IGS Station Height Tests
The IGS permanent GNSS stations are the backbone of the IGS as they provide the overwhelming bulk of the satellite observations used for the IGS products; orbits, clocks, station positions, TZD, etc. The station GNSS receivers record the measurements from the satellite constellations (GPS and GLONASS) as correctly as possible. This means following the IGS guidelines so as not to introduce unknown effects into the recorded observations. If measurements are not properly recorded by the receiver+antenna installation at each station the IGS Analysis Centers and other interested researchers will not be able to properly correct or compensate the effects and the data will be rendered useless in the estimation processes, or even worse the data will negatively affect the IGS products.
No matter how much care is taken into the installation of a permanent GNSS station it is possible that the data are biased due to local effects, either horizon blocks, reflective surfaces, uncontrolled radio-emitting sources, etc can affect the record measurements at the GNSS receiver making them less useful for the IGS.
One of the biggest concerns is the possible dependence of a station’s position calculation to the antenna’s near-field environment. If there are elements close to the antenna that disturb the signal reception then biased measurement can be recorded by the receiver. Since the antenna is a relatively small element compared to its surroundings it is possible to try to see the effect of the surroundings on the measurements by limiting how many of the measurements we actually use in any given calculation. By limiting the observations at low elevations we remove the measurements most likely to be affected by the surroundings.
Taking random days over all seasons during 2009 more than 325 are processed with fixed orbits . The runs go from 5 to 40 deg elevation cut-off angles. The height differences are calculated with respect to the 5 deg solution. It would be expected that the position solutions at a station not be affected too much by excluding low elevation data. Indeed this is correct for most of the stations, but some stations show very large differences when limiting the low elevation data the height estimate changes by decimeters sometimes! Here is the example of one of our (ESA/ESOC) stations MAS1, in Maspalomas, Gran Canaria. It can be seen by the red line that as the elevation cutoff is increased the station height estimate changes by up to 2 cm, and the error band increases of course as less data is used in the estimation.
This kind of processing is also helpful when the observation residuals are analyzed in azimuth and elevation. Taking all the residuals over each of the processed days in 2009 a “station fingerprint” (Huisman, et a., 2009) can be interpolated from the un-differenced residuals. In the case of MAS1 the station fingerprint shows the direction in which blocks and reflective surfaces are located, causing either a hole in the observations or significant negative/positive residuals.
These kind of investigations are important when evaluating the performance of a GNSS permanent station, and is the kind of in-depth analysis I have been doing as part of the IGS Infrastructure Committee.
Happy positioning
Ignacio Romero (Nacho of the IGS!)
