As explained in the blog post “…and a phone to steer her by,” satellite positions on mobiles are impressively accurate nowadays. Here’s how I confirmed that for my own phones, developing thoughts in a post last February. As a result of many further checks, I’m now much more confident in my phones’ accuracy and usefulness on board than I was then.
Immediate checks on accuracy are straightforward. Creating a log of position error for longer term checks is a little more complicated.
Position error: the simplest way to check your phone’s immediate level of accuracy is to download a free app such as GPS Essentials, GPS Status and GPS Test, which were available on PlayStore for my Samsung phones. Below is a screenshot of one of the apps, GPS Test, taken in the open air in inner London and displayed on a Samsung S20+.
As the image shows, there were 40 satellites visible, and 34 were available for fixes. On the left hand side you will see it says the error is 4 metres, which meant I was inside a circle of that radius 95% of the time. In the middle of the bottom of the screen, there are three sets of initials over numbers, and the middle one is the most significant, HDOP (Horizontal Dilution of Precision). The lower the number the better the fix, and while Yachtmaster textbooks say an HDOP of 1.4 is good, here it is well below that at 0.8.
A Royal Institute of Navigation web lecture notes that while it is common knowledge that at least four satellites are needed for a good position, above 10 there is very little increase in accuracy, so there is clearly no shortage of options here. My phone is receiving from the Russian GLONASS satellites (shown as rectangles), Chinese BEIDOU satellites (crosses), European GALILEO SATELLITES (plus signs) as well as the US GPS satellites (discs). Note that we should not in fact be calling the worldwide satellite system GPS, which is just one of its components. GNSS (Global Navigation and Satellite System) is now a much more accurate term.
There is much more information on the screen. For example, green shows satellites used to determine the location, yellow is where all the orbital data is available but not used, blue is where there is only rough orbital data, and grey is where it is unavailable. Not all phones yet support GALILEO satellite signals.
To be sure of testing satellite signals only, go to settings and switch off the so called high accuracy location. My five-year-old Samsung S7 had a ‘satellite only’ setting which was unambiguous, though my 6 month old S20+ only says ‘WiFi and Bluetooth scanning off’. To be doubly sure you are measuring satellite-only positions, you can use your phone in airplane mode.
Tracking accuracy: To check over a longer period, some form of logging device is needed. I expect there are many options for doing this, but I found I had one installed already: my OMN3 mapping app from Anquet. This is excellent for land walking and trekking maps, but I found the position, distance and accuracy logging function operates wherever you are. This is regardless of whether your position is on or off one of Anquet’s land maps, and it went on logging perfectly well out at sea.
OMN3 from Anquet costs about £30 a year and I use the 1:25,000 Ordnance Survey for walking. You need the app in front of you to follow the instructions to set the logging function (called Tracklog) and also the satellite accuracy measure contained within it, so I won’t explain how to do it here.
With two lockdowns in 2020 and 2021, we were walking between five and 10 miles most days for exercise, in flat and wooded countryside. From the end of 2020, I kept the tracklog working on the old Samsung S7 (2016 vintage) through to spring this year, so I had a full record of its radius of accuracy every 100 metres or 60 seconds on many different walks, with the phone set to satellite only position finding or airplane mode. I have continued to run logs on the new Samsung S20+ I bought in April.
Even when I was near wifi and phone masts, switching on ‘high accuracy’ location to use them made no difference except among buildings or indoors.
You can choose higher or lower frequrncies for recording, which means you can test accuracy even on a motorway by spacing the log entries further apart.
Most open-air readings while walking showed 3, 4 or 6 metres accuracy and sometimes 8 metres, and only very rarely did they rise to as much as 12 metres.
That’s all very well on land, but a test at sea would be better. In 2020, we did not get much chance to go far, and by the time we got going on the boat this year, I had switched from my old Samsung S7 to a much newer S20+. Among the seagoing tests I did was running the tracking function with high accuracy location switched off over two 180 mile non-stop coastal passages, plus several shorter ones of 40 or 50 miles.
To confirm I really was measuring satellite only, I switched to airplane mode for periods during the passages but found it made no obvious difference. The result of these tests was a log showing readings largely of 4 or 6 metres radius of accuracy and only occasionally as high as 8 metres.
There is no similar accuracy log on our chartplotter, but when I look from time to time at position error readings I have not found anything to beat the performance of an ordinary phone.
I left the phone in my oilskin pocket much of the time, and otherwise just lying on the chart table or with me in my bunk. I made no effort to keep it facing the sky. It might be a surprise to Anquet, but their app makes a fine source of lat, long and distance travelled at sea as well as satellite accuracy, though I hope the weather’s never rough enough for the ascent and descent measurements to be useful!
Conclusion: a 5 year-old and a new Samsung phone both consistently showed a level of satellite accuracy comparable with my fixed chartplotter, over hundreds of land and sea miles.