Bookmark this page:
- Geotiff-with-UTM-from-more-than-one-zone
- 03-17-2008
  Re: Geotiff with UTM from more than one zone?
| Anton Betzler | 03-17-2008 |
  Re: Geotiff with UTM from more than one zone?
| Luca Morandini | 03-17-2008 |
 Re: Geotiff with UTM from more than one zone?
| Uffe Kousgaard | 03-17-2008 |
If you were Registered and logged in, you could reply and use other advanced thread options
> > On Mon, 17 Mar 2008 20:15:23 +0100, Luca Morandini
> > >Martin wrote:
> > >> Hi,
> > >> is it possible to create a world file with UTM, that covers more than=
> > >> Hi,
> > >> is it possible to create a world file with UTM, that covers more than=
> > >> one zone? In my example I would like to georeferenciate raster maps
> > >> from whole Germany, covering the UTM zones 31 to 33 (well, the map
> > >> images do...). So the upper left corner would be in the 31N zone.
> > >You can cover as many zones as you like, though distortion will reach
> > >levels not admitted in the UTM system (the further east or west you go
> > >outside the zone's extent, the higher the distortion).
> > >Regards,
> > >--------------------
> > > =A0 =A0Luca Morandini
> > >www.lucamorandini.it
> > >--------------------
> > >> from whole Germany, covering the UTM zones 31 to 33 (well, the map
> > >> images do...). So the upper left corner would be in the 31N zone.
> > >You can cover as many zones as you like, though distortion will reach
> > >levels not admitted in the UTM system (the further east or west you go
> > >outside the zone's extent, the higher the distortion).
> > >Regards,
> > >--------------------
> > > =A0 =A0Luca Morandini
> > >www.lucamorandini.it
> > >--------------------
> > Well, actually no. You can certainly create a map covering several UTM
> > zones, but it won't be in UTM. UTM has the following properties which,
> > I hope, will make it clear why what you ask is actually nonsensical.
> > 1) UTM is a series of Transverse Mercator projections on central
> > meridians at 6 degree intervals. UTM zones are 6 degrees wide, neither
> > more nor less, and "extending" a zone is simply not UTM.
> > 2) Each zone has its own coordinate system. The same coordinates can
> > and do appear in each zone. If you simply plot UTM coordinates from
> > several zones, they wil plot on top of each other.
> > 3) The mathematics of the Transverse Mercator projection underlying
> > UTM is such that it is difficult to provide an accurate transformation
> > much beyond 10 degrees from the central meridian.
> > The point is that UTM isn't a single projection, and can't be used as
> > such. The way to do what you want is to decide a sensible projection
> > for such a large area (and that won't be Transverse Mercator because
> > of 3), project your data to it and the create your geotiff.
> > UTM is a spectacularly nasty kludge created by the military; it is
> > pretty much useless for serious small-scale (e.g. national scale)
> > mapping. It was only ever meant for large-scale mapping, where it
> > provides the illusion of ease of use and consistency (it is actually
> > neither).
> > Paul
> > zones, but it won't be in UTM. UTM has the following properties which,
> > I hope, will make it clear why what you ask is actually nonsensical.
> > 1) UTM is a series of Transverse Mercator projections on central
> > meridians at 6 degree intervals. UTM zones are 6 degrees wide, neither
> > more nor less, and "extending" a zone is simply not UTM.
> > 2) Each zone has its own coordinate system. The same coordinates can
> > and do appear in each zone. If you simply plot UTM coordinates from
> > several zones, they wil plot on top of each other.
> > 3) The mathematics of the Transverse Mercator projection underlying
> > UTM is such that it is difficult to provide an accurate transformation
> > much beyond 10 degrees from the central meridian.
> > The point is that UTM isn't a single projection, and can't be used as
> > such. The way to do what you want is to decide a sensible projection
> > for such a large area (and that won't be Transverse Mercator because
> > of 3), project your data to it and the create your geotiff.
> > UTM is a spectacularly nasty kludge created by the military; it is
> > pretty much useless for serious small-scale (e.g. national scale)
> > mapping. It was only ever meant for large-scale mapping, where it
> > provides the illusion of ease of use and consistency (it is actually
> > neither).
> > Paul
> Hi Paul,
> thank you and the others for your answers.
> I now decided to undo the German reunification and let the eastern
> part have the zone of its own (UTM 33N) in
> its own map set. Not very comfortable, but as I think the only way
> (for me) to avoid any confusion as I also have to make
> transformations from Gau=DF-Kr=FCger and geographic coordinates.
> thank you and the others for your answers.
> I now decided to undo the German reunification and let the eastern
> part have the zone of its own (UTM 33N) in
> its own map set. Not very comfortable, but as I think the only way
> (for me) to avoid any confusion as I also have to make
> transformations from Gau=DF-Kr=FCger and geographic coordinates.
