M.A. Holdaway [1], S.M. Foster [1], Darrel Emerson [1], Jingquan Cheng [1], Fred Schwab [2]
August 9, 1996
Keywords: wind speed, wind fluctuations, site testing, weather, chile, Chajnantor, Mauna Kea, antenna pointing, pointing calibration, slant axis antenna, conventional antenna
We analyze wind statistics on the Mauna Kea/VLBA and Chajnantor, Chile sites for the one year period from May 1995 through April 1996 for the purposes of understanding the effects of the winds on pointing errors. Both the Chilean and the Mauna Kea sites display seasonal changes in the wind speed, with winds blowing strongest during the local winter. There is a very weak diurnal variation of wind speed on Mauna Kea, and a very strong diurnal variation of wind speed on the Chilean site with day time winds typically being 2-3 times higher than the night time winds. Since the opacity and phase stability on the Chilean site also settle down to their best values at night, high frequency observations and many of the most demanding mosaics will be performed at night. Median night time winds on the Chilean site are a benign 3.5~m/s, lower than the Mauna Kea night time median of 4.2~m/s.
From model distributions of observed AZ and EL and the pointing error calculations of Cheng (1996, in preparation), we calculate 9~m/s winds typically produce 3.5~arcsecond pointing errors in the conventional design antenna. Although Chilean winds are often high, they are often not highly variable over 10 minutes. The wind power spectrum in Chile is flatter than the Davenport spectrum (1962). If pointing is calibrated every 10 minutes, wind induced pointing errors are reduced to 0.25 - 0.5 of the static wind pointing error, bringing the conventional antenna close to meeting the pointing specification of 1~arcsecond or better in the best half of the wind conditions.
In contrast, the slant-axis mount with an on-axis paraboloid (Cheng, 1993) is able to point to 1.3~arcseconds in 9~m/s winds (Cheng, 1996). The MMA's 1~arcsecond pointing spec is extremely important for mosaicing at millimeter wavelengths and all observations in the submillimeter. We need to scrutinize antenna pointing and calibration in the future to ensure that the MMA antennas will be able to meet this pointing specification. If it is not believed that the conventional antenna can get close to the pointing spec, the slant- axis antenna must be reconsidered.
[1] NRAO/TUC
[2] NARO/CV
View an HTML version of MMA Memo #159.
Download a 0.2MB gzipped postscript version of MMA Memo #159 from
Last modified: 09 December, 1999
kweather@nrao.edu