Short baseline tests of SCIGN radomes

reported by Ken Hudnut, Jeff Behr, Shannon Van Wyk and Aris Aspiotes
(last update December 30 at 3:05 p.m.)

Summary of tests

This page describes experiments conducted by the SCIGN-USGS group to see whether the tall and short models of the SCIGN radome cause significant shifts in the apparent phase center of standard Dorne-Margolin Choke Ring (DMCR) antennas. We conducted these tests using the first-article parts of these SCIGN radomes, mounted on Ashtech DMCR antennas. The tall radome model was mounted on top of a SCIGN antenna adaptor. All tests were done on the roof of the North Mudd building on the Caltech campus.

In our tests, SCIGN radome part numbers 0010 (short radome), 0020 (tall radome) and 0030 (base plate for use with tall radome) - all with serial numbers 00002, 00002, and 00002 were used. The main anticipated difference between these units and the 'production' SCIGN radomes is that the radius of the 'ring' will be increased so as to improve how both the tall and short model radomes fit together. This and any other less significant changes are not at all likely to change the main results of GPS signal propagation or phase pattern testing on the first-article parts.

The 'rock-pin' mounts around the North Mudd roof parapet were used in these tests, in particular NMD3 (for the tall radome and base plate) and NMUD (for the short radome). During the time of the tests, an additional station called NMD1 was also operated with the same type of Ashtech equipment installed.

Also, a short baseline is available to existing SCIGN station CIT1 , only a few meters away, but with a different antenna mount and AOA turborogue receiver and AOA DMCR antenna. Data from that receiver are in the SCIGN archive and available for this time from the SCIGN archive at SIO. As part of our tests on North Mudd at this time, the turborogue receiver on CIT1 was exchanged for an Ashtech Z-12 receiver. This change was made at the time of the GMT day boundary between days 351 and 352 (SCIGN mail message #0487 & 0488). At that time, no changes were made in the CIT1 antenna, antenna cable, or antenna mount at the time of the receiver change. This change was made after most of the radome testing had been well advanced, so it is hoped that this change should not interfere with the radome test results.

Photos of the experiment setup are available here . On this index of small photos, click on any image to see an enlarged (and higher resolution) view of that photograph. The various photos show the SCIGN adaptor and how it is mounted on NMD3 (using a 5/8-11 to pipe thread adaptor) and then the assembly of the base plate and tall radome onto the SCIGN adaptor. The short radome, set up on the NMUD station, had a pattern of dirt (from water droplets, evidently) on it when it was removed, and a couple of the photographs show that dirt accumulation. Also, the short dome had some spiders living in it after just the week or so of its deployment (we think they enter through the drain holes in the choke ring).

Preliminary results indicate that each of the two radomes tested introduced marginally significant shifts in the vertical phase center but no significant shifts in the horizontal phase center. It appears that the vertical shift introduced by the short radome may be slightly larger than that introduced by the tall radome (it is not clear why that would happen - the radome portions of each are manufactured identically - the only differences are physical and have to do with attachment to the antenna or adaptor).

Results from other Investigations:

In e-mail received on 12/21/98 from Yehuda Bock at Scripps, SOPAC has analyzed data from a similar yet independent short baseline test of the tall radome and base plate on the PIN1-PIN2 baseline and he finds similar results. The Scripps group is also testing the short model radome on their SIO2-SIO4 baseline currently.

In e-mail received on 12/29/98 from Frank Webb at JPL, the JPL group has also analyzed data from the PIN1-PIN2 baseline and conclude that the radomes have no significant effect on carrier phase. The JPL group has posted their results onto the web here.

In a second e-mail received on 12/30/98 from Frank Webb at JPL, he mentions some additional new results that have been updated on the JPL group's web page. The radome appears to increase the usual dependence of change in vertical phase center with higher (>20 or 25 degree) elevation angle cutoff.

Data and Observation Schedule:

Data from the relevant stations, CIT1, NMUD, NMD1 and NMD3 are available from the archive at Scripps. The data from CIT1 and NMD3 are downloaded via modem whereas data from NMUD and NMD1 are down- loaded manually (so there is a delay in posting data from those two stations). Files can be retrieved as discussed previously, or from the Scripps ftp site.

To receive an up-to-date tabulation of the data availability from these experiments, use this web tool to access SCIGN data directly from the Scripps archive. By clicking on a link in the tables generated by this web tool, you can download the data files. For example, to list data file links for the stations of interest for the time period of day 325 through 356, here is the table. Now, just click on one of the X's (or other symbols) to retrieve the RINEX data files using your web browser.

Here is a summary of when changes were introduced at each site:

Other than these changes, no other modifications have been done to the equipment on the North Mudd test range during the time interval day 329 - 357(today). Data from station NMD3 for most of day 349, all of 350 and nearly all of 351 were over-written and are not recoverable, unfortunately.

Processing Results:

At this point, Jeff Behr has made first-cut runs of data from these experiments, processing daily solutions (0000-2400 GMT) for each day. Sub-daily solutions and other processing strategies will be tried and reported on soon.

Daily solution results

The results from 5 different solution methods are available in separate text files:

The first group of solutions were done in L1 only mode:

  • with tight a priori coordinate constraints
  • with loose a priori coordinate constraints; 15 degree cutoff
  • with loose a priori coordinate constraints; 30 degree cutoff

    The second group of solutions were done in L2 only mode:

  • 15 degree cutoff
  • 30 degree cutoff

    These reports are preliminary and have not been reviewed for conformity with U. S. Geological Survey editorial standards. Any use of trade, product, or firm names is for descriptive use only and does not imply endorsement by the U. S. Government.

    Return to USGS Pasadena - GPS Index