National Radio Astronomy Observatory
Site Studies
Atacama Large Millimeter Array/US

Chajnantor GPS Receiver

In 1998 June, a commercial Global Positioning System navigation receiver (Garmin GPS 35PC) was installed on the roof of the NRAO instrument container at Chajnantor to provide an absolute time reference for the instrument computers. The receiver antenna is mounted at the top of a solar panel bracket, about 3.26 m above the ground level.

Starting at 1998 November 26 UT 19:30, the GPS receiver position has been recorded every 15 minutes. Intentional degradation of GPS signals (Selective Availability) was discontinued at 2000 May 2 UT 4:00. Measurements recorded through 1999 April 1 UT 17:30 are analyzed in Memo 261 and measurements through 2000 June 7 UT 16:56 are analysed in Memo 312.

The figures below update the corresponding figures from Memo 312 with data recorded 2000 May 2 UT 4:00 through 2001 February 6 UT 14:13. See Memo 312 for discussion.

Table 1. Chajnantor Position
WGS 84 † SAm 56
Latitude [South] 23° 1¢ 22.42² ± 50 mas 23° 1¢ 9.42²
Longitude [West] 67° 45¢ 17.74² ± 30 mas 67° 45¢ 11.44²
UTM Zone 19: Northing 7 453 408 m 7 453 773 m
Easting 627 583 m 627 772 m
Altitude receiver 5062 m ± 2.5 m
ground 5058.7 m
† Uncertainties estimated from Figure 6.

Figure 3. Measured positions of the NRAO equipment container on Chajnantor, Chile, after 2000 May 2 UT 4:00, when SA was discontinued. In the right panels, the histograms show the actual distribution of the measured positions and the curves are normal (Gaussian) distributions for the parameters derived from the data. Note scale difference between latitude-longitude and altitude.

Figure 4. Positions measured after discontinuation of SA and linear cross correlation coefficients. Note scale difference between latitude-longitude and altitude; 0.1² corresponds to 3.1 m.

Figure 5. Diurnal variation of positions measured after discontinuation of SA. Note scale difference between latitude-longitude and altitude; 0.1² corresponds to 3.1 m.

Figure 6. Variance in the mean positions for successively longer sets of measurements after SA was discontinued. The error bars illustrate the sampling uncertainty for a normal distribution, ss/s = [2/(N-1)]1/4 (Bailey 1971). The dashed lines show the decrease with N1/2 expected for sequentially uncorrelated data.

Figures in gif and compressed postscript (.ps.gz) format.

Last modified 2001 March 7
Simon Radford