The RCADC and Atmospheric Refraction

For wide-field use, especially with Hydra, a new corrector has been installed at the R/C focus of the Blanco Telescope. It is referred to as the "RCADC" (Ritchey-Chrètien Atmospheric Dispersion Compensator) corrector. It is located in the telescope chimney. Click here to see just where it is.

Hydra MUST be used with this corrector. Any optical R/C instrument can use the RCADC if desired, but there is not much justification. The R/C and echelle spectrographs do not need its wide field though they sometimes might benefit from the ADC function. The RCADC can only be installed by Observer Support personnel via a motorized system operated from the Cass. cage.

The RCADC has six elements in four groups. This Optical Diagram shows its configuration. It contains two meniscus "corrector" elements of fused silica at the front and back surfaces of the assembly. They provide images with D80 less than .3 arcsec over the entire 42 arcminute Hydra field. The corrector also makes the image "telecentric", which means that the pupil is located at the center of curvature of the field so that the optical axis of the images is perpendicular to the focal surface over the entire field. This minimizes light lost due to focal ratio degradation (FRD) in the fibers.

Between the two corrector elements, there are two cemented doublet prisms of silica and a light flint glass (LLF6). All surfaces of these prisms are plane, inclined appropriately so that the light passes through with zero deviation at an intermediate wavelength (4200A). Each prism provides a small amount of dispersion and rotates under control of the TCS though an angle of 360 degrees. When the two elements are oriented 180 degrees apart, their dispersions cancel so that the prisms have essentially no effect on the images. Orienting them at different angles can provide an artifical dispersion in any direction desired, which can compensate for atmospheric dispersion up to the limiting power of the prisms which in the case of the RCADC is at Air Mass 2.4 (65 degrees zenith angle).

The RCADC is coated with sol-gel over MgF2 on all eight surfaces. Sol-gel over MgF2 has very low reflectivity over a broad wavelength range. Although it has not been directly measured, the overall transmission of the corrector is believed to be above 95% at all wavelengths from 4000-10000A. Transmission falls in the UV due to the LLF6 elements in the ADC prism. Throughput of the corrector is roughly 85% at 3500A, 60% at 3340A and 20% at 3200A.

If dispersion correction is not desired or has been disabled for some reason, the ADC elements MUST be set in the neutral position. This is easily done via the TCS. Zero and 180 degrees is the standard setting but any orientation of the prisms 180o apart is equivalent. Normally, the ADC is left on and dispersion correction is automatic.

Observers sometimes ask when ADC should be used. The safest answer is "always". If there is significant dispersion in the field, correcting for it will improve the efficiency of the observation. It will never make it worse. The only reason not to use the ADC function is to avoid any possible effect on the pointing accuracy or if the control system is malfunctioning. Optical analysis indicates that rotating the elements does not significantly alter the field model, though for lack of time this has not been explicitly verified.

The expected effect of atmospheric refraction on the observing efficiency can be estimated from the following diagrams.

1. Differential refraction at 2km altitude
2. Flux captured by Hydra fibers as a function of seeing and centering
3. Image movement during exposure caused by refraction

The first diagram quantifies the effect of refraction while the second lists fiber efficiency using the standard (Wolff) model of the profile of images degraded by seeing. Using the two diagrams it is relatively easy to estimate the effect of refraction on system efficiency.

For example, when the seeing is 1.0 arcsec, 85% of the incident light will enter a perfectly centered Hydra fiber. If the image is decentered by 0.5 arcsec, the efficiency falls to 73%. A decenter of 1.0 arcsec decreases the efficiency to 39%. Thus, if refraction decenters a star by 0.5 arcsec, in 1" seeing, overall system efficiency will decrease by approximately 14% (.85-.73/.85). Correspondingly at this seeing the efficiency will decrease by 54% if there is a 1 arcsec centering error.)

One can study the table as a function of seeing and estimate how much effect seeing might have on overall efficiency in a particular observing situation. If (say) the 10-15% efficiency degradation produced by an .5 arcsec offset is deemed acceptable, then an overall dispersion of 1 arcsec could be tolerated. Diagram 1 then tells us that someone observing from 3500-5000A could observe to an air mass of 1.3 without using the corrector. Observations from 4000-6000A could be done to an air mass of 1.45 while observations from 6000-9500A could be made at any air mass up to 2.40.

Important! Note that these tables can be used to determine the optimum central wavelength for positioning the fibers. If the corrector is not used, centering the guide star(s) on the wrong wavelength will offset the entire field. Either a filter must be used in the FOPS guide camera or FOPS stars of appropriate color must be used. Of course if the ADC function is enabled, no filter need be installed in the guide camera and the spectral type of the FOPS stars will have no significant effect on the positioning accuracy of the system.

Yet another consequence of refraction is to cause an apparent relative movement of points in the image as the field moves across the sky. This effect is quantified in the third diagram above.

Here, the Hydra field is shown with the locations of star images at 9 points in the field at -70 degrees during ten hours as the telescope tracks from 5 hours east to 5 hours west of the meridian. As can be seen, the image appears to rotate about a point approximately on the edge of the field with an amplitude of about .5 arcsec per hour of telescope motion at the other side of the field and the images drift with respect to the overall rotaion.

This effect is relatively small though it can be significant under some circumstances. Differential motion is the reason that Hydra asks for the approximate time of the middle of the exposure before positioning the fibers. In some situations it is desirable to reposition the Hydra fibers between exposures and to select the location of the guide star(s) in the field with care.

Deciding on what, if anything to do about this effect is up to the observer. One can use the information given in the first diagram to make an estimate as to the relative size of refraction effects at different zenith angles. The information in the third diagram can be used to make an educated guess as to how the images will drift during and between exposures. Since the effect is small, this is all that is ever necessary.

29 May 2000
by T. Ingerson