GPS Satellite Predicted Visibility and Performance in Afghanistan ----------------------------------------------------------------- I have taken a look at predicted GPS visibility and DOP performance for Kabul on 11 October 2001. For receiver coordinates, I used the coordinates of the intersection of highways A76 and A01 in the city (34-31-12 N, 69-9-36 E) with an approximate height of 1800 metres. As far as I have found out, all 28 operational GPS satellites were healthy and transmitting usable signals all day on 11 October. The signal-in-space performance report for 11 October from the GPS Support Center is in this directory (see below for further comments on this report). Also in this directory is a QuickTime movie showing the elevation and azimuth of the GPS satellites over Kabul at 10 minute intervals on 11 October. Positions are based on the GPS almanac for week 111 (a.k.a. week 1135 = 1024 + 111) which began on 7 October. The mask angle is 10 degrees. Also in this direcxtory is a PDF file of a plot of GDOP, PDOP, and HDOP at 0.5 hour intervals based on all visibile satellites above the mask angle. To see the DOP values corresponding to the best 4 satellites in view, you can use the China Lake on-line predictor . The DOP values are quite low and I don't think they would change very much for other locations within Afghanistan. The predicted best-4 world PDOP assessment for 11 October by the GPS Support Center is also in the directory. It shows for all of Afghanistan for a 5-degree mask angle and assuming a 0 metres altitude, that maximum PDOP over all of Afghanistan is between 2 and 4. To get estimated horizontal position errors (2 d.r.m.s roughly 95%), multiply the HDOPs by twice the representative total user equivalent range error (UERE). Currently, the typical PPS signal-in-space only range error is about 1.5 metres. To this must be added tropo delay error, multpath, and receiver noise. It is hard to give a definitive figure (especially for multipath) but we might use 2 to 3 metres for the total UERE. Using 3 metres for UERE and an HDOP of 1 would give a 2 d.r.m.s. value of 6 metres. At good U.S. sites with high quality civil receivers, the SPS is currently providing 2 d.r.m.s. errors of 5 to 7 metres. So 6 metres for an all-in-view PPS receiver is a bit pessimistic. The corresponding CEP value would be about 2.5 metres. The 3-dimensional r.m.s. error is obtained by multiplying the total UERE by the PDOP. Using a PDOP value of 2, we get a typical PPS 3D r.m.s. error of 6 metres. The corresponding SEP would be about 4.7 metres. In the directory, you will find the GPS Support Center global performance reports for 11 October (the performance reports specific to SW Asia appear to be classified). For 11 October, the global GPS CEP is given as 1.4 metres, based on the errors noted at the Control Segment tracking stations, a mask angle of 5 degrees and the use of the best-4 satellites. This figure is about 1/2 of my conservative estimate (1.4 m vs. 2.5 m) and may result from the fact that the range errors at the tracking stations likely have low multipath and receiver noise components and perhaps even a lower residual tropo delay error giving a lower total UERE than might be typically seen in a battle environment. The published SEP is also lower; 2.0 m vs. my conservative estimate of 4.7 m. Furthermore, the Support Center values are helped by the use of a lower mask angle (5 degrees vs. my 10 degrees) which potentially lowers the DOPs but only the signals from the best-4 satellites are taken rather than all in view as I assumed, which has the opposite effect but perhaps not of the same magnitude. Note that since there hadn't been any recorded GPS outages between 7 and 11 October, the results for the 11th likely apply to all of the bombimg days up to Friday, the 12th. On the Friday, PRN28 was repositioned (slight, normal orbit adjustment) and was unusable for most of the day: "GPS SATELLITE SVN44 (PRN28) WAS UNUSABLE ON JDAY 285 (12 OCT 2001) BEGINNING 0745 ZULU UNTIL JDAY 285 (12 OCT 2001) ENDING 2234 ZULU". This would have had a small effect on GPS performance on that day. Updated 14 November 2001. =============================================================================== Richard B. Langley E-mail: lang@unb.ca Geodetic Research Laboratory Web: http://www.unb.ca/GGE/ Dept. of Geodesy and Geomatics Engineering Phone: +1 506 453-5142 University of New Brunswick Fax: +1 506 453-4943 Fredericton, N.B., Canada E3B 5A3 Fredericton? Where's that? See: http://www.city.fredericton.nb.ca/ ===============================================================================