6 replies to this topic

[Martin Gamache]

I'm trying to find a piece of software that can reliably and accurately project both vector and raster datasets into an interrupted Mollweide projection without any artifacts. For reference this is the projection on p. 57 of USGS PP1453.

Manifold 8 lists this projection but can't seem to do it cleanly for either vectors (Natural earth 10M Oceans) or rasters (ETOPO1) I havent't wasted too much more time trying to get it to work with other datasets. Arcmap, Global Mapper and Geocart don't seem able to do it. I don't have a lot of experience with the various flavors of open source GIS out there but I'm hoping someone here does or has a suggestion as to what piece of code can get these global datasets cleanly projected.

I am considering the Goodes Homolosine Ocean implementation in Arcmap as an alternative but the Mollweide is more attractive given the width to height ration of the world once projected, i.e. it fills a large sheet better.

Thanks in advance,

MG

[Strebe]

I'm trying to find a piece of software that can reliably and accurately project both vector and raster datasets into an interrupted Mollweide projection without any artifacts. For reference this is the projection on p. 57 of USGS PP1453.

Manifold 8 lists this projection but can't seem to do it cleanly for either vectors (Natural earth 10M Oceans) or rasters (ETOPO1) I havent't wasted too much more time trying to get it to work with other datasets. Arcmap, Global Mapper and Geocart don't seem able to do it. I don't have a lot of experience with the various flavors of open source GIS out there but I'm hoping someone here does or has a suggestion as to what piece of code can get these global datasets cleanly projected.

I am considering the Goodes Homolosine Ocean implementation in Arcmap as an alternative but the Mollweide is more attractive given the width to height ration of the world once projected, i.e. it fills a large sheet better.

Actually, unless I misunderstand something about your requirements, what you describe is exactly what Geocart is made to do. Please contact me directly or (even better) open a topic on the Geocart forums.

Thanks & regards,
— daan Strebe
dstrebe.atsign.mapthematics.com

[Martin Gamache]

Hi Daan,

I agree it would seems Geocart should be ideally suited for this but it seemed that it wasn't available based on this list :

http://www.mapthemat...on=69#Mollweide

I did not see the interrupted Mollweide listed. If it's available wonderful, problem solved. I'll contact you off list to confirm.

Martin

I'm trying to find a piece of software that can reliably and accurately project both vector and raster datasets into an interrupted Mollweide projection without any artifacts. For reference this is the projection on p. 57 of USGS PP1453.

Manifold 8 lists this projection but can't seem to do it cleanly for either vectors (Natural earth 10M Oceans) or rasters (ETOPO1) I havent't wasted too much more time trying to get it to work with other datasets. Arcmap, Global Mapper and Geocart don't seem able to do it. I don't have a lot of experience with the various flavors of open source GIS out there but I'm hoping someone here does or has a suggestion as to what piece of code can get these global datasets cleanly projected.

I am considering the Goodes Homolosine Ocean implementation in Arcmap as an alternative but the Mollweide is more attractive given the width to height ration of the world once projected, i.e. it fills a large sheet better.

Actually, unless I misunderstand something about your requirements, what you describe is exactly what Geocart is made to do. Please contact me directly or (even better) open a topic on the Geocart forums.

Thanks & regards,
— daan Strebe
dstrebe.atsign.mapthematics.com

[Strebe]

I agree it would seems Geocart should be ideally suited for this but it seemed that it wasn't available based on this list :

http://www.mapthemat...on=69#Mollweide

I did not see the interrupted Mollweide listed. If it's available wonderful, problem solved. I'll contact you off list to confirm.

Martin

Hi, Martin. The reason the interrupted form of the projection is not listed specifically is because of Geocart’s generality: You can interrupt any projection many different ways as long as the projection itself will tolerate it. Hence it’s not practical to list the permutations exhaustively. Geocart can certainly do this particular projection, center, and interruption scheme.

Regards,
— daan Strebe

[frax]

For the record, it might be possible to do it in ArcGIS (and others) by projecting each lobe separately. It is quite a task to set up a script to do it, then you need to have the specs for each lobe as well. Before Interrupted Goode's Homolosine was available in ArcGIS there was a script that I used that did that.

[Dale A]

Manifold 8 lists this projection but can't seem to do it cleanly for either vectors (Natural earth 10M Oceans) or rasters (ETOPO1) I havent't wasted too much more time trying to get it to work with other datasets.

Did you post this question over on the Manifold forum? What do you mean by cannot do it cleanly? Could you post a screen grab?

<edit> Did you check the clip coordinates box as per the manual? That might explain the appearance of artifacts?
Cheers
D.

[Martin Gamache]

Manifold 8 lists this projection but can't seem to do it cleanly for either vectors (Natural earth 10M Oceans) or rasters (ETOPO1) I havent't wasted too much more time trying to get it to work with other datasets.

Did you post this question over on the Manifold forum? What do you mean by cannot do it cleanly? Could you post a screen grab?

<edit> Did you check the clip coordinates box as per the manual? That might explain the appearance of artifacts?
Cheers
D.

Vectors reprojected more cleanly with the clip coordinate box checked. Thanks for the tip. As you can see below the same can't be said for the raster data. Here is a screen grab of Etopo1 data reprojected with clip coordinates checked:

Attached Files

•  etopo1_moll_intrpt.JPG   42.01K   38 downloads