Determining an Accurate Reference Point with Delorme PostPro and Garmin Accuracy Estimate

A. Post-Processed Accuracy - Delorme GPSPostPro

Overview - Differential GPS: Since I was aiming to measure the impact of SBAS, I needed an accurate reference point, but GPS accuracy is question that divides, with some people claiming they achieve a 3 meter accuracy based on estimated readings on their GPS receiver. Experts have shown though that since SA was lifted in May 2000, "simple" GPS with no outside assistance can achiever an accuracy varying between 7 and 10 meters at a given point in time, mostly due to the errors generated when the signals travel through the ionosphere. The good news is that this error can be calculated at a reference station whose position is known very accurately. This is beyond the scope of this article but the idea is that this computed error is then used by the GPS receiver to recompute the position as if it hadn't been present, i.e. differential GPS. Now this can be done either in real time with expensive equipment (or SBAS to some extent in theory) or after the fact with the post-processing of a recorded log for sub-metric accuracy, a vast improvement. There are many professional packages around, by Trimble for instance, but there is also a consumer solution, Delorme's GPSPostPro 2.0 that can either process a dynamic baseline or a fixed point, and this is what I was interested in

Tools : Delorme Bluelogger, GPSPostPro v2 and Delorme Earthmate
Location: IGN Geodetic point in Trilport, centimeter level accuracy, position P1
Reference Station: MLVL Marne la Vallée

Method : Data logged for 30 minutes using the automatic mode programmed with the BlueLogger Manager PC Program. This allows you to have the logger switch itself off automatically after x minutes of operation, I chose 30 minutes as this is the "sweeet spot" for post-processing per Delorme's testing

Results: GPSPostPro gave me a post-processed position (P2) with a stated accuracy of 84cm. I then calculated the distance from P2 to P1 and obtained a distance of 49cm consistent with the GPSPostPro report.

Reference point calculation: having now gained confidence in the accuracy provided in the GPSPostPro report, I used the same method to determine the position of the reference point I would use for SBAS testing. I logged data for 30 minutes during 3 separate sessions and obtained accuracies of 93, 77 and 59cm.

Conclusion

Thanks to Delorme's unique and affordable Post-Processing suite (BlueLogger, PostPro software and Earthmate Receiver, I was able to determine a reliable position I could now use for my SBAS testing project in the next pages.

B. Non Post-Processed Accuracy - Garmin receivers

During the geodetic marker testing I also recorded data with a Garmin 60Cx and a Garmin Forerunner 205, here's what I obtained:
1. Garmin 60Cx - measured instant accuracy: 7 meters - Per Garmin : 7 meters
2. Garmin 60Cx - measured average accuracy: 6.8 meters - Per Garmin : 3 meters
3. FR205 - measured instant accuracy: 3 meters - Per Garmin: 7 meters

While the differences are not large, these results do go to show that Garmin's estimated accuracy is just that, estimated. I was particularly disappointed in the unrealistic "average accuracy" indicated by the 60Cx.

C. Links

1. Forum discussion about the BlueLogger - http://www.gpspassion.com/forumsen/topic.asp?TOPIC_ID=10099
2. Coyotebush's Kinematic Project.