Center for Mapping

Short Baseline Test Results

 

During the three field sessions presented here seven pairs of receivers were tested, some of them multiple times. In general, all receivers were tracking the same satellites, however, some anomalies were observed on various receivers. For example, during the first test (August 27) brief losses of lock were present on L2 frequency for Trimble 4000SSE, while Leica 399, although maintained uninterrupted lock, started tracking PRN 15 with about 2-minute delay. Topcon/JPS Legacy and Leica 9500 maintained uninterrupted lock; data were entirely lost on Trimble 4000SSI (as a result of a torrential rain at the end of the session). During the September 3^rd test brief losses of lock were present in Leica 399 data for PRN 26 and 29, and Trimble 4000SSE experienced numerous brief losses of L2 lock on different PRNs. Topcon Legacy, Trimble 400SSI and Ashtech Z-12 maintained uninterrupted lock to all satellites in view.  During the October 1^st test both pairs Trimble 4700 and Legacy maintained uninterrupted lock.

 

Figure 1 illustrates the L1 single difference residual plots for a high satellite for all pairs of receivers tested. The residual scatter varies among the receiver types, and is higher than for the zero baselines, as expected. There are clearly some systematic effects present in the residual plots. Those are most likely due to multipath. It should be mentioned here that variations in the antenna phase center might be a reason for errors in the most precise geodetic applications. However, it is rather unlikely that this effect is present in the residual plots shown here, as we were dealing with very short baselines, and all antennas were North-oriented, virtually resulting in error cancellation. Moreover, the data analysis indicates that the radio interference and ionospheric effects have not significantly contaminated the data, even though we are at the peak of the 11-year sunspot cycle. The gap in the residual plot for Leica 399 receiver is caused by the fact that the memory card had to be replaced during the survey.

 

 

 

 

Figure 1. L1 phase residuals for PRN 23 (high satellite) on 10-m baseline; first row: Trimble 4000SSE (left), Topcon/JPS Legacy (right), (GPS week 1078); second row: Trimble 4000SSI (left) (GPS week 1078), Leica 9500 (right) (PRN 21 (high), GPS week 1077); third row: PRN 23, Leica 399 (left), and Ashtech Z-12 (right) (GPS week 1078); fourth row: PRN 9 (high), Trimble 4700 (left), Topcon/JPS Legacy (right) (GPS week 1082).

 

 

 

The short baseline summary statistics, covering all tests, is presented in Table 1 in terms of residual scatter, and the standard deviations, based on L1 single difference least squares residuals for low and high satellites. The level of standard deviations indicates the measurement precision (on single difference level), which is more likely to show the real receiver performance, as opposed to the zero baseline, which might provide slightly too optimistic estimates.

 

 

 

 

 

Table 1.  Short baseline test, low satellite (left) and high satellite (right) – survey statistics. Results for the receivers tested multiple times are included.