4.7 Improved Measurement Accuracy Plans for FY97

The UNAVCO Boulder Facility has proposed to subcontract under the Smithsonian Institution to continue test and development work to improve GPS measurement accuracy that were formerly funded directly through the UNAVCO Facility NSF grant. The level of proposed effort in months/year is divided between Chuck Meertens, an Associate Scientist, and an Engineer and is summarized below in the order they appear in the main Smithsonian proposal:

(1) The proposed Multipath Calibration System (MCS) will play a critical part in the GPS community's ability to achieve the highest possible precision at a GPS site and greatly enhance the ability to measure, model, characterize and mitigate site effects. Chuck Meertens is proposed to participate directly in this effort at 0.5 months per year.

(2) Investigation of two problems using ray approximations have led to a better understanding of fundamental error sources. The first is the "Role of Multipath in Antenna Height Tests at Table Mountain" which showed that large (up to 2 cm) elevation cutoff dependent vertical errors were caused by ground multipath and correlation to the estimated tropospheric parameters. A second modeled error was the presence of an antenna raydome. This problem was identified by noticing phase pattern differences for the Ball chamber patterns for an antenna with and without a dome. First order ray tracing can partially model this effect and the influence on baseline solutions. The result has been practical recommendations on antenna setup heights and use of domes. As noted in the body of the main (Smithsonian) proposal, however, a higher level of modeling is needed to better understand and mitigate other mounting and antenna setup problems arising from multipath and modifications to receiver phase patterns. The UNAVCO Boulder Facility proposes to participate in this effort primarily through field acquisition of data and testing of model results from the moment calculations and related models. Proposed budgets include Chuck Meertens at 0.5 months and an Associate Scientist at 2.0 months per year.

(3) The effects of radio frequency interference (RFI) have become increasingly important due to the proliferation of microwave communications, even in the most remote locations. The Boulder Facility proposes to develop, in conjunction with Goddard and the National Institute of Standards and Technology (NIST), a strategy for testing and possible mitigation of RFI effects on GPS receivers and antennas/preamplifiers. Laboratory versus field range tests will be examined in order to determine the most effective quantitative RFI characterization approach. This test will be adopted for ongoing GPS system tests. Proposed budgets include Chuck Meertens at 0.5 months, and an Associate Scientist and Engineer at a combined 2.5 months per year.

(4) By far the largest activity of the Boulder Facility will be in antenna phase center offset measurements and phase pattern characterization and calibration. For this application, general antenna phase offset and pattern models, as well as specific antenna serial number calibrations are required. Developing and testing these models are also a high priority of the International GPS Service for Geodynamics (IGS) special working group on improving accuracy. These calibrations will also be critical in the validation of the MCS.

The UNAVCO Boulder Facility has over the years developed a receiver and antenna testing strategy using direct GPS field observations in calibrated baselines and zero-baseline experiments. A primary concern has been effects of antenna and receiver mixing and site effects such as antenna mounts, antenna setup heights, snowfall, multipath and domes. A typical test involves one to two weeks of field measurements using three to 15 receivers. A control setup is followed by the test of interest, such as addition of a raydome or mixed antenna pair. Analysis consists of multipath and receiver performance evaluated using UNAVCO's QC program as well as high precision baseline processing and examination of residuals and solutions. Each test from initiation to analysis and presentation of results is on the order of effort of conducting, processing and analyzing a small GPS field campaign.

Proposed are a nearly continuous set of field experiments at the UNAVCO roof antenna range and Table Mountain calibrated baseline monuments. Receivers and antennas from the UNAVCO pool will be used along with those provided from manufacturers for special testing. The initial experiments will be to verify antenna phase offset and patterns generated by the relative field calibrations by the National Geodetic Survey (NGS) and the IGS. This effort will be in support of the IGS special subcommittee addressing antenna mixing issues of which Chuck Meertens is a member. The UNAVCO/Ball tests in Boulder will be designed with the assistance of Tom Clark from Goddard. Mixed data will come from new field tests as well as an extensive set of data from past tests which resides in the UNAVCO data archives. (At the suggestion of the IGS, some of the effort will be to compile a set of standard files which can be processed by other investigators using other software in order to verify antenna phase models).

