The image on the left displays height scatter after correcting for ionosphere derived from L1 and L2 pseudoranges. The image in the center displays height scatter after applying an additional correction for seasonal troposphere delay. The image on the right displays height scatter when IGS rapid orbits and clocks are substituted for the broadcast orbit and clock data.
 

Date	Dual Freq. Measured Ionosphere correction 95%	Dual Freq. Seasonal Troposphere correction 95%	Dual Freq. Rapid Orbit & Clock 95%
May 03	19.88 m	7.14 m	5.49 m
May 04	21.02 m	6.92 m	5.48 m
May 05	27.91 m *	16.98 m *	18.21 m *
May 06	20.73 m	7.44 m	5.37 m
May 07	20.59 m	7.23 m	5.65 m
May 08	21.25 m	7.00 m	5.40 m
May 09	19.89 m	7.00 m	5.45 m
May 10	20.50 m	7.29 m	5.47 m
May 11	20.78 m	7.76 m	5.17 m
May 12	21.36 m	7.74 m	5.33 m
May 13	-- data set incomplete --
May 14	19.46 m	7.48 m	5.21 m
May 15	20.36 m	7.94 m	5.58 m
May 16	20.71 m	6.83 m	5.37 m
May 17	19.50 m	8.37 m	5.23 m
May 18	21.02 m	8.14 m	5.42 m
May 19	23.35 m *	11.72 m *	11.46 m *
May 20	20.95 m	7.14 m	5.70 m
May 21	19.84 m	7.46 m	5.25 m
May 22	21.33 m	8.40 m	5.15 m
May 23	21.11 m	7.39 m	5.39 m
May 24	22.84 m	8.71 m	5.02 m
May 25	21.36 m	7.31 m	5.13 m
May 26	21.28 m	7.48 m	5.36 m
May 27	21.36 m	7.79 m	6.00 m
May 28	21.01 m	6.99 m	5.66 m
May 29	20.82 m	6.43 m	5.54 m
May 30	20.92 m	7.72 m	5.32 m
May 31	22.40 m	8.43 m	5.45 m
* Note: May 5 & May 19 data sets have L2 disturbance.

The table at left summarizes the accuracies when progressively including the different corrections described above. Click on any of the results for a full-size image of the underlying 24-hour data set. The numbers in the table represent the +/- limit in meters bounding 95% of the computed heights.

Now that SA has been removed, it is possible to assess the improvement that is possible, and measure the contributions to the overall error seen in a GPS height. The table shows that profound improvement to height accuracy can be obtained if a general model of troposphere error, not based on actual surface measurements, is applied to the data.

The third column shows the improvement if precise orbit and clock data could be used in place of the broadcast information, or if broadcast orbits and clocks could be upgraded. The third column indicates the possibilities if precise GPS orbit and clock data are quickly gathered in a global network, computed and predicted over short time intervals, and made available through the Internet. The third column also gives an indication of the level of L1 and L2 pseudorange multipath present at ERLA, and provides a bound on residual troposphere error.

The time series of May 5 and May 19 show disturbances that are not present in the single frequency, L1-only, data.  These disturbances are most evident after the troposphere correction is applied.  It is unlikely that a local influence, such as an object, could create a multipath reflection on L2 but not on L1.  Some form of radio frequency interference or receiver malfunction may be the cause.

For additional information:

Dr. Richard Snay
National Geodetic Survey, NOAA
Richard Snay

Return to the GIAC REMOVAL OF GPS SELECTIVE AVAILABILITY page for more results.