[GRASS-dev] TGRASS: granularity question concerning MODIS LST maps

GRASS GIS GRASS GIS developer
1

Hi,

I am registering in GRASS GIS 7.0 some 20000 MODIS LST maps following this scheme (4 maps per day):

…
MYD11A1.A2014232.LST_Night_1km.reconstruct|2014-08-20 01:30
MOD11A1.A2014232.LST_Day_1km.reconstruct|2014-08-20 10:30
MYD11A1.A2014232.LST_Day_1km.reconstruct|2014-08-20 13:30
MOD11A1.A2014232.LST_Night_1km.reconstruct|2014-08-20 22:30
…

I fed this list into t.register.

All runs fine I think, yet I don’t understand the “Granularity” output here:

t.info modis_lst_reconstructed
±------------------- Space Time Raster Dataset -----------------------------+
| |
±------------------- Basic information -------------------------------------+
| Id: … modis_lst_reconstructed@modis_lst_reconstructed
| Name: … modis_lst_reconstructed
| Mapset: … modis_lst_reconstructed
| Creator: … metz
| Temporal type: … absolute
| Creation time: … 2015-09-12 12:10:49.603736
| Modification time:… 2015-09-12 12:15:04.975590
| Semantic type:… mean
±------------------- Absolute time -----------------------------------------+
| Start time:… 2000-03-12 10:30:00
| End time:… 2015-04-02 13:30:00
| Granularity:… 180 minutes
| Temporal type of maps:… point
±------------------- Spatial extent ----------------------------------------+
| North:… 5447750.0
| South:… 770000.0
| East:… … 7716750.0
| West:… 2168000.0
| Top:… 0.0
| Bottom:… 0.0
±------------------- Metadata information ----------------------------------+
| Raster register table:… raster_map_register_ad09fa9530ea41bd826fc5fe56b0e47c
| North-South resolution min:. 250.0
| North-South resolution max:. 250.0
| East-west resolution min:… 250.0
| East-west resolution max:… 250.0
| Minimum value min:… 10610.0
| Minimum value max:… 13852.0
| Maximum value min:… 14426.0
| Maximum value max:… 19283.0
| Aggregation type:… None
| Number of registered maps:… 20156
|
| Title:
| Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h
| Description:
| Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h
| Command history:
| # 2015-09-12 12:10:49
| t.create output=“modis_lst_reconstructed”
| type=“strds” semantictype=“mean” temporaltype=“absolute”
| title=“Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h”
| description=“Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h”
| # 2015-09-12 12:15:05
| t.register -i input=“modis_lst_reconstructed”
| type=“rast” file=“/tmp/filenames_modis_lst_maps.txt.tgrass.txt”
|
±---------------------------------------------------------------------------+

Why a granularity of 180 minutes? If RTFM, where to find it?

thanks,

Markus

2

Hi,

From 10:30 to 13:30 are 180 minutes. This is the smallest gap size between the time instances. And it is the greatest common divider between all gaps in the time series.

Ciao
Sören

Am 12.09.2015 12:29 schrieb “Markus Neteler” <neteler@osgeo.org>:

Hi,

I am registering in GRASS GIS 7.0 some 20000 MODIS LST maps following this scheme (4 maps per day):

…
MYD11A1.A2014232.LST_Night_1km.reconstruct|2014-08-20 01:30
MOD11A1.A2014232.LST_Day_1km.reconstruct|2014-08-20 10:30
MYD11A1.A2014232.LST_Day_1km.reconstruct|2014-08-20 13:30
MOD11A1.A2014232.LST_Night_1km.reconstruct|2014-08-20 22:30
…

I fed this list into t.register.