One option would be to add a shift in the eastings of the 32N and 33N
parts so that these abut in series. The coordinates wont be true UTM
but the projection for each part will be as good. The map will of
course be distorted and perhaps unfamiliar, but it is a map and that
is what maps are. If one were to use 32N as base then 31N would
require negative coordinates which is not to be recommended.
> > > On Mon, 17 Mar 2008 20:15:23 +0100, Luca Morandini
> > > >Martin wrote:
> > > >> Hi,
> > > >> is it possible to create a world file with UTM, that covers more th=
an
> > > >Martin wrote:
> > > >> Hi,
> > > >> is it possible to create a world file with UTM, that covers more th=
> > > >> one zone? In my example I would like to georeferenciate raster maps=
> > > >> from whole Germany, covering the UTM zones 31 to 33 (well, the map
> > > >> images do...). So the upper left corner would be in the 31N zone.
> > > >You can cover as many zones as you like, though distortion will reach=
> > > >> images do...). So the upper left corner would be in the 31N zone.
> > > >You can cover as many zones as you like, though distortion will reach=
> > > >levels not admitted in the UTM system (the further east or west you g=
o
> > > >outside the zone's extent, the higher the distortion).
> > > >Regards,
> > > >--------------------
> > > > =A0 =A0Luca Morandini
> > > >www.lucamorandini.it
> > > >--------------------
> > > Well, actually no. You can certainly create a map covering several UTM=
> > > >Regards,
> > > >--------------------
> > > > =A0 =A0Luca Morandini
> > > >www.lucamorandini.it
> > > >--------------------
> > > Well, actually no. You can certainly create a map covering several UTM=
> > > zones, but it won't be in UTM. UTM has the following properties which,=
> > > I hope, will make it clear why what you ask is actually nonsensical.
> > > 1) UTM is a series of Transverse Mercator projections on central
> > > meridians at 6 degree intervals. UTM zones are 6 degrees wide, neither=
> > > 1) UTM is a series of Transverse Mercator projections on central
> > > meridians at 6 degree intervals. UTM zones are 6 degrees wide, neither=
> > > more nor less, and "extending" a zone is simply not UTM.
> > > 2) Each zone has its own coordinate system. The same coordinates can
> > > and do appear in each zone. If you simply plot UTM coordinates from
> > > several zones, they wil plot on top of each other.
> > > 3) The mathematics of the Transverse Mercator projection underlying
> > > UTM is such that it is difficult to provide an accurate transformation=
> > > 2) Each zone has its own coordinate system. The same coordinates can
> > > and do appear in each zone. If you simply plot UTM coordinates from
> > > several zones, they wil plot on top of each other.
> > > 3) The mathematics of the Transverse Mercator projection underlying
> > > UTM is such that it is difficult to provide an accurate transformation=
> > > much beyond 10 degrees from the central meridian.
> > > The point is that UTM isn't a single projection, and can't be used as
> > > such. The way to do what you want is to decide a sensible projection
> > > for such a large area (and that won't be Transverse Mercator because
> > > of 3), project your data to it and the create your geotiff.
> > > UTM is a spectacularly nasty kludge created by the military; it is
> > > pretty much useless for serious small-scale (e.g. national scale)
> > > mapping. It was only ever meant for large-scale mapping, where it
> > > provides the illusion of ease of use and consistency (it is actually
> > > neither).
> > > Paul
> > > The point is that UTM isn't a single projection, and can't be used as
> > > such. The way to do what you want is to decide a sensible projection
> > > for such a large area (and that won't be Transverse Mercator because
> > > of 3), project your data to it and the create your geotiff.
> > > UTM is a spectacularly nasty kludge created by the military; it is
> > > pretty much useless for serious small-scale (e.g. national scale)
> > > mapping. It was only ever meant for large-scale mapping, where it
> > > provides the illusion of ease of use and consistency (it is actually
> > > neither).
> > > Paul
> > Hi Paul,
> > thank you and the others for your answers.
> > I now decided to undo the German reunification and let the eastern
> > part have the zone of its own (UTM 33N) in
> > its own map set. Not very comfortable, but as I think the only way
> > (for me) to avoid any confusion as I also have to make
> > transformations from Gau=DF-Kr=FCger and geographic coordinates.
> > thank you and the others for your answers.
> > I now decided to undo the German reunification and let the eastern
> > part have the zone of its own (UTM 33N) in
> > its own map set. Not very comfortable, but as I think the only way
> > (for me) to avoid any confusion as I also have to make
> > transformations from Gau=DF-Kr=FCger and geographic coordinates.
