[GRASS-user] (kein Betreff)

GRASS GRASS user
1

hello everybody!
i have a little problem:
i want to set vertices on lines in a specified space (e.g. 2 km) in a lat long coordinate system.
i tried it with "v.to.points -vi .... dmax=0.03" and it works. the problem is, that in the equatorial zone the space between the new added points is about 1,7 km but up to the polzones the spacing is rather smaller and about 600 m!
can anybody help me with this problem? a want to have an equal space for all vertices on my polylines

michael
--
GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
Alle Infos und kostenlose Anmeldung: http://www.gmx.net/de/go/freemail

2

Lat-Long is not good to do distance measurements!

Why don't you reproject your lines in a "metric" projection system and
check the distances again.

On Thu, 2007-12-13 at 14:03 +0100, Michael Misun wrote:

hello everybody!
i have a little problem:
i want to set vertices on lines in a specified space (e.g. 2 km) in a lat long coordinate system.
i tried it with "v.to.points -vi .... dmax=0.03" and it works. the problem is, that in the equatorial zone the space between the new added points is about 1,7 km but up to the polzones the spacing is rather smaller and about 600 m!
can anybody help me with this problem? a want to have an equal space for all vertices on my polylines

michael

--
Nikos Alexandris
.
Department of Remote Sensing & Landscape Information Systems
Faculty of Forestry & Environmental Sciences, Albert-Ludwigs-University Freiburg
.
Tel. +49 (0) 761 203 3697 / Fax. +49 (0) 761 203 3701 / Skype: Nikos.Alexandris
.
Address: Tennenbacher str. 4, D-79106 Freiburg i. Br., Germany

3

Nikos Alexandris <nikos.alexandris@felis.uni-freiburg.de> writes:

>> hello everybody! i have a little problem: i want to set vertices on
>> lines in a specified space (e.g. 2 km) in a lat long coordinate
>> system. i tried it with "v.to.points -vi .... dmax=0.03" and it
>> works. the problem is, that in the equatorial zone the space between
>> the new added points is about 1,7 km but up to the polzones the
>> spacing is rather smaller and about 600 m! can anybody help me with
>> this problem? a want to have an equal space for all vertices on my
>> polylines

> Lat-Long is not good to do distance measurements! Why don't you
> reproject your lines in a "metric" projection system and check the
> distances again.

  Or, rather, why not to create the vector feature you need in a
  location which preserves the distance better, and `v.proj' it to
  the latlong location afterwards?

4

I'm curious about the statement that "Lat-Long is not good to do distance
measurements" Someone else made a similar observation in a different
conversation recently too. I'm not a geographer so I'm probably missing
something but doesn't lat long just give you a point on the surface of the
earth and if you have two of these don't you more or less automatically know
the distance between them?

Regards, Jerry

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org
[mailto:grass-user-bounces@lists.osgeo.org] On Behalf Of Nikos Alexandris
Sent: Thursday, December 13, 2007 10:55 AM
To: Michael Misun
Cc: grass-user@lists.osgeo.org
Subject: Re: [GRASS-user] (kein Betreff)

Lat-Long is not good to do distance measurements!

Why don't you reproject your lines in a "metric" projection system and
check the distances again.

On Thu, 2007-12-13 at 14:03 +0100, Michael Misun wrote:

hello everybody!
i have a little problem:
i want to set vertices on lines in a specified space (e.g. 2 km) in a lat

long coordinate system.

i tried it with "v.to.points -vi .... dmax=0.03" and it works. the problem

is, that in the equatorial zone the space between the new added points is
about 1,7 km but up to the polzones the spacing is rather smaller and about
600 m!

can anybody help me with this problem? a want to have an equal space for

all vertices on my polylines

michael

--
Nikos Alexandris
.
Department of Remote Sensing & Landscape Information Systems
Faculty of Forestry & Environmental Sciences, Albert-Ludwigs-University
Freiburg
.
Tel. +49 (0) 761 203 3697 / Fax. +49 (0) 761 203 3701 / Skype:
Nikos.Alexandris
.
Address: Tennenbacher str. 4, D-79106 Freiburg i. Br., Germany

5

The width of a degree of longitude varies by latitude; meridians of longitude converge to
a single point at the poles. Latitude varies as well, although considerably less so, due to
the rotation of the earth slightly squashing the poles, and bulging the Equator.

