[GRASSLIST:7485] Re: ERDAS Atcor vs. i.atcor

[Martin Wegmann]

Dear Wolfang, 

thanks for clarifying these points, sorry for the possible misunderstandings, 

best regards, Martin

On Wednesday 06 July 2005 17:57, Wolfgang wrote:
> Dear list, dear Martin
> Just a few comments to avoid misunderstandings at a later stage.
>
> Most Radiative Transfer Models (RTM's) as implemented in remote sensing
> software such as ATCOR 2, 3, (4*) in Erdas, PCI, ENVI*, FLASH for ENVI or
> 6S as independend module or loose or tight coupled in GRASS or OSSIM, make
> use of look up tables for band models. In the case of 6S these tables were
> generated with HITRAN, as far as I know Richter used MOTRAN to generate his
> tables for ATCOR. This is done for spectral intervals, in the case of 6S
> this interval is 10cm-1. It would take several days to run an atmospheric
> correction for an whole Landsat image would MODTRAN be used dirrectly. 
>
> I am thus not doing a comparison between 6S and MODTRAN but rather 6S and
> ATCOR by Richter as implemented in PCI. The limitation of ATCOR implemented
> in proprietary software is, that it is not as flexible for adaptation and
> automation.
>
> I would also like to make it clear that Christo Zietsman should get full
> credit for translating the 6S code from FORTRAN to C++, I just provided the
> funds and gave guidence. The modifications  and integration with MSG
> products are my doing.
>
> With kind regards
>
> Wolfgang
>
>
>
>  ----- Original Message ------
>  From: Martin Wegmann
>  Sent: Wednesday, July 06, 2005 15:58
>  To: GRASSLIST(engl) GRASSLIST at baylor.edu;
>  Cc: wolfluck at mweb.co.za;
>  Subject: Re: [GRASSLIST:7413] Re: ERDAS Atcor vs. i.atcor
>
>
>
> Hello,
>
> Maciek referred me to Wolfgang Lück who did a comparison between MODTRAN
> and 6S. He will publish the results end of the year. However he received
> the permission by Vermote to rewrite the Fortran code of 6S to C but not
> yet to release it under the GPL.
>
> please see this quote of his response for further informations:
>
> "Thank you for your interest in i.atcor (6S). I am busy validating and
> comparing results from this module with Atcor 2 & 3 as implemented in PCI.
> We intend to improve it so that atmospheric products from MSG can be
> integrated to define correction models more accurately. We also intend to
> improve the HOT model for relative visibility extraction with a possible
> combination of TC4 and HOT taking differences for land cover types (i.e
> water is problematic with HOT) into account. I do not expect our
> implementations to be quite as good as Richters models because we are no
> experts in atmospheric physics. I hope to publish my thesis end of the year
> and will thereby then present a quantitative comparison of the two models."
> (wolfluck_AT_mweb.co.za)
>
> regards, Martin
>
> On Friday 01 July 2005 17:36, Jonathan Greenberg wrote:
> > Martin:
> >
> > I'd recommend also posting this question to the Remote Sensing listserv
> > I moderate at UC Davis: http://www.cstars.ucdavis.edu/starserv . We have
> > a lot of people use ATCOR and other image calibration programs on this
> > listserv! Please post your responses if and when you get them to this
> > list, I'm interested in hearing what you find out!
> >
> > --j
> >
> >
> > On 7/1/05 5:02 AM, "Grit Schuster"
> >
> > wrote:
> > > Hi Martin,
> > >
> > > I can't tell you anything about the goodness or performance.
> > >
> > > I just know that the moduls use different algorithms.
> > > Atcor for ERDAS uses MODTRAN. The Grass module uses 6S.
> > >
> > >
> > > As far as I know, MODTRAN is quite widely used.
> > >
> > > The method to use for an atmospheric correction depends on
> > > different factors:
> > > on the kind of the atmospheric input data, the sensor, the spectral
> > > bands you want to correct and of course on the purpose of your data
> > > analyses.
> > >
> > > An interesting document about the atmospheric correction of ASTER data
> > > is provided by the NASA
> > > http://eospso.gsfc.nasa.gov/eos_homepage/for_scientists/atbd/docs/ASTER
> > >/a tbd-a st-09t.pdf.
> > > I think this gives a good overview about the relevance of the different
> > > parameters used for atmospheric corrections and their influence on the
> > > accuracy of the results.
> > >
> > > You could post this topic to the Applied GIS & RS mailing list
> > > (http://www.matox.com/agisrs/).
> > >
> > > If you'll find out something, I would quite interested in it :-)
> > >
> > > Sorry that I can't tell you more. Hope this helps a bit!
> > >
> > > Grit
> > >
> > > Martin Wegmann wrote:
> > >> Hello,
> > >>
> > >> I wonder if somebody can judge the goodness of performance for the
> > >> output of ERDAS Atcor (http://www.geosystems.de/atcor/) vs. GRASS
> > >> i.atcor (http://www.cs.sun.ac.za/~caz/6S/i.atcorr.html). Are their
> > >> methods comparable at all? I am not experienced in atmospheric
> > >> correction and hence all the algorithm are rather meaningless to me.
> > >>
> > >> Any links, papers, comparisons etc. are welcome, thanks in advance,
> > >> Martin

-- 
Martin Wegmann

DLR - German Aerospace Center
German Remote Sensing Data Center
@
Dept.of Geography
Remote Sensing and Biodiversity Unit
University of Wuerzburg
Am Hubland
97074 Würzburg

phone: +49-(0)931 - 888 4797
fax:   +49-(0)931 - 888 4961
http://www.biota-africa.org
http://www.biogis.de