC, running Agilent Vee Pro 6 is used to control
all the mechanical functions of the instrument, the settings of the lock-in
amplifiers and ADCs, and also acquires and displays the data. Communication
with the ARCOM computer is via an Adderlink KVM Extender, using a dedicated
ethernet line to a remote monitor and keyboard in the observatory’s control
room. Data reduction is carried out using a laptop running IDL, which shares
files with the ARCOM over the LAN.
In order to measure fractional polarisations of 10–6 or less,
it is essential to determine the telescope polarisation (TP) so that the
absolute error in TP is much lower than 10–6. This
was achieved by observing nearby stars (typically within ~20 pc) with the
telescope de-rotator on, causing the TP to rotate while any other contributions
to the polarisation are fixed. Measuring the polarisation of these nearby
stars as a function of parallactic angle should produce, in the absence of
any intrinsic stellar polarisation, interstellar polarisation, and instrument
polarisation, a sinusoidal curve in U and in Q with an amplitude
equal to the telescope polarisation, and phase shifted by 45°. Figure 3 shows the measured U and Q
polarisations as a function of parallactic angle. In practice, even for very
nearby stars, there will be some interstellar polarisation and a curve was
fitted to the data set taking this into account. The best fit curves give
a fractional polarisation for the TP of (16.45±0.20)×10–6.
This very low TP makes the WHT an ideal telescope for making very high sensitivity
Figure 3. U and Q polarisations for nearby stars
at different parallactic angles. U and Q are not shown in the equatorial coordinate
system. Fractional polarisations are in units of 10–6. [ JPEG | TIFF ].
In the two observing runs to date observations were made of τ Boo and
ν And, although poor weather has meant that we have only limited data. Nonetheless,
the very good performance of PlanetPol gives us confidence that we can determine
the polarisation signature of the hot-Jupiter EXPs.
There are also several other observational programmes that will be possible
with PlanetPol, with its very high sensitivity. Of course, very high sensitivities
require large numbers of photons with a fractional polarisation of 1×10–6,
requiring 2×1012 detected photons.