GPS accuracy

[markpierce]

I'm more concerned with the accuracy of charts.

[phiggins]

OK so we finally figured out the pattern is just the highlighting. If I turn off track highlighting and raise the Tracking Precision to high and run my GPS test for a few hours I get the attached results. It indicates the worse GPS reading was off by 9 meters for this GPS at this position at this time. Just nice to know.

[transmitterdan]

Not really that useful though. That particular "accuracy" is not going to be the same on the high seas. So depending on it is not a good idea.

[colemj]

Originally Posted by phiggins

It indicates the worse GPS reading was off by 9 meters for this GPS at this position at this time. Just nice to know.

Hopefully it didn't cause you to take a dump in the sink or anything too bad…

Mark

[phiggins]

Not really that useful though. That particular "accuracy" is not going to be the same on the high seas. So depending on it is not a good idea.

Well I would think it would be a lot better where the GPS is open to the skies. Here the GPS is in the house, surrounded by trees and 100 meters from high voltage power lines.

[bcn]

Right, open sky (even with a GPS indoors but in cases of a steel/aluminium/carbon fibre house), no obstructions and no multi path.
Should do better.

[transmitterdan]

Maybe it will be better or maybe worse. You just can't be sure is all I'm saying. Therefore, there is no point in knowing the accuracy for it will surely change. There are no decisions that can be made on the basis of GPS accuracy in the range of meters. For close quarters work it's better to use other means such as depth and eyes.

I'm not saying GPS is useless. It is very good and more reliable than any other technique.

[bcn]

Dan,

we agree: no automatic docking please

Hubert

[barnakiel]

Originally Posted by transmitterdan

What method of determining speed and position has greater accuracy than GPS?

Off the top of my head I could imagine at lest two:

- visual,

- mixed gnss.

Still, mixed gnss will likely incorporate gps as a component.

Possibly, maybe, an inertial system could be more accurate. But this one is only a guess.

b.

[HappySeagull]

9 meters is pretty good, given where you are but If you are looking for trouble, consider the Vertical plane.. That'll get you trying to choose between geoid models. Most boats are airborne or submarines if you ask the gps on wgs84.

agree usually faster,better fix on water or prairie. I have drifted into dark coves at night with a pikepole feeling for steep shore and have found gps to be pretty darn good

[bcn]

Yes, a combined Inertial/GNSS IMU will give a quite precise and high bandwidth solution.

Positioning will not be better than the specs of the GPS - so depending on the type of receiver and differential corrections but you can get rid of those spikes of the GPS shown on the screen shots = 3m bare/ 1.80m WAAS/ < 1m with differential corrections in real time.
And you will get a very good heading signal.
Not a inexpensive solution though...

Hubert

[bcn]

Originally Posted by HappySeagull

9 meters is pretty good, given where you are but If you are looking for trouble, consider the Vertical plane.. That'll get you trying to choose between geoid models. Most boats are airborne or submarines if you ask the gps on wgs84.

agree usually faster,better fix on water or prairie. I have drifted into dark coves at night with a pikepole feeling for steep shore and have found gps to be pretty darn good

For the Geoid correction you will need fine tuned gravity models. Do exist but as we are on the surface not that big issue.
Vertical precision is aprox. the double of horizontal precision. Or 2 times the error.

[barnakiel]

Originally Posted by bcn

Yes, a combined Inertial/GNSS IMU will give a quite precise and high bandwidth solution.

Positioning will not be better than the specs of the GPS - so depending on the type of receiver and differential corrections but you can get rid of those spikes of the GPS shown on the screen shots = 3m bare/ 1.80m WAAS/ < 1m with differential corrections in real time.
And you will get a very good heading signal.
Not a inexpensive solution though...

Hubert

We are talking of two possibly similar takes, what I meant exactly is a mixed GLONASS/GPS or a mixed GLONASS/BEIDOU (GLONASS/BEIDOU/GPS, etc.) method.

I do not see where they get expensive. My sub-USD100 does it as well as a handful of brand devices from say Garmin, etc. (e.g. Garmin etrex10)

For a standalone (one gnss based, not a combined signal positioning system) positioning MAY be better, depending on where you are and how the other two systems develop (soon: three other systems, when Galileo will (?) become available).

As you noted: today positioning seems on pair with the spikes removed. I think the spikes are of minor importance anyways when smoothing algorithms are employed (which is the case with gps anyways).

b.

[bcn]

The receivers that can take several GNSS systems (GPS, GLONASS, Beidou, one day Galileo) are not improving precision so much. You have more satellites in favorable positions and thus a more stable reception and less spikes.
This helps especially in complicated locations like cities or roads with a some vegetation. Or in a fjord.
GPS + GLONASS is already a widespread feature in modern chip sets.

People promoting Galileo expect improved precision using the additional L5 frequency. We shall see...

Don't expect the same performance for a GNSS sensor as from a combined inertial/GNSS IMU.

[Lepke]

From the point of view of an old sailor that crossed many oceans, many times with a sextant. On a ship in mild weather with a 3 star fix, I was lucky to plot a triangle with sides less than a mile.
GPS isn't perfect, but neither are your charts. Most charts are based on old manual inputs. Sighting positions were probably off more than 15 meters. The field people didn't have anything as accurate as gps until recently.
Be happy with what you've got.