Main Page | Report Page

 

  Computers Forum Index » Computer Languages (IDL-PVWAVE) » Divide the world into hexagons...

Author Message
Ed Hyer...
Posted: Mon Oct 18, 2010 3:20 pm
 
I recall reading a paper some time back where the rectangular lat/lon
grid was replaced with a hexagonal grid, whose polygons did not change
size with distance from the poles. Does anyone know where there might
be some code to create these grids? Not thinking specifically of
graphics, although any routine that could generate the graphical part
could also do the part I'm interested in.

Any leads welcome,

--Edward H.
 
Ed Hyer...
Posted: Mon Oct 18, 2010 5:33 pm
 
Thanks Ken!

Looks like there's some relevant code out there in MATLAB:

http://people.sc.fsu.edu/~jburkardt/m_src/sphere_grid/sphere_grid.html

But still a lot of work to turn that into even a rudimentary Earth
coordinate system.

What I am actually hoping to get out of this is a cheap way to bin
observations into equal-size areas across the globe. This does not
require a full geoid or other complications.

--Edward H.

On Oct 18, 9:02 am, "Kenneth P. Bowman" <k-bow... at (no spam) null.edu> wrote:
Quote:
In article
3493fa57-c263-4cdb-a375-8f06146cd... at (no spam) 42g2000prt.googlegroups.com>,
 Ed Hyer <ejh... at (no spam) gmail.com> wrote:

I recall reading a paper some time back where the rectangular lat/lon
grid was replaced with a hexagonal grid, whose polygons did not change
size with distance from the poles. Does anyone know where there might
be some code to create these grids? Not thinking specifically of
graphics, although any routine that could generate the graphical part
could also do the part I'm interested in.

Any leads welcome,

--Edward H.

There are a number of global meteorological models that use a
grid based on an icosahedron, which has 20 faces, each face is an
identical equilateral triangle.

Try googling "icosahedral atmospheric model", e.g.,

   http://www.wrfportal.org/CIRA-Magazine-GIMTool.pdf

Note that the resulting grid is not perfectly uniform.  Most cells
are hexagons.  Some cells near the vertices of the original
icosahedron are pentagons.  Also, it is not possible to make the
hexagons complete regular, but they are close.

Ken Bowman

 
Kenneth P. Bowman...
Posted: Mon Oct 18, 2010 8:02 pm
 
In article
<3493fa57-c263-4cdb-a375-8f06146cd67c at (no spam) 42g2000prt.googlegroups.com>,
Ed Hyer <ejhyer at (no spam) gmail.com> wrote:

Quote:
I recall reading a paper some time back where the rectangular lat/lon
grid was replaced with a hexagonal grid, whose polygons did not change
size with distance from the poles. Does anyone know where there might
be some code to create these grids? Not thinking specifically of
graphics, although any routine that could generate the graphical part
could also do the part I'm interested in.

Any leads welcome,

--Edward H.

There are a number of global meteorological models that use a
grid based on an icosahedron, which has 20 faces, each face is an
identical equilateral triangle.

Try googling "icosahedral atmospheric model", e.g.,

http://www.wrfportal.org/CIRA-Magazine-GIMTool.pdf

Note that the resulting grid is not perfectly uniform. Most cells
are hexagons. Some cells near the vertices of the original
icosahedron are pentagons. Also, it is not possible to make the
hexagons complete regular, but they are close.

Ken Bowman
 
jkeller...
Posted: Tue Oct 19, 2010 8:18 am
 
On Oct 18, 7:33 pm, Ed Hyer <ejh... at (no spam) gmail.com> wrote:
Quote:
Thanks Ken!

Looks like there's some relevant code out there in MATLAB:

http://people.sc.fsu.edu/~jburkardt/m_src/sphere_grid/sphere_grid.html

But still a lot of work to turn that into even a rudimentary Earth
coordinate system.

What I am actually hoping to get out of this is a cheap way to bin
observations into equal-size areas across the globe. This does not
require a full geoid or other complications.

--Edward H.


A while back I wrote IDL code to plot such a grid. If you are
interested I could send you a copy.

Sincerely,
Jan
 
 
Page 1 of 1    
All times are GMT
The time now is Sat Aug 30, 2014 2:39 pm