| View previous topic :: View next topic |
| Author |
Message |
Adakite
Ra (Power Member)
Joined: 29 Mar 2005
Posts: 98
Location: IPGP, Paris
|
 Posted: Thu Jan 25, 2007 4:51 pm Post subject: MOLA 128 MEGDR from pds (SOLVED) |
|
|
I tried to import Mola 128 MEGDR from pds into Isis, and I get an error : because pds file are binary and ascii. Is there any solution ?
thanks. |
|
| Back to top » |
|
 |
janderson
Isis Support Team
Joined: 27 Feb 2004
Posts: 919
Location: USGS Astrogeology Research Program, Flagstaff, AZ
|
| Posted: Fri Jan 26, 2007 9:17 am Post subject: |
|
|
Can you post where we can find the file you are using?
Thanks, Jeff |
|
| Back to top » |
|
|
Adakite
Ra (Power Member)
Joined: 29 Mar 2005
Posts: 98
Location: IPGP, Paris
|
| Posted: Sun Jan 28, 2007 10:38 am Post subject: |
|
|
Sure,
fellowing, informations related to PDS Mola data:
http://pds-geosciences.wustl.edu/missions/mgs/mola.html
and then I get data from Mission Experiment Gridded Data Records (MEGDRs): topographic maps (128px/deg):
http://pds-geosciences.wustl.edu/geodata/mgs-m-mola-5-megdr-l3-v1/mgsl_300x/meg128/
I used the megt00n270hb.img file.
I make a
pds2isis from=megt00n270hb.img to=mola_128.cub
and I get:
**I/O ERROR** Unable to read PVL file [/Volumes/VMX/themis/megt00n270hb.img]
**PVL ERROR** Invalid PVL format
**PARSE ERROR** ASCII data expected but found unprintable (binary) data
(NB. I am using a MacIntel (Tiger) platform with Isis 3.1.5)
Thanks for help. |
|
| Back to top » |
|
|
thare
GIS Support Team
Joined: 22 Mar 2004
Posts: 657
Location: USGS Astrogeology Research Program, Flagstaff, AZ
|
| Posted: Mon Jan 29, 2007 8:15 am Post subject: |
|
|
The MOLA DEMs are actually raw binary files (*.img) with a detached PDS header (*.lbl) which points to the *.img. Download the *.lbl files, place them in the same directory as the *.img files. Now try pointing to the label files for importing.
let us know if it works.
Trent |
|
| Back to top » |
|
|
thare
GIS Support Team
Joined: 22 Mar 2004
Posts: 657
Location: USGS Astrogeology Research Program, Flagstaff, AZ
|
| Posted: Mon Jan 29, 2007 9:11 pm Post subject: |
|
|
One more item to be aware of if you are hoping to use the PDS MOLA files with ISIS generated files.
The imported MOLA PDS label has a few slightly different values than those used in ISIS (including the 64ppd MOLA radius DEM in the $ISIS3DATA). The MOLA team has always defined Mars as a sphere with a radius value of 3396.0 km. ISIS uses the latest IAU 2000 elliptical defined radius of 3396.19km (& semi-minor of 3376.2km). This has the effect of slightly changing the pixel size and registration in meters. I am not sure all the repercussions of this using it in ISIS with other files generated by ISIS. The method we used for ISIS2 was to bring in the PDS MOLA files as raw and then manually add the mapping parameters into the ISIS cube. I don’t think there is a program yet in ISIS3 to define mapping labels on a raw image. Thus a possible method would be to edit the detached PDS label before importing. This would require a change in the radius, cellsize, and the sample/line projection offsets. The projection offset values might be a little tough to figure out because I am not 100% sure if they are based in the center of the top left pixel.
Otherwise, I can upload the ISIS2 MOLA defined 128ppd topographic files which should work in ISIS3 as well.
