[gdal-dev] 'gdalwarp.exe' for reprojection

Yuta Sato yutaxsato at gmail.com
Sun Jul 19 07:40:13 PDT 2015


Dear Even:

Thanks for your response. I'm sorry to specify the MODIS data I'm using.
It's not the swath product the thread is talking that GDAL is not
appropriate. Yes, in case of swath product which requires gridding, thanks
for your improvement of the gdal_grid, I gonna use of it for swath data.


My question is about the MODIS data which was already gridded. The HDF data
has 2400 by 2400 pixels in sinusoidal projection. My concern is during
conversion into GCS, what parameters gdalwarp uses?
PIXEL_STEP = ?
LINE_STEP = ?
ULX and ULY = ? (HDF file's topleft pixel's topleft corner or center corner)

I am using nearest interpolation.





On Sun, Jul 19, 2015 at 9:02 PM, Even Rouault <even.rouault at spatialys.com>
wrote:

> Yuta,
> >
> > When I convert a HDF file in Sinusoidal Projection into Geographic
> > Coordinate System using gdalwarp.exe, how can I specify the following
> > parameters?
> >
> > PIXEL_OFFSET = 0
> > LINE_OFFSET = 0
> > PIXEL_STEP = 1
> > LINE_STEP = 1
>
> You could override the values in a VRT produced by gdal_translate -of VRT
> source_dataset out.vrt
> But it is dubious that changing the original values reported by the driver
> is
> going to produce something correct.
>
> >
> > I want to make sure that there are not any pixel shifts during
> > re-projection.
> > If gdalwarp.exe is not a good option for it, what are other methods for
> it?
> > I want to reproject a MODIS file.
>
> There have been recent discussions about why GDAL warping is not
> appropriate
> for datasets that have bow-tie :
> http://osgeo-org.1560.x6.nabble.com/gdal-dev-
> gdalwarp-problems-with-MODIS-L1B-td5206583.html
>
> There are a few hints in
> http://stackoverflow.com/questions/27968890/reproject-
> and-mosaic-modis-level2-using-gdal-python
>
> At that time, I see you were looking for a fast(er) gdal_grid. Recently
> I've
> added in trunk (post 2.0) a linear interpolation method which runs several
> magnitude of order faster than the other methods.
>
> Even
>
> --
> Spatialys - Geospatial professional services
> http://www.spatialys.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20150719/cd21e688/attachment.html>


More information about the gdal-dev mailing list