GPS to grid reference

Quoted message said:

But how are datums (ugh! data is plural of datum) defined? That's why it's
a lousy article. It says there are lots of them but not why.

Why ? For the same reason why there were different videosystems : everyone
wants to make its own and wants to be worldleader :-(

--
"Beannachd leibh"

Theo
www.theosphotos.fotopic.net

In article <[email hidden]>, Bernard Hill
<[email hidden]> writes

Quoted message said:

In article <[email hidden]>, Dominic Sexton
<[email hidden]> writes

Quoted message said:

In article <[email hidden]>, Bernard
Hill <[email hidden]> writes

Quoted message said:

As a mathematician, I know there is one axis of rotation (it may move
slowly over millennia but for now there is one). There is one
centroid of the earth, and together these define the equator and the
two poles.

You have assumed that the earth is a perfect ellipsoid or sphere. It
isn't.

No I have not. Whatever the shape, there is one centroid and one axis
of rotation.

How do you determine the centroid of the earth? Is it the mathematical
centre of the body formed by the shape of the surface of the earth? Does
it take into account the uneven distribution of mass in the earth?

Originally the shape, rotation and centroid location were determined by
celestial observations. The observations were obviously made from the
ground and those from different observatories were combined to model the
size, shape, rotation and location of the earth. These models were
specific to the locations for the observations. That is why when they
were made elsewhere they got different results.

As observations improved the models were refined. There have been a
number of realisations of the zero degree meridian going through
Greenwich observatory. Some history:

http://gpsinformation.net/main/greenwich.htm
http://www.nmm.ac.uk/server/show/conWebDoc.416

Surveyors need a model of the earth to make meaning from their
measurements. The models determined by the astronomers were in that
country / region were used because they naturally fitted the shape of
the earth in that area.

The model will most probably not tie in with the reality for the whole
earth but it does tie in with the local shape of the earth.

I hope that explains why we have a number of geodetic datums.

--

Dominic Sexton

Bernard Hill said:

But they can't be. So don't assume they must and define them in a more
logical way.

Well what is your "more logical way"?

Quoted message said:

Quoted message said:

so given angular distances along it are constant,

??

x degrees along a parallel, angular distance should always be a
constant c km lomg. But if the surface is irregular (as it is),
that won't be the case. This is Not Good in a deterministic
measurement system...

Quoted message said:

Of course. But that doesn't mean to say that there isn't a set of
coordinates which describe points on the surface. In mathematical terms
it's called Spherical Polar Coordinates: radius, latitude, longitude.
Conventionally called R, theta, phi. All you need to define are where
the zeroes of theta and phi are.

This isn't new stuff: it's centuries old.

It isn't new, but what it isn't is much use in practice on the
Earth. WGS84 gives a much better account of the real shape of the
place than an ellipsoid, which is what you need for sperical polars
to work consistently, accurately and usefully across the surface.
Since maps are of the surface, and not some reference ellipsoid, it
makes a lot of sense to use a reference model that more accurately
reflects the surface you're mapping.

Quoted message said:

As I said, there can be defined one unique set of lat/long.

The issue isn't whether you can do it, it's is it actually useful
in practice. Lat/Long breaks down as a fully consistent system
because its assumed geometry is not the same as the earth's.

You addressed this yourself with the original question: if it's
really simple in practice, why has so much effort been put in to
incredibly complicated and unwieldy alternatives? The answer is
because they actually work better in practice for consistent,
accurate measurement and mapping.

Quoted message said:

In fact I
fail to see how any other method could work.

If it doesn't, then neither does GPS. Which suggests a great deal
more maritime disasters should be happening than do! ;-)

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

Bernard Hill said:

It doesn't need to be a perfect anything. A stony asteroid can have
latitude and longitude.

It can, but think about how useful it will be as a general co-ordinate
system. You can locate anything, yes, but you won't have useful
information about other points. How far away is x from y, given you
have lat and long of each? You have no idea... one of the points of a
useful mapping system is to tell you information like that.

Quoted message said:

But how are datums (ugh! data is plural of datum) defined? That's why
it's a lousy article. It says there are lots of them but not why.

As Dominic pointed out, the WGS84 shape is a good /general/ fit on a
global scale, but it may not be the best thing on a local scale, and it
(or indeed a "simple" Lat/Long model) may not be optimum for your
particular mapping purpose. If you have a local map, and want to get
about locally using it, it will make excellent sense to use a model that
fits your locale particularly well for your local maps, and/or uses
convenient assumptions that don't carry over well to a global model.

