GPS & Altitude.... Who to trust?

[Mav359]

I have a Samsung S4, and Bluenext GPS dongle and i have used google earth

Whilst they all agree the GPS location they all disagree as to my altitude

S4       = 122m

GPS    =   80m

Gog E =   74m

So i want to know who if anyone is telling me the truth and is there a better way of obtaining altitude measurements (for free)

Cheers

[Steve Ward]

OS Map ... contours never lie ... 

[AstroTux]

GPS fixes for altitude are notoriously unreliable. For this reason good hand-held GPS units for hill-walkers have an integrated barometer. I would trust the OS map more than any GPS.

hope this helps,

Alan

[Mav359]

The OS maps online seem to suggest that Google earth is right so i am going to use that as my reference from now on

73m it is!!!

Cheers Guys

[perfrej]

It is not that the GPS is unreliable, rather that it uses the WGS84 ellipsoid approximation of the earth's surface. Said ellipsoid is most accurate over North America (surprise!). In the Stockholm area, where I live, it differs by about 32 meters - and that's one of the smaller ones.

Some commercial GPS navigators contain compensation tables for areas which make them report a false GPS altitude, but one that corresponds more to reality.

The only true data for altitude is the SRTM data sets acquired from accurate radar satellites. Upon that data (and other sources), our national geological survey authorities base their highly accurate altitude grids - some with down to 5 m resolution.

Babbling away here... Anyway, never trust a GPS' reported altitude, but do look at Google Earth data, which incidentally is based to the most part on SRTM data over the water-line and bathymetric data from IBCAO and GEBCO (the former of which I have actually participated somewhat in the acquisition of on my scientific journeys on the Swedish icebreaker Oden).

73m it is

/per

[russ.will]

'Scuse the question, but at such low altitudes, or given that the range of answers thrown up cover about 50m, why does it matter?

Or to put it another way, how high do you have to be before it matters and to what?

Russell

[D4N]

What you are using the altitude for matters too, if you are using it in an application that uses WGS84 then feeding it a corrected height will induce errors. Be especially wary if it is using a 2D calculation for position and using manual height to complete the fix.

Height totally depends on the reference you are using it against, even sea level is arbitrary. Sea level could mean HAT (highest astronomical tide), or MHWS (mean high water springs), or LAT (lowest astronomical tide), or MSL (mean sea level) or it could just refer to the zero height on the spheroid you are using.

There are other spheroids and datums that you can use, normally ED50 is preferred in Europe but there are many others. The most important thing is to ensure that all your equipment is either using the same variables or converting them correctly.

Sent from my iPad using Tapatalk

[ollypenrice]

I love Per's post but, then again, how on earth are these differences going to affect you?

Olly

[russ.will]

That's my point.

I've heard of all sorts of astro loonies that live up mountains, but seldom heard altitude discussed as an issue.

Russell

[perfrej]

Ah, it affects you when you involve variables such as:

* atmospheric refraction

* conversion between J2000 and local topocentric coordinates

Then again, this is mostly important for you if you do unguided imaging with sky modeling and expect your mount to hit your target unaided.

There is also the thrilling technological aspect. Take my GPS unit in my boat. It uses IALA differential corrections, an advanced atmospheric model and can decide the boat's position to an accuracy of 10 cm - all the time. Charts are 10m precision so you may ask yourself if this matters. It does in the Stockholm Archipelago where the rocks are frequent and you need to eliminate all your errors. In more statistical environments, the smaller error is averaged out by the larger one. In boat navigation terms, that means that you will miss the rock, statistically, just about every time out of one million passes, but in reality terms it means you may very well hit it if you navigate too close.

I'm at it tonight

/per

[ollypenrice]

Heheh, my dear friend Per loves technology. I love living up a mountain!!! I have the good sense to guide since not doing so would mean selling my fun car, buying a mount like Per's and then paying Per to make it work. He's extravagently expensive, I warn you!!!         

Olly

[perfrej]

Haha!

I kind of like both the mountain and the technology, which is why a good part of my technology is currently on a mountain in Provence...while I am in my living room sofa!

We both drink the same stuff and as it happens I am currently ingesting a glass of Cotes de Provence Rosé. The technology calls and I see that Olly's site is dark and reasonably clear, so time for a few vcurves

/per

[ollypenrice]

Too much moon for me! Bed time.

Olly

[Mav359]

WOW did that get out of hand.....

I only asked because ASCOM and other apps like align master, stellarium etc ask the question so i figured it mattered to the accuracy of those applications and so i just wanted to be as accurate as possible.

Cheers guys

[perfrej]

If a topic doesn't get out of hand it may not warrant discussion

/p

[StarryEyed]

Aarrgghhhhhhhh........

ED80 ETRS89 UTM. Been an issue at work.

Did you know there are three systems for longitude and latitude. Also that due to relativity and GPS sats moving through time slower than us the clocks on them are slower and if this was not taken into account you GPS would be a mile out this time tommorrow.

[perfrej]

It is ore complex than that... There's projection, ellipsoid, reference datums and more. The Einstein sympton, however, is not one of GPS' problems as it is all handled properly at all ends.

/per

[StarryEyed]

I think we are moving slower I cant remember. If your curious there is a great spreadsheet on the OS for doing the calcs for the translations.

Always the OS their the best.

[perfrej]

OS?

[AstroTux]

OS?

Ordnance Survey. The original map makers.

Hope this helps,

Alan

[perfrej]

OK, I get it. Your government agency for geographical surveys. We have them too in Sweden. Coordinate transforms are rather tricky to get right, and sometimes you need to use 7-parameter tranformations to get reasonable accuracy. In Sweden, old coordinate systems are difficult to get rid of, and as late as in the seventies most municipal areas had their own references. They finally agreed aön SWEREF2000, which is very cose to WGS84 but not quite... Bummer.

Interesting fact; in my boat, in which I am sitting right now, the nav system sports an accuracy of less than 100mm. I KNOW where the boat is

/per

[Alien 13]

WOW did that get out of hand.....

I only asked because ASCOM and other apps like align master, stellarium etc ask the question so i figured it mattered to the accuracy of those applications and so i just wanted to be as accurate as possible.

Cheers guys

I only ever use a polarscope with my kit no fancy software/hardware and i wouldnt let a laptop near it but i do rely on stellarium to give me the hour angle of polaris but it never agrees with any of the other polar finder apps out there so my only reason for trusting stellarium is that its one of the few that do ask for altitude info. 

I would love to know which app/software can give the real answer to the position of polaris.

Alan

[D4N]

You could just look in an almanac and work it out for yourself. The maths involved is not difficult.

If you want to convert it to alt az it's a little more complicated but still doable.

Sent from my iPhone using Tapatalk