The accuracy of GPS [brief summer]

I'm always being asked - and have been for ages - how much confidence we can put in the GPS on our tablets and smartphones. At the risk of repeating myself, I persist in asserting that our mobile devices are provided with the most accurate GPS receivers that can be found in civilian use, especially for pleasure boating.

The reason is simple. GPS receiver modules built into our wonderful little mobile gadgets, ultra miniaturized, are made by tens of millions of copies for the leaders of the mobile "smart things" : Apple, Samsung, Sony, etc.  While the receivers fitted to marine instruments designed by Garmin, Furuno, Raymarine and some others, are not sold, for the best, only a few tens of thousands units each year. And as everyone in this world knows, in economics the price of things is inversely proportional to the volume. Tablet and smartphone manufacturers can therefore get the best products at the lowest costs., and even the relatively low margin rates are more than offset by the sales volume. While our marine electronics guys have to buy more expensive, and compensate for the low level of sales by higher margins - and therefore sales rates -. The equation is easy to understand !

What impact on the accuracy of devices ?

In reality : very weak ! Below I have compared, a few days ago, the position given on board my boat parked at the pontoon by three devices :

  • Furuno GP32, external antenna, end of production 2003
  • MLR SP24 portable, integrated antenna, end of production 2000 (company liquidated on June 28 June 2001)
  • iPhone 8 integrated antenna, end of production 2020

Once stabilized after about ten minutes of waiting, here are the results :

To the hundredth of a minute, the position in latitude / longitude is identical, either on a scale of 18-20 meters. The MLR SP24 does not give thousandths of a minute, we can not compare it more finely.

To the thousandth of minute, the iPhone gives a position in latitude of 0,002 mn further south than GP32, so 3.70 meters, and in longitude 0,003 mn further west, i.e. 3.64 meters at Lat 49°N (¹). Which gives us a gap of only 5 meters in the SW. Which of the two is right, I let you choose. However, the gap seems negligible to me, because I'd be very unwise to be shaving rocks with a GPS within 5 meters of the hazards (²) !

(¹) After a refreshing dive into calculating distances in longitude and the Pythagorean theorem
(²) ENCs, Routing and sailing shaves stone


4 Replies to “The accuracy of GPS [brief summer]”

  1. GPS gives good information on position and speed : ground speed that Air France pilots ignored during the crash in the South Atlantic . On the other hand, general public GPS are doubtful when at altitude

  2. GPS accuracy is one thing, that of the cards date of their last statement.

    Most of our maps were raised before the existence of GPS, by triangulation methods. This there have been corrections, adaptations and other adjustments. But still be more or less 5m, it's a half boat length for most of us.

    And to add to the confusion, let's not forget the geodesic reference grid. The one on our map is not necessarily the same as the American WGS84 grid which serves as a reference for the GPS. Have you ever wondered what is the benchmark grid of the Chinese Beidou or Russian GLONASS or European Galileo system?? Today the receivers in our latest iPhones use at least 3 of these systems in parallel… To increase the precision they say.

    1. All IHO member hydro services have adopted the WGS84 coordinate system. GPS operators also for civilian frequencies which are common to GNSS, including Beidou.

  3. Thank you Francis
    new consumer GPS chips will hit the market in new smartphones, on Android for now. The precision will be better and above all more stable and less sensitive to environmental noise between the satellites and our equipment.
    For example the series 8 from u-Bloc
    I expect one for the end of September with a GNSS NEO-M8N chip, which thanks to the large distribution has this one for less than 10 € (manufacturer price).
    I will give you the result of my tests.

Comments are closed.