For example, all N/S lines in the UK National Grid appear to be
parallel, where clearly that can't be the case on a spherical surface
with a single point North. But it makes getting about on an OS map a
lot easier than marking it with lat/long, so we use that rather than
Lat/Long.

There are lots of datums because there are lots of scales, locations,
co-ordinate systems and functional uses that just work better in
practice with different datums. Would you really want OS Landrangers to
be marked with lines of latitude and longitude rather than the OS UK
National Grid, and supplied slightly curved rather than flat?

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

Theo said:

Quoted message said:

But how are datums (ugh! data is plural of datum) defined? That's why it's
a lousy article. It says there are lots of them but not why.

Why ? For the same reason why there were different videosystems : everyone
wants to make its own and wants to be worldleader :-(

No, it's like asking why are there so many different internal combustion
engine formats. Different configurations and working methods work
better in different places. Diesel mini motor scooter, anyone? Two
stroke HGV? Rotary piston engine on a motorbike? And so with map
datums, you use the best one for your area and purpose.

The basic problem comes down to this: maps are flat, but they portray a
surface which isn't. That is actually very difficult to do
consistently, usably and accurately.

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

"Peter Clinch" <[email hidden]> schreef in bericht
news:[email hidden]...

Quoted message said:

The basic problem comes down to this: maps are flat, but they portray a
surface which isn't. That is actually very difficult to do consistently,
usably and accurately.

Holland is flat but we don't have a system of our own. Normal maps don't
even have lat/long :-(

--
"Beannachd leibh"

Theo
www.theosphotos.fotopic.net

Bernard Hill said:

As a mathematician, I know there is one axis of rotation (it may move
slowly over millennia but for now there is one). There is one centroid
of the earth, and together these define the equator and the two poles.
So we have well-defined latitude lines and one longitude line which goes
through a particular spot in Greenwich.

So why more than one datum?? If there is more than one system of
lat/long then they are all (with possible exception of 1) in error.

Apart from the historical reasons for having more than one datum,
discussed elsewhere in this thread, there is the need to deal with the
practical problems associated with plate tectonics.

For example, Australia is moving NNE at approximately 6cm per year -
see:
http://en.wikipedia.org/wiki/Image:Global_plate_motion.jpg

This means that Australia has moved about 2 metres towards the equator
since 1984. It is not practical to define a position in a global
context and then have the coordinate values of that position change
over time. The solution is to define a datum for the tectonic plate
and then the coordinates in the 'Plate' datum do not change over time.
In the case of Australia, this is defined as the "Geocentric Datum of
Australia" (GDA94). This was identical to WGS84 in 1994, but is now no
longer the same.

In the case of the UK, which is on the Eurasian Plate, positions are
now referenced to the "European Terrestrial Reference System"
(ETRS89). This means that positions referenced to this datum do not
change over time, as would be the case if the WGS84 datum was used
(approx 1.8cm North East movement per year).

The constant movement of the earth's tectonic plates means that is not
possible to define a time unvarying global datum for practical use. In
fact it is believed that, as a result of the 2005 Asian Tsunami, the
polar axis was shifted at the North Pole by 2.5 cm. See:
http://news.com.com/2100-1008_3-5536983.html
This change, which is readily detectable with modern instruments, did
not take a millennia, but was almost instantaneous.

It would not be practical to have to change all coordinate values for
each and every earthquake.

I believe (and I could be wrong on this one) that for accurate work,
different datums/coordinate systems are used on each side of the San
Andreas fault in California. Perhaps someone else can clarify this.

All this means that in this changing world, there is still a practical
need for more than one datum.

Alan Murphy

Theo said:

Holland is flat

Only in a gravitational sense. The Earth's curvature makes it less flat
than a map of the place.

Quoted message said:

but we don't have a system of our own.

Not everywhere does. Norwegian maps use WGS 84, for example. If an
existing model fits well enough there's nothing to gain from doing it
yourself a different way.

Quoted message said:

Normal maps don't even have lat/long :-(

Would you want them to? OS maps have it marked on the borders for
reference, but I've yet to come across any isntance of anyone actually
/using/ that on the ground

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

Message-ID: <[email hidden]> from Peter Clinch
contained the following:

Quoted message said:

Two
stroke HGV?

Some tanks use a two stroke diesel engine.