~ Eric.

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org on behalf of Gerald Nelson
Sent: Thu 12/13/2007 3:06 PM
To: 'Nikos Alexandris'; 'Michael Misun'
Cc: grass-user@lists.osgeo.org
Subject: RE: [GRASS-user] (kein Betreff)

I'm curious about the statement that "Lat-Long is not good to do distance
measurements" Someone else made a similar observation in a different
conversation recently too. I'm not a geographer so I'm probably missing
something but doesn't lat long just give you a point on the surface of the
earth and if you have two of these don't you more or less automatically know
the distance between them?

Regards, Jerry

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org
[mailto:grass-user-bounces@lists.osgeo.org] On Behalf Of Nikos Alexandris
Sent: Thursday, December 13, 2007 10:55 AM
To: Michael Misun
Cc: grass-user@lists.osgeo.org
Subject: Re: [GRASS-user] (kein Betreff)

Lat-Long is not good to do distance measurements!

Why don't you reproject your lines in a "metric" projection system and
check the distances again.

On Thu, 2007-12-13 at 14:03 +0100, Michael Misun wrote:

hello everybody!
i have a little problem:
i want to set vertices on lines in a specified space (e.g. 2 km) in a lat

long coordinate system.

i tried it with "v.to.points -vi .... dmax=0.03" and it works. the problem

is, that in the equatorial zone the space between the new added points is
about 1,7 km but up to the polzones the spacing is rather smaller and about
600 m!

can anybody help me with this problem? a want to have an equal space for

all vertices on my polylines

michael

--
Nikos Alexandris
.
Department of Remote Sensing & Landscape Information Systems
Faculty of Forestry & Environmental Sciences, Albert-Ludwigs-University
Freiburg
.
Tel. +49 (0) 761 203 3697 / Fax. +49 (0) 761 203 3701 / Skype:
Nikos.Alexandris
.
Address: Tennenbacher str. 4, D-79106 Freiburg i. Br., Germany

_______________________________________________
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

6

Just another piece of information (although don't remember where I read
it): In general every 0.0001° of latitude equals ~ 11m. Longitude I
think is the tough one... :wink:

On Thu, 2007-12-13 at 14:12 -0500, Patton, Eric wrote:

The width of a degree of longitude varies by latitude; meridians of longitude converge to
a single point at the poles. Latitude varies as well, although considerably less so, due to
the rotation of the earth slightly squashing the poles, and bulging the Equator.

~ Eric.

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org on behalf of Gerald Nelson
Sent: Thu 12/13/2007 3:06 PM
To: 'Nikos Alexandris'; 'Michael Misun'
Cc: grass-user@lists.osgeo.org
Subject: RE: [GRASS-user] (kein Betreff)

I'm curious about the statement that "Lat-Long is not good to do distance
measurements" Someone else made a similar observation in a different
conversation recently too. I'm not a geographer so I'm probably missing
something but doesn't lat long just give you a point on the surface of the
earth and if you have two of these don't you more or less automatically know
the distance between them?

Regards, Jerry

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org
[mailto:grass-user-bounces@lists.osgeo.org] On Behalf Of Nikos Alexandris
Sent: Thursday, December 13, 2007 10:55 AM
To: Michael Misun
Cc: grass-user@lists.osgeo.org
Subject: Re: [GRASS-user] (kein Betreff)

Lat-Long is not good to do distance measurements!

Why don't you reproject your lines in a "metric" projection system and
check the distances again.

