Radial velocity stability

- PMS updated the FIES fiber-positions monitoring page to give more useful information: It displays current positions (basically the content of the old .def files) plus to option to list the last 30 split positions for a given fiber.

- dying ThAr diagnostic: added self-updating plot to the trouble shooting section, that is recreated every time fies-calibs is run. The plot helps to define what is normal when the system starts to report warnings about 'few lines found' in the ThAr analysis.

ToDo:

- write a small document that details the physical constraints for the design of a new fiber bundle

As an indication of the long-term radial-velocity stability, observations of a standard star with the medium-resolution fiber as part of observing programs spanning various semesters show a standard deviation of 11.4 m/s. The measurements span 15 months (December 2012 - March 2014), which includes 2 times that the dewar of FIES was pumped.

Lars Buchhave (NBI & StarPlan, Copenhagen, Denmark) has expressed his wish to upgrade the fiber bundle in order to decrease the RV zero-point scatter for the high-resolution fiber from the current level of 8 m/s to an aimed-for level of 3 m/s. This involves the use of octagonal fibers and possibly also a double scrambler, the latter at an expected expense of about 20% of the light.

A questionnaire was sent around to the community to poll for other requested upgrades and to see if the expected loss of 20% of the light was found problematic. The general response was that what we currently offer is what the community wants. In order to meet Lars's and the communities wish we can retain the low-resolution and two medium-resolution fibers, and upgrade the high-resolution fiber for superior RV stability. The medium-resolution fibers can be made octagonal as well (without scrambler, so no additional loss of light).

There has been some discussion about the precise way how such changes might be implement. One issue is the design of the new fiber bundle, but also how to achieve optimal stability. The HARPS-N design incorporates a double scrambler, which may be difficult to install in our case. It has also been noted that directly coupling fibers with different shapes (octagonal, hexagonal, heptagonal or circular) in a face-to-face configuration, gives good scrambling results as well.

We had a meeting with Lars at the NOT office, and agreed that Lars will try to secure funds for a new fiber bundle, and we will provide a document that describes all the constraints that may impact on the design of a new bundle (which is needed to estimate the required cost).