--
Geoff Berrow (put thecat out to email)
It's only Usenet, no one dies.
My opinions, not the committee's, mine.
Simple RFDs http://www.ckdog.co.uk/rfdmaker/

Theo said:

"Peter Clinch" <[email hidden]> schreef in bericht
news:[email hidden]...

Quoted message said:

The basic problem comes down to this: maps are flat, but they portray a
surface which isn't. That is actually very difficult to do consistently,
usably and accurately.

Quoted message said:

Holland is flat but we don't have a system of our own. Normal maps don't
even have lat/long :-(

There are "Amersfoortse coordinaten" but I have only come across those in
a scientific context.

Roos

The message <[email hidden]>
from Peter Clinch <[email hidden]> contains these words:

Quoted message said:

If mechanical engineering isn't required
then reducing features doesn't necessarily make the unit any cheaper.
In fact, by reducing production runs it might actually make things more
expensive!

At one stage the "holier than thou" Army actually paid substantially
over the odds to have black saloon cars, with NO ASHTRAYS!

Richard

In article <[email hidden]>, Geoff Berrow
<[email hidden]> writes

Quoted message said:

Message-ID: <[email hidden]> from Peter Clinch
contained the following:

Quoted message said:

Two
stroke HGV?

Some tanks use a two stroke diesel engine.

Aye when I was working in oil exploration we used some big off-road
buggies powered by Detroit Diesel two stokes.

--

Dominic Sexton

Peter Clinch said:

Theo said:

Quoted message said:

But how are datums (ugh! data is plural of datum) defined? That's why it's
a lousy article. It says there are lots of them but not why.

Why ? For the same reason why there were different videosystems : everyone
wants to make its own and wants to be worldleader :-(

No, it's like asking why are there so many different internal combustion
engine formats. Different configurations and working methods work
better in different places. Diesel mini motor scooter, anyone? Two
stroke HGV?

There are such beasts, Foden FD6 being one. Also I think a lot of
those yellow school buses over the other side of the pond are two
stroke diesels.

The biggest diesels in the world are two stroke
http://people.bath.ac.uk/ccsshb/12cyl/
--
Phil Cook, last hill: Angletarn Pikes via Beda Fell.
http://www.p-t-cook.freeserve.co.uk/May07/martin9.htm

"Peter Clinch" <[email hidden]> schreef in bericht
news:[email hidden]...

Quoted message said:

Theo said:

Holland is flat

Only in a gravitational sense. The Earth's curvature makes it less flat
than a map of the place.

If that's the case (and I know it is) why is there so much water on top of
Holland ?
.......
Wait a minute...., there's also a lot of water on top of Scottish hills :-)

--
"Beannachd leibh"

Theo
www.theosphotos.fotopic.net

Phil Cook said:

There are such beasts, Foden FD6 being one. Also I think a lot of
those yellow school buses over the other side of the pond are two
stroke diesels.

Well, I bow my head reverently when it comes to folk knowing more about
diesel than I do! ;-)

Though hopefully you see my point...

Pete.
--
Peter Clinch Medical Physics IT Officer
Tel 44 1382 660111 ext. 33637 Univ. of Dundee, Ninewells Hospital
Fax 44 1382 640177 Dundee DD1 9SY Scotland UK
net [email hidden] http://www.dundee.ac.uk/~pjclinch/

Bernard Hill said:

There has to be a cheap GPS which has a readout as a grid ref.

I don't need maps, got those. I don't need track saving, I got maps. I
don't need direction, I got a compass.

But just now and again I get "lost" on a featureless flat top in the
cloud and would love to know where I am as a simple grid reference.

My Garmin GPS III+ give GB OS grid refs. I don't know about the newer Garmin
models but I bet they still make models that do what you need.

In fact thinking about it, as most of us use a GPS in conjunction with a map
it would seem to be a basic function that all GPS units designed for walking
should be able to do. Isn't that why mine offers so many different options,
to cope with the map systems used in other countries? I count 104 on the
III+
--

Regards
Nick

Bernard Hill said:

In article <[email hidden]>, Dominic Sexton
<[email hidden]> writes

Quoted message said:

In article <[email hidden]>, Bernard
Hill <[email hidden]> writes

Quoted message said:

Quoted message said:

Quoted message said:

As a mathematician, I know there is one axis of rotation (it may move
slowly over millennia but for now there is one). There is one centroid
of the earth, and together these define the equator and the two poles.

You have assumed that the earth is a perfect ellipsoid or sphere. It
isn't.

