Both differential image motion observations and direct measurements of
turbulence show that the seeing during the winter at the South Pole is
approximately 1.7 arcseconds at visual wavelengths at a height of 12 meters
above the snow surface. Assuming Kolmogarov turbulence, this implies seeing
of approximately 1.2-1.1 arcseconds at thermal infrared wavelengths between
2.4 and 5 µm. Almost all of the turbulence causing the seeing is
the lowest 200 meters of the atmosphere, where wind velocities are
typically less than 7 meters/sec when the sky is clear. In these
conditions, simple tip-tilt adaptive optics should allow 2-meter telescopes
to achieve diffraction-limited imagery at 2.4-5 µm over large fields of
view (Wild et al. 1998).
Seeing: Optical and Infrared
Differential Image Motion Measurements
Two telescopes made HDIMM (Hartmann Differential Image Motion Measurement)
observations at a wavelength of 500 nm during the winter of 1995-96. The
measurement technique is similar to the differential image motion method
described by Sarazin and Roddier (1990),
but employs multiple apertures to
increase the amount of data which can be accumulated during a given
measurement period (Bally et al. 1996).
The telescopes were the 60-cm
SPIREX telescope (using a 48-aperture HDIMM mask) and a 28-cm Celestron
telescope (using a 12-aperture HDIMM mask). The 28-cm telescope was
attached to the side of the 60-cm telescope. They were mounted on a tower
approximately 12 m above the snow surface. A set of observations consisted
of measurements of five stars, Alpha Centauri, Beta Centauri, Sirius, Alpha
Eridani, and Alpha Carina. Each set took about 75 minutes. There were 28
sets of observations during the course of the winter, comprising a total of
274 individual stellar measurements (Loewenstein
et al. 1998). The
differential-motion measurements yield a robust measure of the seeing. The
two telescopes give similar values, with medians of 1.71 arcseconds for the
60-cm telescope and 1.64 arcseconds for the 28-cm telescope. Plotting the
seeing data against wind speed and direction revealed no strong
correlations. The HDIMM data can also be analyzed to obtain the common-mode
motion of the images. The common-mode motion is comprised of a component
due to common-mode seeing and a component due to vibrations of the
structure supporting the telescope. These results set upper limits on the
separate contributions of the two components.
The mean rms common-mode motion for the 60-cm telescope was 0.84
arcseconds. Power spectra of the motions show that most of the power is
concentrated at frequencies below 5 Hz. The common-mode motion correlates
weakly with wind speed, with higher values observed above 14.5 mph (the
mean winter wind speed). There was no correlation with date of
observation over the May-September duration of the experiment.
The figure below shows the accumulated probability distribution based
on the individual measurements for each telescope. 50% of the time
the seeing was better than 1.64 arcseconds in the C-11 and 1.71
arcseconds in SPIREX. The average seeing values of each telescope
are 1.53 +/- 0.053 arcsec and 1.42 +/- 0.058 arcsec, respectively.
The median seeing for each telescope is 1.64 arcseconds and 1.71
Direct measurements with microthermal sensors on a 27-m tower
al. 1996) and during 15 balloon flights (Marks et al. 1997) yielded results
which are consistent with the HDIMM observations. In addition, the data
show where the turbulence responsible for the seeing is located. The mean
value of the seeing calculated by integrating the balloon microthermal
measurements from the surface to the maximum balloon altitude was 1.86
arcseconds. The data show that the turbulence causing the seeing is
strongly concentrated in the lowest few hundred meters of the atomosphere,
as expected from the altitude profiles of temperature and wind shear. The
mean seeing above approximately 200 meters is 0.37 arcseconds. The balloon
microthermal measurements were made and analyzed using techniques identical
to those employed by Jean Vernin and his collaborators at a number of other
astronomical sites, including Parenal in Chile. Above 200 meters, the Polar
seeing is better than the seeing above Parenal. Microthermal observations
from a 27-m tower (Marks et al. 1996)
indicate that the contribution of
the lowest 27 meters of the boundary layer to the total seeing is
approximately 0.6 arcseconds.
Summary of integrated seeing and boundary layer data, from 16 balloon
launches between 20 June and 18 August 1995. The "free atmosphere"
refers to the entire atmosphere excluding the boundary layer.
Values are quoted for a wavelength of 0.5 µm.
For more information, see Marks et al. 1999.
Acoustic backscatter (also called echosonde or SODAR) measurements also
give information about the vertical distribution of turbulence.
Measurements at the South Pole by Neff in 1993-94 (personal communication)
confirm the result that most of the turbulence lies below an altitude of
200 m. During the summer of 1998-99, the University of New South Wales
group installed an acoustic sounder in the AASTO (Automated Astrophysical
Site Testing Observatory) at the South Pole. This instrument measures the
altitude profile of turbulence, wind speed, and wind direction to a height
of 300-600 m. It will continuously monitor the structure of the
boundary-layer turbulence at the South Pole during the coming winter. After
characterizing the seeing and sky brightness at the South Pole, the AASTO
will be deployed to make similar measurements at other sites on the
Seeing (arcsec): total
Seeing (arcsec): free atmosphere
r0 (cm): total
r0 (cm): free atmosphere
Boundary layer height (m)