Inherent in the field phase estimations is the need for a well calibrated reference antenna, in this case a JPL designed chokering with Dorne Margolin element. UNAVCO antenna consistency tests have shown that there is variation in the horizontal phase center offsets of up to 3 mm (L3 ionosphere-free combination) for this type of antenna. A high priority will be placed on identifying a set of antennas which have the minimum offset using rotation tests from which can be determined the phase offset for individual antennas. These antennas will also be tested in chambers at Ball and Goddard and be available for future IGS field phase calibrations to be conducted at NGS, UNAVCO and Bern. These types of field calibrations are essential as it will not be possible to test all antenna models, let alone individual antennas by serial number in the chambers.

An additional task will be to use field rotations to calibrate the UNAVCO Community antenna pool. Three or more antennas can be used to determine the mean horizontal phase center offset for an individual antenna. We propose, as part of our ongoing field experiments, to calibrate as many choke ring antennas as possible before they get permanently installed and then work back into the Trimble antenna pool for which absolute calibrations have not yet determined. This field calibration will be interspersed with other proposed field tests. It is not practical to do many more than five antennas at a time so an ongoing activity is suggested.

Additional tests are proposed for antenna raydomes. A new phase concentric acrylic dome has been developed which reduces the vertical error to less than 2 mm (+- 1mm) and removes the apparent elevation cutoff dependence of the baseline solution. Longer term statistics are not available. A new compressed styrofoam dome of the same shape which, from preliminary tests, has even smaller effects needs further testing. There are also a number of other domes in use which need to be examined as well as looking into issues such a snow and moisture buildup which also effect accuracy. The proposed budget for antenna measurements is 0.5 month for Chuck Meertens and a combined 4.5 months for an Associate Scientist and Engineer per year.

(6) Finally, an ongoing program of testing new receivers and antennas for performance and mixing with the existing UNAVCO and IGS GPS pool will continue. GPS systems undergo major changes by the manufacturers with a cycle of about two years. The UNAVCO Community ARI purchase in 1995/1996 coincided with the release of the Trimble SSi with chokering antenna, the Ashtech Z-12 with chokering, and the Leica full-wavelength L2 SR399, all of which were purchased by the community. NSF investigators achieve a great deal of leverage on their projects by taking advantage of existing GPS systems, say from local, state and national surveyors and existing continuous GPS infrastructure. Competition also reduces system prices. The penalty to be paid is reduced standardization and the measurement errors related to mixing equipment. Purchasing pressure from the UNAVCO Community and previous receiver/antenna performance test reports have been successful in getting at least some of the manufacturers to offer a more standard antenna with superior performance. Unfortunately, there are other choke ring antennas being supplied by other important manufacturers. Novatel, for example, has a new dual-frequency receiver which will be used with the North American FAA continuous GPS network for navigation and the data will also be distributed to the geodetic community. Receiver tests at the mm-level will require a better understanding of tracking loops and measurements and careful testing of mixing equipment. Work in this area will be coordinated with Tom Clark (GSFC). Tests will include field and laboratory observations, data processing, and data analysis using QC and various GPS processing software. Receiver tests are proposed at a budget of 0.5 month for Chuck Meertens and a combined 3.0 months per year for an Associate Scientist and Engineer. Test results will be posted on-line and presented at meetings such as AGU and various IGS meetings.

[Next] [Previous] [Up] [Contents]

Last modified Tuesday, 08-Nov-2005 02:34:55 UTC

Home | About Us | Contact Us | Support | Search | Facility | PBO | Education & Outreach

Comments: webmasterunavco.org
© 2009 UNAVCO, Inc.