Quoted message said:

No I have not. Whatever the shape, there is one centroid and one axis of
rotation. Slice the earth with planes perpendicular to the axis of
rotation and you have parallels of latitude. The one going through the
centroid you can define as the equator, the rest can be divided up into
parts for the other ones.

Quoted message said:

Quoted message said:

Quoted message said:

So we have well-defined latitude lines and one longitude line which
goes through a particular spot in Greenwich.

So why more than one datum?? If there is more than one system of
lat/long then they are all (with possible exception of 1) in error.

They are all in error to a more or less extent. The earth is not a
perfect geometrical shape. It simply does not fit an ellipsoid
everywhere no matter how you specify that ellipsoid. However it is
mathematically advantageous to model the earth as though it were an
ellipsoid.

Quoted message said:

It doesn't need to be a perfect anything. A stony asteroid can have
latitude and longitude.

Quoted message said:

Quoted message said:

Therefore the various ellipsoid based datums were developed (with
different centres and sizes) to try and fit the shape of the earth in
the area of application.

With the advent of the space age it became possible to determine datums
with reasonable fit globally such as the native datum of the GPS
WGS-84. The average global fit is good but the fit in any particular
area might be poor. This combined with vast quantities of historical
data referred to local datums around the world means that the local
datums have not suddenly been replaced by a 'superior' global datum.

Quoted message said:

But how are datums (ugh! data is plural of datum) defined? That's why
it's a lousy article. It says there are lots of them but not why.

International politics, surveying practicality, and tectonic plate
drift. Britain is changing in position relative to the US very much
more quickly than London is changing in position relative to
Edinburgh, and both US and UK mapping authorities care about local
accuracy much more than the care about getting the width of the ocean
right. So it makes very good sense to relate all national measurements
to a national reference datum point, and then move that (and the whole
country) around as other surveys and politics dictate.

--
Chris Malcolm [email hidden] DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]

Nick Mason noemail@nospam said:

Bernard Hill said:

There has to be a cheap GPS which has a readout as a grid ref.

I don't need maps, got those. I don't need track saving, I got maps. I
don't need direction, I got a compass.

But just now and again I get "lost" on a featureless flat top in the
cloud and would love to know where I am as a simple grid reference.

Quoted message said:

My Garmin GPS III+ give GB OS grid refs. I don't know about the newer Garmin
models but I bet they still make models that do what you need.

Quoted message said:

In fact thinking about it, as most of us use a GPS in conjunction with a map
it would seem to be a basic function that all GPS units designed for walking
should be able to do. Isn't that why mine offers so many different options,
to cope with the map systems used in other countries? I count 104 on the
III+

I think most GPS units designed for walkers are fine. What you need to
avoid are the ones designed for car drivers, since for some reason the
manufacturers think car drivers are very stupid and scared off by TMI
so car driving units are generally pretty dumbed down.

--
Chris Malcolm [email hidden] DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]

Bernard Hill said:

In article <[email hidden]>, Peter Clinch
<[email hidden]> writes

Quoted message said:

Bernard Hill said:

I'll look out for that. But I was hoping to get under £50 since the
electronics obviously comes that sort of price.

eBay is probably your friend here.

Quoted message said:

From a quick trawl it's possible to get a basic unit (say, an eTrex or
Explorist) for sensible quantities of money.

Exactly. But finding out if they do OS GR is why I asked the question.

The Yellow Etrex does OS GR and you should be able to pick up one for
under 50 GBP but why not go for the blue Etrex Legend for under 80 GBP
I got mine for 77 GBP and it was a US grey import so no base maps but
hey you can download the mapsource for europe from bittorrent. 30 GBP
difference and you get maps sounds good to me. Why make thing
difficult for yourself.
--
Gwyn. [email hidden]
Remove rem if replying

On Wed, 1 Aug 2007 19:39:38 +0100, Roger <[email hidden]>

Quoted message said:

I think there would be too much loss of face for the OS ever to admit
that they had got it wrong these last 60 odd years.

Roger I think your missing the point. Who cares if it means the SW
corner of a square or a point as long as your intelligent enough to be
able to convert a 10 figure grid ref into a six figure one so you can
point the mountain rescue services where you are. I think its
intelligent of the OS to mean a square so you can just truncate and
not have to round your coordinates.
In any case a 10 figure grid ref is just too small to matter.
--
Gwyn. [email hidden]
Remove rem if replying