All runs fine I think, yet I don’t understand the “Granularity” output here:

t.info modis_lst_reconstructed
±------------------- Space Time Raster Dataset -----------------------------+
| |
±------------------- Basic information -------------------------------------+
| Id: … modis_lst_reconstructed@modis_lst_reconstructed
| Name: … modis_lst_reconstructed
| Mapset: … modis_lst_reconstructed
| Creator: … metz
| Temporal type: … absolute
| Creation time: … 2015-09-12 12:10:49.603736
| Modification time:… 2015-09-12 12:15:04.975590
| Semantic type:… mean
±------------------- Absolute time -----------------------------------------+
| Start time:… 2000-03-12 10:30:00
| End time:… 2015-04-02 13:30:00
| Granularity:… 180 minutes
| Temporal type of maps:… point
±------------------- Spatial extent ----------------------------------------+
| North:… 5447750.0
| South:… 770000.0
| East:… … 7716750.0
| West:… 2168000.0
| Top:… 0.0
| Bottom:… 0.0
±------------------- Metadata information ----------------------------------+
| Raster register table:… raster_map_register_ad09fa9530ea41bd826fc5fe56b0e47c
| North-South resolution min:. 250.0
| North-South resolution max:. 250.0
| East-west resolution min:… 250.0
| East-west resolution max:… 250.0
| Minimum value min:… 10610.0
| Minimum value max:… 13852.0
| Maximum value min:… 14426.0
| Maximum value max:… 19283.0
| Aggregation type:… None
| Number of registered maps:… 20156
|
| Title:
| Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h
| Description:
| Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h
| Command history:
| # 2015-09-12 12:10:49
| t.create output=“modis_lst_reconstructed”
| type=“strds” semantictype=“mean” temporaltype=“absolute”
| title=“Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h”
| description=“Reconstructed LST for AQUA and TERRA with 4 overpasses per 24h”
| # 2015-09-12 12:15:05
| t.register -i input=“modis_lst_reconstructed”
| type=“rast” file=“/tmp/filenames_modis_lst_maps.txt.tgrass.txt”
|
±---------------------------------------------------------------------------+

Why a granularity of 180 minutes? If RTFM, where to find it?

thanks,

Markus

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

3

On Sat, Sep 12, 2015 at 12:42 PM, Sören Gebbert
<soerengebbert@googlemail.com> wrote:

Hi,
From 10:30 to 13:30 are 180 minutes. This is the smallest gap size between the time instances.
And it is the greatest common divider between all gaps in the time series.

Sure but in the end that's not of great relevance when dealing with
irregular, absolute data, right?
So I can basically ignore it here? Just to be sure.

thanks
Markus

4

Hi Markus,
Am 12.09.2015 14:43 schrieb “Markus Neteler” <neteler@osgeo.org>:

On Sat, Sep 12, 2015 at 12:42 PM, Sören Gebbert
<soerengebbert@googlemail.com> wrote:

Hi,
From 10:30 to 13:30 are 180 minutes. This is the smallest gap size between the time instances.
And it is the greatest common divider between all gaps in the time series.

Sure but in the end that’s not of great relevance when dealing with
irregular, absolute data, right?
So I can basically ignore it here? Just to be sure.

I am not sure what you expect from the granularity? In case it is of no relevance for you, then you can simply ignore it. If you dont need this kind of temporal information in an aggregation process, ignore it.

However, you can use it as indicator of the import procedure and time stamp quality of your dataset. In this case indicates the granularity correctly alligned time stamps.

Ciao
Sören

thanks
Markus

5

On Sat, Sep 12, 2015 at 8:08 PM, Sören Gebbert
<soerengebbert@googlemail.com> wrote:

Hi Markus,
Am 12.09.2015 14:43 schrieb "Markus Neteler" <neteler@osgeo.org>:

On Sat, Sep 12, 2015 at 12:42 PM, Sören Gebbert
<soerengebbert@googlemail.com> wrote:
>
> Hi,
> From 10:30 to 13:30 are 180 minutes. This is the smallest gap size
> between the time instances.
> And it is the greatest common divider between all gaps in the time
> series.

Sure but in the end that's not of great relevance when dealing with
irregular, absolute data, right?
So I can basically ignore it here? Just to be sure.

I am not sure what you expect from the granularity? In case it is of no
relevance for you, then you can simply ignore it. If you dont need this kind
of temporal information in an aggregation process, ignore it.

Not sure myself: I have four irregular overpasses (i.e. maps) per day,
from which I want to generate e.g. weekly averages.
I suppose that it is not relevant then?

However, you can use it as indicator of the import procedure and time stamp
quality of your dataset. In this case indicates the granularity correctly
alligned time stamps.

ok, thanks.

(I added a granularity note in r66182 to the manual)

ciao
Markus