[Qgis-user] GPS issue on android tablet

[Mike Shand]

Zoltan,

I suggest you create a Csv file and plot the points on a large scale map 
or hi-res satellite image.
You say "varies very widely", a quick check indicated the max. variation 
between your points is approx 8m and the min. approx 1m.

I would say if you are getting 3m GPS accuracy that is pretty good if 
you are outside the US or Europe.  I assume you are using a dedicated 
GPS unit attached to your Tablet and not a built-in GPS on your Tablet.
Bear in mind that the 3m accuracy deviation (imagine a 3m radius circle 
from where you mark the point) can be in any direction, so if you are 3m 
E out with first reading and 3m W with the second reading the error 
between the point could be 6m.

Mike


On 04/06/2014 10:14, Zoltan Szecsei wrote:
> On 2014/06/04 11:07, Giuseppe De Marco wrote:
>> hello, I've been using Qgis on android tablet (samsung galaxy tab 2)
>> and I've been experimenting GPS connection.
>> The issue is that even if Qgis finds and use several satellites (9 or more)
>> precision never gets under 3.0 meters but what is strange is that
>> coordinates value fro the same spot varies very widely:
>> Time                Coordinates                       Altitude
>> 14:39:36       41.69528959 13.26464738      343.1
>> 14:39:54       41.69529880 13.26466761      341.8
>> 14:40:06       41.69528519 13.26470222      340.4
>> 14:41:18       41.69528944 13.26472418      341.1
>> 14:41:42       41.69529608 13.26471821      341.8
>> 14:41:54       41.69530621 13.26470892      342.5
>>
>> I have colleagues who have experienced the same.
>> Has anyone else experienced the same issue?
>> Is that depending on GPS device or could it be a software related
>> issue?
>>
> There are probably 10000 reasons for this.
> I am surprised that you even think you are getting sub 5m accuracy.
> Please plot the points on a cadastral map (or some basemap that is
> "guaranteed" accurate) and see what you are really getting.
> You also don't say whether you are in a tree'd or urban (or whatever)
> area - but observing 9 satellites suggests somewhere in the open.
>
> For professional GPSes, to get sub 10cm accuracy, you need to observe
> for 30-60 minutes, and do post processing.
>
> I am not a surveyor, so my info may be not 100% - Please can a surveyor
> or someone well experienced in these issues chip in.
>
> Regards,
> Zoltan
>