On Thu, 2007-12-13 at 14:03 +0100, Michael Misun wrote:
> hello everybody!
> i have a little problem:
> i want to set vertices on lines in a specified space (e.g. 2 km) in a lat
long coordinate system.
> i tried it with "v.to.points -vi .... dmax=0.03" and it works. the problem
is, that in the equatorial zone the space between the new added points is
about 1,7 km but up to the polzones the spacing is rather smaller and about
600 m!
> can anybody help me with this problem? a want to have an equal space for
all vertices on my polylines
>
> michael

--
Nikos Alexandris
.
Department of Remote Sensing & Landscape Information Systems
Faculty of Forestry & Environmental Sciences, Albert-Ludwigs-University Freiburg
.
Tel. +49 (0) 761 203 3697 / Fax. +49 (0) 761 203 3701 / Skype: Nikos.Alexandris
.
Address: Tennenbacher str. 4, D-79106 Freiburg i. Br., Germany

7

Although I'm not a geographer, having looked at a globe I do know about
longitude meridians running through the poles. :wink:

If you only had two longitude values you wouldn't even be able to calculate
EW distance because of this effect. But if you know both the value for
latitude and longitude at each of say two points, you know how close each is
to the equator and so should be able to correct for the effect, doing some
kind of great circle route type calculation. Or...

Jerry

-----Original Message-----
From: Nikos Alexandris [mailto:nikos.alexandris@felis.uni-freiburg.de]
Sent: Thursday, December 13, 2007 2:27 PM
To: Patton, Eric
Cc: Gerald Nelson; Michael Misun; grass-user@lists.osgeo.org
Subject: RE: [GRASS-user] (kein Betreff)

Just another piece of information (although don't remember where I read
it): In general every 0.0001° of latitude equals ~ 11m. Longitude I
think is the tough one... :wink:

On Thu, 2007-12-13 at 14:12 -0500, Patton, Eric wrote:

The width of a degree of longitude varies by latitude; meridians of

longitude converge to

a single point at the poles. Latitude varies as well, although

considerably less so, due to

the rotation of the earth slightly squashing the poles, and bulging the

Equator.

~ Eric.

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org on behalf of Gerald Nelson
Sent: Thu 12/13/2007 3:06 PM
To: 'Nikos Alexandris'; 'Michael Misun'
Cc: grass-user@lists.osgeo.org
Subject: RE: [GRASS-user] (kein Betreff)

I'm curious about the statement that "Lat-Long is not good to do distance
measurements" Someone else made a similar observation in a different
conversation recently too. I'm not a geographer so I'm probably missing
something but doesn't lat long just give you a point on the surface of the
earth and if you have two of these don't you more or less automatically

know

the distance between them?

Regards, Jerry

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org
[mailto:grass-user-bounces@lists.osgeo.org] On Behalf Of Nikos Alexandris
Sent: Thursday, December 13, 2007 10:55 AM
To: Michael Misun
Cc: grass-user@lists.osgeo.org
Subject: Re: [GRASS-user] (kein Betreff)

Lat-Long is not good to do distance measurements!

Why don't you reproject your lines in a "metric" projection system and
check the distances again.

On Thu, 2007-12-13 at 14:03 +0100, Michael Misun wrote:
> hello everybody!
> i have a little problem:
> i want to set vertices on lines in a specified space (e.g. 2 km) in a

lat

long coordinate system.
> i tried it with "v.to.points -vi .... dmax=0.03" and it works. the

problem

is, that in the equatorial zone the space between the new added points is
about 1,7 km but up to the polzones the spacing is rather smaller and

about

600 m!
> can anybody help me with this problem? a want to have an equal space for
all vertices on my polylines
>
> michael

--
Nikos Alexandris
.
Department of Remote Sensing & Landscape Information Systems
Faculty of Forestry & Environmental Sciences, Albert-Ludwigs-University
Freiburg
.
Tel. +49 (0) 761 203 3697 / Fax. +49 (0) 761 203 3701 / Skype:
Nikos.Alexandris
.
Address: Tennenbacher str. 4, D-79106 Freiburg i. Br., Germany

8

Nikos Alexandris wrote:

Just another piece of information (although don't remember where I
read it): In general every 0.0001° of latitude equals ~ 11m.

1852m in a nautical mile. 60 nautical miles in a degree latitude.
so 1 degree latitude = 111.12km

Longitude I think is the tough one... :wink:

dist*cos(lat)

Jerry:

> I'm curious about the statement that "Lat-Long is not good to do
> distance measurements" Someone else made a similar observation in
> a different conversation recently too. I'm not a geographer so
> I'm probably missing something but doesn't lat long just give you
> a point on the surface of the earth and if you have two of these
> don't you more or less automatically know the distance between
> them?

