Focus On: Deployment of the IRMA Infrared Water Vapor Radiometers
  Matthias Schoeck

The higher you go, the drier it gets. That is the rule of thumb concerning the average amount of precipitable water vapor (PWV) above an astronomical site. It works pretty well as a guideline, but there are many aspects of PWV that are of interest for TMT in a more quantitative way. How often is the PWV content of the atmosphere above the site lower than a certain value? What are the conditions one gets during the best 10% of the time? Are there diurnal or seasonal patterns that are specific to a site? And so on. The answers to these questions will not only affect site selection, but might also determine certain TMT operations strategies. For example, if the PWV is always high during a certain time of the year, you would not schedule mid infrared observations for that time. Operation of a PWV monitor on the TMT candidate sites was therefore always considered highly desirable.

In mid 2005, TMT contracted with the Astronomical Instrumentation Group of the University of Lethbridge in Alberta, Canada, to build three updated versions of its Infrared Radiometer for Millimetre Astronomy, IRMA. These radiometers measure the infrared flux of the sky in a band around 20 microns. In this band, water lines are the only significant emitters and the flux depends directly on the amount of PWV along the line of sight. An atmospheric model is used to determine the PWV value from the measured flux. Several units of two previous versions of IRMA existed already and had been tested at several sites, including Mauna Kea. The units are small and compact and therefore well suited for operation on a remote site.

The three IRMAs were first deployed side-by-side on Paranal for two weeks in January 2006, and then moved to Armazones. The comparison of the results showed that the IRMAs do, indeed, all measure the same thing, but that there were systematic offsets between them that needed to be addressed by a refinement of the calibration procedure. In addition, we discovered that some modifications would be needed to make the units more robust for operation at the TMT candidate sites. After approximately 3 months of data had been taken on Armazones, all units were sent back to Lethbridge in the middle of 2006 for additional work. Matthias Schoeck went to Lethbridge for 5 weeks in October and November to assist with the modifications and to work on the calibration. By the end of his stay, and thanks to a great, concentrated effort by everybody in the IRMA Group, all three units had been upgraded, all were working reliably, and a new calibration concept had been worked out. The units were shipped back to Chile on December 15, 2006, and were again tested side-by-side on Armazones for two weeks in January, 2007. There, the units produced PWV values that agreed to better than 10% at high PWV levels and to approximately 0.25mm at low levels. One unit remains on Armazones, one was deployed to the 13N site on Mauna Kea at the end January, and the third is being transferred to Tolonchar as this article is written.

More details on the IRMA radiometers can be found here: http://research.uleth.ca/irma