> One option would be to add a shift in the eastings of the 32N and 33N
> parts so that these abut in series. The coordinates wont be true UTM
> but the projection for each part will be as good. The map will of
> course be distorted and perhaps unfamiliar, but it is a map and that
> is what maps are. If one were to use 32N as base then 31N would
> require negative coordinates which is not to be recommended.- Hide quoted =
text -
> parts so that these abut in series. The coordinates wont be true UTM
> but the projection for each part will be as good. The map will of
> course be distorted and perhaps unfamiliar, but it is a map and that
> is what maps are. If one were to use 32N as base then 31N would
> require negative coordinates which is not to be recommended.- Hide quoted =
> - Show quoted text -
I have done a similar task with the 3 UTM zones that cover Texas (13,
14, 15) and 1 meter NAIP data. I created .DBF image catalogs in
ArcGIS for the images in each UTM zone and then created a layer file
(.LYR) that references these 3 image catalogs. I used this script as
the image catalog tools in ArcCatalog do not work well (http://
arcscripts.esri.com/details.asp?dbid=3D13129). I hope this helps.
On Mon, 17 Mar 2008 20:38:32 GMT, Paul Cooper
Paul,
So what you are saying is that distortion will be "acceptable" over an
area of 20 degrees wide. So if he wants to cover 3 UTM zones with
some sort of UTM-like projection, that would be just fine - it's only
18 degrees wide.
In most applications I have seen where people ask this sort of
question, they want to do gps mapping with receivers that are not
accurate to better than 5-10 meters anyway, so what's the point of
being fussy about a few more parts per million at the far edges?
The OP should specify his purpose, and state whether he prefers his
accuracy to be on the flat map because he is producing paper, or on
the curve, because he is producing a gps map, and/or working from gps
track/point data.
Perhaps he should retain use of lat/lon for all data in/out of his
application, and project to his pseudo-utm 3-zone thingee for any
"flat" displays - on screen or on paper. This would of course mean he
would have to "de-project" anything he draws on-screen before
outputting to any file.
As long as he is not using a total station, and trying to set bridge
bearings or even property bars and road alignments at the far corners
of his area based on the one central meridian, his inaccuracies
shouldn't ever even get noticed.
- Tom
>3) The mathematics of the Transverse Mercator projection underlying
>UTM is such that it is difficult to provide an accurate transformation
>much beyond 10 degrees from the central meridian.
>The point is that UTM isn't a single projection, and can't be used as
>such. The way to do what you want is to decide a sensible projection
>for such a large area (and that won't be Transverse Mercator because
>of 3), project your data to it and the create your geotiff.
>UTM is a spectacularly nasty kludge created by the military; it is
>pretty much useless for serious small-scale (e.g. national scale)
>mapping. It was only ever meant for large-scale mapping, where it
>provides the illusion of ease of use and consistency (it is actually
>neither).
>UTM is such that it is difficult to provide an accurate transformation
>much beyond 10 degrees from the central meridian.
>The point is that UTM isn't a single projection, and can't be used as
>such. The way to do what you want is to decide a sensible projection
>for such a large area (and that won't be Transverse Mercator because
>of 3), project your data to it and the create your geotiff.
>UTM is a spectacularly nasty kludge created by the military; it is
>pretty much useless for serious small-scale (e.g. national scale)
>mapping. It was only ever meant for large-scale mapping, where it
>provides the illusion of ease of use and consistency (it is actually
>neither).
Paul,
So what you are saying is that distortion will be "acceptable" over an
area of 20 degrees wide. So if he wants to cover 3 UTM zones with
some sort of UTM-like projection, that would be just fine - it's only
18 degrees wide.
In most applications I have seen where people ask this sort of
question, they want to do gps mapping with receivers that are not
accurate to better than 5-10 meters anyway, so what's the point of
being fussy about a few more parts per million at the far edges?
The OP should specify his purpose, and state whether he prefers his
accuracy to be on the flat map because he is producing paper, or on
the curve, because he is producing a gps map, and/or working from gps
track/point data.
Perhaps he should retain use of lat/lon for all data in/out of his
application, and project to his pseudo-utm 3-zone thingee for any
"flat" displays - on screen or on paper. This would of course mean he
would have to "de-project" anything he draws on-screen before
outputting to any file.
As long as he is not using a total station, and trying to set bridge
bearings or even property bars and road alignments at the far corners
of his area based on the one central meridian, his inaccuracies
shouldn't ever even get noticed.
- Tom
news:03e9329a-c551-4722-8243-
Yes, but you will have to decide for using a single zone for the whole
country, most likely zone 32 then. Of course distortion will be bigger to
the east and west.
> is it possible to create a world file with UTM, that covers more than
> one zone? In my example I would like to georeferenciate raster maps
> from whole Germany, covering the UTM zones 31 to 33 (well, the map
> images do...). So the upper left corner would be in the 31N zone.
> one zone? In my example I would like to georeferenciate raster maps
> from whole Germany, covering the UTM zones 31 to 33 (well, the map
> images do...). So the upper left corner would be in the 31N zone.
Yes, but you will have to decide for using a single zone for the whole
country, most likely zone 32 then. Of course distortion will be bigger to
the east and west.
- More than 12 satellites
- Satellite Navigation
- 2005-11-16
- Air Travel With Roadmate 800
- Magellan GPS
- 2006-07-19
