[Talk-GB] accurate GPS

Warin 61sundowner at gmail.com
Wed Oct 9 22:43:46 UTC 2019


On 09/10/19 22:57, Simon Ritchie wrote:
> > You'll need a GPS receiver with the capability of outputting carrier 
> phase data (u-blox receivers will do this) and ideally a 
> well-characterised external antenna (these are quite expensive).
>
> That's very useful.  Thanks.  I am indeed using a uBlox device as my 
> base station.  I ran it for 48 hours and got accuracy to maybe half a 
> metre, but I haven't tried post-processing thedata.

What do you mean by the term 'accuracy'?

Ubox use the term 'precision' which in metorology means repeatability in 
common language.

Something can be highly repeatable but be a long way from the true value.

In this instance ubox are only evaluating their chip.
In order to obtain the best measurement you need an expensive antenna, 
pay for the correction data at the time of measurement and apply that 
correction data.

Claiming a 2 cm accuracy without stating the need to apply the 
corrections is misleading.
Ubox claim a 'precision' - not the same thing as 'accuracy'.

Be very careful in your use of terms, some people will become very upset 
if you claim a 2 cm 'accuracy' when all you are giving them is 'precision'.
I would suggest you research metrology terms.
You might start with https://www.npl.co.uk/resources/gpgs ?

> However, that still leaves the fundamental problem:   I can (and will) 
> publish the kit of parts for making your own base station.  You could 
> use something similar to build a rover or you could buy one off the 
> shelf.  According to the ads this will give you an accuracy of 2 cm, 
> but how will you check that you really are getting that accuracy?

True 'accuracy' can only be determined when you know the 'true value'.

Ultimately no one can really determine the 'true value' they can only 
estimate it and, if they are metorologists, give an uncertainty 
statement on that value.





More information about the Talk-GB mailing list