BJ none of the gps units actually record doppler speed, what they record is SOG (speed over ground) derived from the doppler shift of the satellites signal as it is received by the gps. It calculates the speed relative to the gps satellites in orbit. Doppler speed is the velocity through space relative to a satellite, the gps unit uses complex algorithms to calcualte the speed over the ground from this information and records it along with the Course over Ground COG .

All gps units calculate SOG and some like the NAVi store it for retreival. The NMEA standard was developed so gps and computers could exchange information, the NMEA information contains the SOG data which is read by all navigation software. There are numerous gps units available that store and output this information but the NAVi is one of a few which display the data on built in screens.

Some units like the NAVi also store and output the data in binary format which needs to be parsed and decoded to extract the SOG information but essentially it is all derived from the same doppler calculations the gps unit calculates.

DGPS and WAAS are corrections to the positional data transmitted from the satellites. The positional data (lat and long) rely on the doppler frequency shift in poor reception situations to derive a dead reckoning calculation of where it thinks the gps should be based on course over ground and speed over ground. The algorithm determines if it uses this data or accepts the positional data to log a fixed position. Where positional data lacks in accuracy is in the rounding of the data to a set decimal level, this leads to the so called "grid effect" you have no doubt heard about. Binary data gives you access to an extra decimal place with a resultant improvement in accuracy. With SOG the decimal level of accuracy is the same for both NMEA and Binary data. I first looked at binary data as a method of post processing the data to improve the accuracy to sub meter levels, this still remains possible with other gps units available that can be configured to produce the information required for post processing.

Hope this gives you an idea why the NAVi is more accurate than the Garmin units that cant save the NMEA or Binary data for analysis.

The SOG is derived from the 3 dimensional Velocity calculated by the change in frequency as the gps moves in relationship to the satellites. It is independant of lat long position.

The reason the NAVi is more accurate is because it can store the SOG information as well as the status of the satellites it derived the calculations from. The Foretrex for instance saves a .gpx file which contains time, lat, long and elevation. The trackpoint accuracy is not that different on the NAVi but the velocity is more accurate as it is not derived from positional data. If the Foretrex could save the same data it would probably be to the same level of accuracy as well.

We are also talking of two different types of speed. As BJ said if you use the lat/long and the time taken to move from one trackpoint to another you get true speed, by definition distance divided by time; speed is a scalar quantity and does not keep track of direction. Using the doppler frequency shift results in velocity, a vector quantity that has a directional component. Where the confusion arises is when you try and prove where the velocity occured as the position was not part of the equation in measuring velocity. It is essentially movement through space.

BJ said...

OK, so if I am understanding correctly, all GPS units display a speed over ground that comes from the satellites. That is each brand displays a speed on the display and that speed is accurate no matter what the brand.

The Maximum speed shown on the display is the maximum spped as determined from the satellite and should be the same on each brand of GPS and therefore be as accurate as each other.

yes and no.

The actual calculations used for computing the SOG value as seen on the display, (often) uses a proprietary algorithm - but for all intense purposes, the implementation details are close enough so that the display value can be very similar between different manufacturers.

That said, from experience it appears that the Foretrex uses a 5-second average for its "max" value, while the Navi doesn't have a filter on it max value.

Originally the Navi made logged data available using the NMEA format (look on the web for more info about this format). One realisation was that this data contained the extra decimal place of precision - for Lat/Lon, compared to the data that was available from the Foretrex's logged data.

ie: by looking at a lat/lon track from a ForeTrex, it is easy to see that the data saved is limited in precision, simply because the spacing between adjacent data points is far to large; if you look at a Navi lat/lon track, you can see an order of magnitude better precision.

Some time passes and Roo finds that the SOG was also available. Further reading and testing found that the SOG values (which are derived from the doppler calculations, not the lat/lon calcs) contained at least an order of magnitude less error for the calculation of speed for each second - than would be possible from calculating speed from adjacent lat/lons.

Also importantly, the doppler calculations of speed are less affected by atmospheric conditions, than the equivalent lat/lon calculations - thus for our requirements, worst case error is more predictable for the doppler speed.

Now if you take a recent Garmin ForeRunner, its lat/lon values are actually more accurate than the Navi's lat/lons as it uses a better/more-sensitive GPS chip - but is still less accurate than the SOG values.**

The difference is that the Navi saves the satellite data which can be downloaded but other brands only save the track points (series of lat/longs over time) and then uses these to determine average speeds etc etc. This is where the difference is because the raw data is not as accurate.

So to answer => the Navi saves more stuff, with (usually) more precision, than the most of the other consumer devices. Its also kinda cheap and has more features.

And all this ignores the "real-time" data which can be retrieved from the devices, when they are permanently connected to a computer.

** This applies for short-time periods, say up to about 2 minutes, that windsurfing uses for its speedsailing. Using a longer timeframe, the accumulated error of doppler may be more or less than the lat/lon errors, depending on how you choose calculate your run.

Aside: AFAIK, the Navi is very much capable of using DGPS and WAAS (as is the ForeTrex), if the signals were available and close enough to the devices.