While this topic has taken on a life of its own, most SEESAT users do not need accuracies better than 100m for satellite observing applications. Certainly if you are even miles away from your actual position, you can compute satellite predicts good enough to see with binoculars. However, if there is a valid requirement for accuracies less than this, then there must be enough data taken over a long enough period to justify statistical statements for accuracy purposes. Ron's recommendation for this is correct as noted below, though perhaps the length of time does not nearly have to be so long. Based on my personal testing at known benchmarks here after S/A was turned off, this period of time does not have to be more than a few minutes to get to 5-7m provided the satellite geometry is giving a PDOP that is low enough to maximize accuracy (generally less than 4.00). One might even argue that in a matter of seconds you could reach this goal. For newbies to SEESAT, one should not over emphasize GPS for making valid sat predicts. GPS simply is not necessary for successfully predicting most satellites. Last week I was in Hersonissos, Crete and on the island of Corfu; in both places I used a Garmin GPS III Plus to ascertain my instantaneous position to feed into heavens-above.com to get Iridium flare predicts and CGRO look angles. I took data for 10 seconds and the position generated was perfectly OK for this. In April, I surveyed several sites west of Lusaka, Zambia in preparation for our NASA JSC Astronomical Society June 2001 solar eclipse expedition and the errors using the same receiver were about twice that I am getting now after S/A was turned off. Of course, the GPS readouts were used to enter my location into heavens-above and to use QUICKSAT to generate low altitude sat predicts. All the predicts were right on target; I could have used a coarse map before I left on my trips and the results would have been the same. I have conducted numerous such expeditions when differential correction was needed from a base station to reduce the single receiver error and I am extremely glad that this is no longer mandated. For general astronomical applications, a better than 30 m accuracy was required for and 15m or even less was a goal of ours. With S/A off, these applications are truly achievable with inexpensive single frequency receivers. Before S/A was turned off, taking readings every 2 seconds, for example, and averaging them over 4 hours was enough to get us to a 30m accuracy (by collecting lots of data and averaging) with 95% confidence using a single standalone receiver under normal geometry from the GPS constellation. One has to be careful even now because the accuracy does change with time as the satellite geometry moves to a less favorable orientation. Some manufacturers such as Trimble have a web site program to predict this geometry, making it easier to know which periods of time at a particular location are best to avoid. Paul - I would caution folks not to accept all the low accuracy values unless you take a full day (or more) of data. Typically accuracy is mentioned as the data points containing 95% of the observed points. At this time I am accepting the 7-9 meter values I quoted earlier because these are 24 hour measurements. I have not yet "surveyed" my normal observing position with my GPS but I suspect I am plenty close as it is. Certainly GPS now allows a person to observe from a new site without having to get topo maps, etc. Ron Lee ----------------------------------------------------------------- Unsubscribe from SeeSat-L by sending a message with 'unsubscribe' in the SUBJECT to SeeSat-L-request@lists.satellite.eu.org http://www2.satellite.eu.org/seesat/seesatindex.html ----------------------------------------------------------------- Unsubscribe from SeeSat-L by sending a message with 'unsubscribe' in the SUBJECT to SeeSat-L-request@lists.satellite.eu.org http://www2.satellite.eu.org/seesat/seesatindex.html
This archive was generated by hypermail 2b29 : Wed Jun 07 2000 - 04:55:06 PDT