[GRASS-dev] Landsat MTL files

GRASS GRASS developer
1

Dear all, in improving https://github.com/NikosAlexandris/i.landsat.import,
new is, for example (assuming a 'Landsat8' directory that
contains 87 decompressed and unpacked Landsat acquisitions),

i.landsat.import pool=Landsat8 -t

which will return 87 timestamps

LO81920282015061LGN00<Suffix>|2015-03-02 09:58:07.5566043 +0000
LC81920282013135LGN01<Suffix>|2013-05-15 10:00:20.1682888 +0000
..
LC81920282013279LGN00<Suffix>|2013-10-06 10:00:15.3022235 +0000
LC81920282014218LGN00<Suffix>|2014-08-06 09:58:23.0169101 +0000

useful to feed t.register (a suffix option yet to implement so as to
define the <Suffix> part).

Not all MTL files have identical strings for what concerns the DATE and
TIME of acquisition. So far, the module treats

TIME_STRINGS = ['SCENE_CENTER_TIME', 'SCENE_CENTER_SCAN_TIME']
DATE_STRINGS = ['DATE_ACQUIRED', 'ACQUISITION_DATE']

Would you know of other strings, let me please know.
Any module-related ideas welcome.

Thank you, Nikos

2

* Nikos Alexandris <nik@nikosalexandris.net> [2017-10-15 17:57:30 +0200]:

...in improving https://github.com/NikosAlexandris/i.landsat.import

The script will import scenes in independent Mapsets. This is contra to
the "convention" to import data (first) in the PERMANENT Mapset. If
there is interest, please see Stefan's valid comments at
https://github.com/NikosAlexandris/i.landsat.import/issues/2.

If such a script will land in the addons repository, it should be in
line with GRASS' best practices.

Thank you, Nikos

3

On Nov 14, 2017 1:24 AM, “Nikos Alexandris” <nik@nikosalexandris.net> wrote:

…in improving https://github.com/NikosAlexandris/i.landsat.import

The script will import scenes in independent Mapsets. This is contra to
the “convention” to import data (first) in the PERMANENT Mapset.

IMHO there is no such convention.
To use different mapsets is perfectly fine.
The PERMANENT mapset is ideal for base cartography shared with all other mapsets.

Markus

4

Hi,

Sorry if I confuse something here or cause confusion…

I do understand that it can be handy to handle Landsat scenes in different mapsets. And creating new mapsets at import can be a smart workflow indeed.

I just thought GRASS modules were supposed not to write data to other mapsets than the ones they are called from (which I understood was e.g. why Sören changed modules of the temporal framework that always wrote to TGIS DB in PERMANENT)?

Cheers

Stefan

5

On Tue, Nov 14, 2017 at 9:22 AM, Stefan Blumentrath <Stefan.Blumentrath@nina.no> wrote:

Hi,

Sorry if I confuse something here or cause confusion…

I do understand that it can be handy to handle Landsat scenes in different mapsets. And creating new mapsets at import can be a smart workflow indeed.

I just thought GRASS modules were supposed not to write data to other mapsets than the ones they are called from (which I understood was e.g. why Sören changed modules of the temporal framework that always wrote to TGIS DB in PERMANENT)?

GRASS modules should not try to write to existing mapsets because they might not have permission to do so. Creating a new mapset and then writing to the new mapset is fine.

Markus M

Cheers

Stefan

From: grass-dev [mailto:grass-dev-bounces@lists.osgeo.org] On Behalf Of Markus Neteler
Sent: tirsdag 14. november 2017 08.36
To: Nikos Alexandris <nik@nikosalexandris.net>
Cc: GRASS developers list <grass-dev@lists.osgeo.org>
Subject: Re: [GRASS-dev] Landsat MTL files

On Nov 14, 2017 1:24 AM, “Nikos Alexandris” <nik@nikosalexandris.net> wrote:

…in improving https://github.com/NikosAlexandris/i.landsat.import

The script will import scenes in independent Mapsets. This is contra to
the “convention” to import data (first) in the PERMANENT Mapset.

IMHO there is no such convention.
To use different mapsets is perfectly fine.
The PERMANENT mapset is ideal for base cartography shared with all other mapsets.

Markus

grass-dev mailing list
grass-dev@lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/grass-dev