[GRASS-dev] i.landsat.toar to be included in Grass7

Yann Chemin yann.chemin at gmail.com
Wed Oct 13 06:14:46 EDT 2010


Yann Chemin
Senior Spatial Hydrologist

www.icwater.org

On 13/10/2010, at 18:23, Hamish <hamish_b at yahoo.com> wrote:

> Yann wrote:
>> I would like to include i.landsat.toar in GRASS7.
>> I believe this will bring the necessary complement to
>> i.atcorr and the subsequent i.albedo/i.vi/etc that are part
>> of the chain processing of Landsat.
>
> I'd like to see i.landsat.acca (cloud masks) travel along with
> it too.
>
+1 for that one too
Do you have a v7 version already ?

>> Is there anybody maintaining it these days?
>
> Yes, Jorge and I have been working on getting it ready recently,
> https://trac.osgeo.org/grass/log/grass-addons/imagery/i.landsat.toar
>
Ok will read that one

> The grass7 patch is out of date now- to avoid wasted work I was
> waiting until the grass6 version settled down before recutting
> that.
>
You mean to include it in v6 too?

>
>> I would be happy to do the necessary to include it in GRASS
>> 7. In the future, I would also like to remove from
>> GRASS-ADDONS the i.dn2* modules that mostly are included into
>> i.landsat.toar but for the .MTL metadata format of some types
>> of Landsat 5 imagery (I checked the landsat_met.c in
>> i.landsat.toar and believe it is feasible to transfer the .MTL
>> parsing code across eventually).
>
> as long as functionality is retained and differences between the
> output of the two modules are for known reasons, I don't mind
> i.dn2* being "merged" into i.landsat.toar. I don't think that
> getting the L5 glovis MTL metadata files to read in i.landsat.toar
> will be too hard, as the L7 versions of the same work ok already.
> The process of making it more robust will take a while though, and
> the hardcoded file length stuff in .toar probably has to go away.

If MTL for L7 it should be ok then
I was thinking more about parsing metadata files of other format as
"merging" the modules. The i.dn* are simple dn to rad to reftoa,
nothing i.Landsat.toar cannot do already.

About the fixed length stuff so you have a proposition or should I see
how it comes up along the way?

>
> thanks,
> Hamish
>
>
>

Thank you Hamish this sounds really good
>


More information about the grass-dev mailing list