We describe here a method to correct interferometer phase variations due to fluctuations in the atmospheric water vapour. The fluctuations are estimated by fitting an atmospheric water line model to antenna temperature values measured at frequencies near the water vapour line at 183 GHz. The fitting is done with the atmospheric water vapour density as a variable, iterating until agreement between observation and prediction is achieved. From the computed value of water vapour density, an integrated value for the precipitable water vapour (PWV) contained in the atmospheric column is derived.
The variations on the amount of PWV are directly related to electrical path length variations of an electromagnetic wave travelling through the atmosphere. This quantity, at millimetre and sub-millimetre frequencies, can be linearly related to a phase variation in an astronomical signal arriving to the elements of an interferometer.
This method to estimate the PWV has been used before to analyse the data from a 183 GHz radiometer operating at Llano de Chajnantor, the selected site for the Atacama Large Millimeter Array (ALMA), creating a time series of the PWV in the area. The derived values have been previously correlated with observations of atmospheric opacity at 225 GHz and radiosonde measurements, with good results .
Here we apply this radiometric method of phase variation estimation to two real cases. One is the data from interferometric observations at Mauna Kea between the James Clark Maxwell Telescope (JCMT) and the Caltech Submillimeter Observatory (CSO), operating at 356 GHz. The second case is at Chajnantor, where we have a 300-m baseline interferometer operating at 11.2 GHz. In both cases we compare the phase estimated by radiometry and the interferometer phase. Figures for correlation and achieved phase correction are given. Some error sources are discussed.
View a PDF version of ALMA Memo #332.
Download a postscriptversion of ALMA Memo #331.
Last modified: November 03, firstname.lastname@example.org