For distance measurements in Lat/Lon GRASS uses the
G_geodesic_distance() function found in lib/gis/geodist.c
to calculate the great-circle distance.

This is widely used by the raster modules and libs (which are generally
very mature [sometimes to the point of geriatrics]) as well as the
display modules (d.measure, d.what.vect, ...)

The vector and DB modules are brand new for GRASS 6 and so many/most
vector modules only consider Cartesian space when doing there
calculations. These will be slowly updated, but it takes time.

So modules like r.buffer will make correct pear shaped circles in a lat
lon location, but v.buffer will just make an incorrect lat=lon even at
high latitudes circle.

> Lat-Long is not good to do distance measurements!
>
> Why don't you reproject your lines in a "metric" projection system
> and check the distances again.

If working with vector modules this is probably your best bet, work in
a projected location then reproject the results back.

for v.to.points, dmax= dist should be made to work in meters for
lat/lon locations and calculate with G_geodesic_distance() as it makes
its way between verticies. (e.g. the d.what.vect Tcl popup window does
this for line length, AFAICT) Feel free to file wish/bug reports as
needed.

Hamish

      ____________________________________________________________________________________
Looking for last minute shopping deals?
Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping

9

hi, first of all thanks a lot for your answers.
that is true wat you say eric, but i hope there would be a solution for my problem. i know, that in grass4 was a module called v.plant which fills in points on a line in a specified space but i doesn't work here in my grass6.2.3

above all i work in a location for mars not for the earth

-------- Original-Nachricht --------

Datum: Thu, 13 Dec 2007 14:12:13 -0500
Von: "Patton, Eric" <epatton@nrcan.gc.ca>
An: "Gerald Nelson" <gnelson@uiuc.edu>, "Nikos Alexandris" <nikos.alexandris@felis.uni-freiburg.de>, "Michael Misun" <doppelM@gmx.de>
CC: grass-user@lists.osgeo.org
Betreff: RE: [GRASS-user] (kein Betreff)

The width of a degree of longitude varies by latitude; meridians of
longitude converge to
a single point at the poles. Latitude varies as well, although
considerably less so, due to
the rotation of the earth slightly squashing the poles, and bulging the
Equator.

~ Eric.

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org on behalf of Gerald Nelson
Sent: Thu 12/13/2007 3:06 PM
To: 'Nikos Alexandris'; 'Michael Misun'
Cc: grass-user@lists.osgeo.org
Subject: RE: [GRASS-user] (kein Betreff)

I'm curious about the statement that "Lat-Long is not good to do distance
measurements" Someone else made a similar observation in a different
conversation recently too. I'm not a geographer so I'm probably missing
something but doesn't lat long just give you a point on the surface of the
earth and if you have two of these don't you more or less automatically
know
the distance between them?

Regards, Jerry

-----Original Message-----
From: grass-user-bounces@lists.osgeo.org
[mailto:grass-user-bounces@lists.osgeo.org] On Behalf Of Nikos Alexandris
Sent: Thursday, December 13, 2007 10:55 AM
To: Michael Misun
Cc: grass-user@lists.osgeo.org
Subject: Re: [GRASS-user] (kein Betreff)

Lat-Long is not good to do distance measurements!

Why don't you reproject your lines in a "metric" projection system and
check the distances again.

On Thu, 2007-12-13 at 14:03 +0100, Michael Misun wrote:
> hello everybody!
> i have a little problem:
> i want to set vertices on lines in a specified space (e.g. 2 km) in a
lat
long coordinate system.
> i tried it with "v.to.points -vi .... dmax=0.03" and it works. the
problem
is, that in the equatorial zone the space between the new added points is
about 1,7 km but up to the polzones the spacing is rather smaller and
about
600 m!
> can anybody help me with this problem? a want to have an equal space for
all vertices on my polylines
>
> michael
--
Nikos Alexandris
.
Department of Remote Sensing & Landscape Information Systems
Faculty of Forestry & Environmental Sciences, Albert-Ludwigs-University
Freiburg
.
Tel. +49 (0) 761 203 3697 / Fax. +49 (0) 761 203 3701 / Skype:
Nikos.Alexandris
.
Address: Tennenbacher str. 4, D-79106 Freiburg i. Br., Germany

