[gdal-dev] MODIS reprojection

[Rutger]

Hey,


Etienne Tourigny-3 wrote
> I use a similar command, without the -a_ullr argument

And does it then capture the correct corner coordinates? If i run your
command it does assign the sinusoidal projection, but takes the pixel
coordinates (eg 0-2048) as the corner coordinates instead of the sinusoidal
ones. 

@Laura, 
Indeed we are talking about different products, sorry for the confusion. You
mentioned in your first email the products MYD14.A2 and MOD14.A2. I thought
you meant the "MYD14 A2" product, but you attached a "MYD14" product, which
is different. Its obvious where the confusion comes from. :)

These are all the fire products:
MYD14 	Aqua 	Thermal Anomalies & Fire 	Swath 	1000m 	5 min
MOD14A2 	Terra 	Thermal Anomalies & Fire 	Tile 	1000m 	8 day
MYD14A2 	Aqua 	Thermal Anomalies & Fire 	Tile 	1000m 	8 day
MOD14A1 	Terra 	Thermal Anomalies & Fire 	Tile 	1000m 	Daily
MYD14A1 	Aqua 	Thermal Anomalies & Fire 	Tile 	1000m 	Daily
MOD14 	Terra 	Thermal Anomalies & Fire 	Swath 	1000m 	5 min

The "MOD14" & "MYD14" products are indeed swath, and should be used together
with the MOD03 with the same timestamp. 

I think the best way to process them is by using "MRT Swath" which you can
download at:
https://lpdaac.usgs.gov/tools

MRT Swath is way beyond the scope of this mailing list, but feel free to
send me a direct email about it if you have questions.

I have used Pyresample as an alternative which works very well, but requires
the use of Python:
http://code.google.com/p/pyresample/

To end with an actual GDAL related solution, there is an option to warp an
image using gdalwarp and so-called "geolocation arrays". I can get it to
work with the MODIS swaths, but i dont like the results at all. It is
previously discussed at:
http://osgeo-org.1560.x6.nabble.com/gdal-dev-reproject-python-numpy-binary-swath-lat-lon-td4978609.html

The MODIS swath products arent equally gridded and still contain what is
called the 'bow-tie' effect. (overlapping scans at high zenith angles). I
dont know if the GDAL geoloc option is supposed to handle this properly. The
image a posted earlier in the thread mentioned above clearly shows artifacts
which i cannot explain or resolve. 


Regards,
Rutger



 




--
View this message in context: http://osgeo-org.1560.x6.nabble.com/MODIS-reprojection-tp5047700p5048019.html
Sent from the GDAL - Dev mailing list archive at Nabble.com.