The Watcher project began in 2002 when funding was obtained from the Irish Research Council for Science, Engineering and Technology (IRCSET). The Principal Investigator is Lorraine Hanlon. We proposed building a robotic telescope system which, through innovative use of readily available commercial products, would be able to respond to gamma-ray burst (GRB) triggers within minutes (or better) and detect optical flashes and afterglows as faint as 19th magnitude in brightness. The system was designed to respond primarily to very precise (arcminute) GRB locations obtained by satellites such as HETE-II, INTEGRAL (in orbit) and SWIFT (in orbit). The design allows a smaller field of view and hence a greater telescope sensitivity than earlier generation robotic experiments. The coordinates are sent directly to the telescope via the Gamma-ray Coordinates Network on an Internet socket and the telescope automatically slews to the GRB location and begins observing.
The telescope is located at the Boyden Observatory astronomical site, near Bloemfontein in South Africa. The site is excellent, with the required Internet infrastructure for coordinates and data transfer. The southern hemisphere at these longitudes is not well covered by rapid response telescopes and hence a unique database of GRB flash and afterglow data is being obtained. Watcher is therefore making a very significant and timely contribution to the knowledge of the environments of GRBs and the physics of their afterglows.
The name Watcher came mainly from the name of the original satellites which discovered GRBs, which were called the Vela series .’Velar’ is the infinitive of the Spanish verb ‘To watch’, hence ‘Watcher’.
Collaboration with Boyden Observatory
The first visit to Boyden was made in December 2002, where we met the astronomers responsible for the site, Dr. Pieter Meintjes, Dr. Matie Hofmann and Hannes Calitz.
We discovered a long-standing Irish connection with the site, as it had hosted the Armagh-Dunsink-Harvard telescope from 1950 to 1978. A detailed account of the ADH telescope and its importance for Irish astronomy is given in this article by Dr. John Butler.
The research and outreach programmes at Boyden observatory were being revitalised at the time of our visit, with a new auditorium being planned (now built) and a gravitational microlensing observing programme underway with the 60” telescope. There was shared enthusiasm for finding a home for Watcher on the site and in 2003 University College Dublin and the University of the Free State in Bloemfontein signed a Memorandum of Understanding for the project.
We received a warm welcome from the ‘Friends of Boyden’ and were invited to join their trip to view the 2002 solar eclipse from close to the Zimbabwean border. The weather didn’t quite co-operate, but glimpses through the cloud made the 2,000km round-trip worthwhile.
Lord of the Rings fans might be interested to know that Bloemfontein is the birthplace of J.R.R. Tolkein.
Refurbishment of the Watcher building
The building chosen to house Watcher has a roll-back roof and laboratory space separate from the telescope building. Significant refurbishment of the building was required to make it ‘fit for purpose’. The old equipment had to be removed from the building and new roof trusses designed in order to allow Watcher to be placed on a tall pier to maximise its sky coverage. The winch system had to be motorised and placed under software control to allow the roof to open and close automatically. The roof electronics system was designed by UCD Electronics technicians Tom McKenna and Victor Litera, who also installed the system. The roof motor system was built and installed by UCD Physics’s principal technician Dave Cooney. In the summer of 2004, major construction work was carried out on the building.
Meanwhile back in Dublin
The complex software for the project was implemented by PhD student John French. We chose to use the Linux-based robotic observatory manager RTS-2 which had been designed by the Czech Ondrejov group and was being used to operate the Czech BART system and the Spanish BOOTES telescopes. Having collaborated with Dr. Rene Hudec and Dr. Alberto Castro-Tirado, we were confident that we could overcome any technical problems with their help. We chose a Linux-based system for the project due to its stability and open-source ethos. New device drivers had to be written for several devices by John which were not originally incorporated into RTS-2 and we got (and still get) a lot of support from Peter Kubanek, who is the main architect of the package. More technical details can be found in here.
The main component of the telescope is the optical tube assembly (OTA) which houses the mirror. Budgetary constraints restricted the size of the mirror to a 16” (40 cm) diameter primary. We chose a classical Cassegrain system from US company Optical Guidance Systems who provided us with an excellent product. There was a long delay on the procurement of the OTA but it arrived in UCD for preliminary testing in March 2004.
A temporary home for Watcher was made on the roof of the School of Physics to allow some preliminary checks to be made on the system. The poor Irish weather made it clear why Ireland is such a bad place to try to do observational astronomy and we were happy when sufficient testing had been carried out to allow the system to be shipped to South Africa.
There and back again (twice)
Veteran PhD student John French (joined November 2002) and rookie Gary Melady (joined September 2004) made their first trip to Boyden in September 2005. They were joined by UCD electronics technician Victor Litera who installed the roof control sub-system.
John and Gary had to overcome very major problems at the site on that first trip and then again on their second expedition later in the Autumn of 2005. There were many issues, such as Internet access and security which were solved on-site. However, it became apparent that integrating the ‘Millennium’ telescope mount into the robotic observatory manager was causing significant technical problems related to telescope pointing which appeared unsolvable.
We decided to bite the bullet and opt for a different mount, the Paramount ME robotic mount and we were happy to find that it lived up to its name. It was integrated almost out of the box with RTS-2 and on the final field trip in March 2006, despite lousy weather (a month’s rainfall in a few days), the jubilant team announced (in an email from John whose subject line read, with typical understatement, ‘We’re in business’) that on March 29th, 2006, Watcher responded to its first gamma-ray coordinates network alert, just 5 minutes after they got the system working. Watcher re-pointed within 18 seconds and started taking data. It turned out to have been a false alert but it showed the system was working.
We had a fantastic run in the summer of 2006, with 190 days of remote operation (mostly unattended) and 149 observing nights between April and September 2006. When the GRB is immediately visible, the response times are typically tens of seconds, exceeding our original design expectation.