Trent
Last edited by thare on Tue Jan 30, 2007 1:25 pm; edited 1 time in total |
|
| Back to top » |
|
|
Adakite
Ra (Power Member)
Joined: 29 Mar 2005
Posts: 98
Location: IPGP, Paris
|
| Posted: Tue Jan 30, 2007 2:10 am Post subject: |
|
|
Ok, thank you very much. I am able to make a cub using PDS Mola files now. I will look at registration point and keep you informed.
thanks. |
|
| Back to top » |
|
|
Adakite
Ra (Power Member)
Joined: 29 Mar 2005
Posts: 98
Location: IPGP, Paris
|
| Posted: Thu Apr 12, 2007 4:58 am Post subject: |
|
|
Hello,
You mentionned that you could import into Isis3 a 128 px/deg DEM. It would be usefull to get a better DEM resolution than 64 px/deg in Isis distribution. Keep me informed if you plan to do that.
Thank you again. |
|
| Back to top » |
|
|
thare
GIS Support Team
Joined: 22 Mar 2004
Posts: 657
Location: USGS Astrogeology Research Program, Flagstaff, AZ
|
|
| Back to top » |
|
|
Adakite
Ra (Power Member)
Joined: 29 Mar 2005
Posts: 98
Location: IPGP, Paris
|
|
| Back to top » |
|
|
thare
GIS Support Team
Joined: 22 Mar 2004
Posts: 657
Location: USGS Astrogeology Research Program, Flagstaff, AZ
|
| Posted: Sat Apr 14, 2007 6:32 pm Post subject: |
|
|
Looks like the ftpext finally sync'd with the input server. Both files should be available now.
The north is 0 to 360 lon and 0 to 88 lat.
The south is 0 to 360 lon and 0 to -88 lat.
ISIS2 only supports files under 2.1GB so they are split into two files. These could be merged into one file in an ISIS3 cube format which supports larger file sizes.
Both are ~463 m/p or 128ppd. There is not really any data from 88 to 90 and -88 to -90. The 64ppd data at these 2deg areas at poles is very generalized.
Trent |
|
| Back to top » |
|
|
Adakite
Ra (Power Member)
Joined: 29 Mar 2005
Posts: 98
Location: IPGP, Paris
|
| Posted: Mon Apr 16, 2007 9:32 am Post subject: |
|
|
Thank you Thare,
Which projection is used for those MOLA files ?
and just a last thing,
I make a gdal_translate on a ISIS cube using -of ENVI option. I get a header readable by ENVI. But, the thing is localisation is not exactly the same that I observed in qview. The is a shift for all data (THEMIS, MOLA, MOC etc).
Is gdal_translate doing something wrong ?
How can I get the same georeferencing in ENVI than in ISIS.
NB. I'm using a Mars 2001 IAU ellipsoid in ENVI
Thanks again ! |
|
| Back to top » |
|
|
thare
GIS Support Team
Joined: 22 Mar 2004
Posts: 657
Location: USGS Astrogeology Research Program, Flagstaff, AZ
|
| Posted: Mon Apr 16, 2007 10:21 am Post subject: |
|
|
How much off? ENVI may use a elliptical equation so you may need to define Mars as a sphere to force the same projection equation ISIS is using. This fortunately also helps with the fact that the MOLA files are released in planetocentric latitudes, which ENVI may not understand.
So use 3396190 for the semi-major and minor axis.
The MOLA files are Simple Cylindrical (also known as Equirectangular or Equidistant Cylindrical) with a clon=180 and clat=0.
You should be able to "more" the *.cubs above to see the projection information defined in the label. Note that it states it is using a semi-major (A) and semi-minor (C) but the actual ISIS projection equation is only using the sphere.
Trent |
|
| Back to top » |
|
|
Adakite
Ra (Power Member)
Joined: 29 Mar 2005
Posts: 98
Location: IPGP, Paris
|
| Posted: Wed Apr 18, 2007 7:33 am Post subject: |
|
|
| Thanks again. Works perfectly now ! |
|
| Back to top » |
|
|
|