_______________________________________________
grass-user mailing list
grass-user@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/grass-user

--
Psssst! Schon vom neuen GMX MultiMessenger gehört?
Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger

10

On 14/12/07 10:01, Michael Misun wrote:

hi, first of all thanks a lot for your answers. that is true wat you
say eric, but i hope there would be a solution for my problem. i
know, that in grass4 was a module called v.plant which fills in
points on a line in a specified space but i doesn't work here in my
grass6.2.3

above all i work in a location for mars not for the earth

Maybe v.lrs could be of help ?

See http://grass.itc.it/grass63/manuals/html63_user/lrs.html

Moritz

11

Michael Misun wrote:

i know, that in grass4 was a module called v.plant which fills in
points on a line in a specified space but i doesn't work here in my
grass6.2.3

it seems available in GRASS 5.4 too:
  http://grass.ibiblio.org/gdp/html_grass5/html/v.plant.html

From that help page:

"Design of v.plant

v.plant has been implemented as a Bourne shell script. The basic
procedure is as follows:

1. Write out the existing map in the dig_ascii format;

2. Use an included awk script to process each vector segment in turn,
moving from point to point along the segment. If a span between two
adjacent points is greater than the specified threshold (in map units)
extra points are inserted collinearly.

3. The modified dig_ascii file is re-imported, overwriting the original
file."

the main difference between GRASS 5 & 6's ascii output format is the
order of the columns (x y in GRASS 6, not y x as in GRASS 5).

so with that adjustment (and probably a few other minor tweaks) the
script could probably be used in GRASS 6. GRASS 4/5 vector modules
using the C api need considerable rewrite to work with GRASS 6 as the
GRASS 6 vector engine is a all new.

But you can use v.split, v.segment, or v.lrs.* in GRASS 6 to do the
same thing as v.plant.

for GRASS 5/6 porting list see:

http://grass.gdf-hannover.de/wiki/GRASS_Module_Porting_List#Vector_modules

above all i work in a location for mars not for the earth

That should not be a problem. If your lat/lon projection definition
defines 'a' and 'es' parameters (or 'a' + defined as a sphere), then
G_geodesic_distance() will use those values in the distance
calculation. It only defaults to Earth values if those are empty.

see G_get_ellipsoid_parameters() in lib/gis/get_ellipse.c which is
called by G_begin_distance_calculations() (lib/gis/distance.c).

G_distance() then calls G_geodesic_distance() if the projection is
lat/lon. [see also lib/gis/geodist.c and lib/gis/geodesic.c]

e.g. d.measure with an xmon should return the correct great circle
distance. (I have no idea about the distance measurement tool in the
GUIs)

Again, be careful with new vector modules. Some may be simplisticly
doing their own sqrt(a^2 + b^2) calucations instead of using
G_distance(). Luckily the results should be so badly wrong for lat/lon
locations that you will quickly notice the error. Post any requests for
specific module updates to the wish/bug tracker.

Hamish

      ____________________________________________________________________________________
Looking for last minute shopping deals?
Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping

12

hi everybody!

i have a question concerning display rastermaps.

the facts:
i have a few tif files and my task is making a mosaic of these files. i imported the files with r.in.gdal and it works. the first problem is, that i can only display one rastermap in a monitor. when i use i.image.mosaic there will be an output mosaic of the first 2 rasters but i also can't display it in a monitor.

have anyone an idea or solution about this problem?

thanks michael

13

On 10/01/08 11:15, doppelM@gmx.de wrote:

hi everybody!

i have a question concerning display rastermaps.

the facts: i have a few tif files and my task is making a mosaic of
these files. i imported the files with r.in.gdal and it works. the
first problem is, that i can only display one rastermap in a monitor.

You can display as many raster maps as you want in one monitor. Just zoom out sufficiently, or set the region with 'g.region rast=map1,map2,map3,...' and then 'zoom to current region'.

when i use i.image.mosaic there will be an output mosaic of the first
2 rasters but i also can't display it in a monitor.
have anyone an idea or solution about this problem?

r.patch

Moritz