Earthdata Forum

Hello,

I have a few question regarding SEADAS Tool re-projection.

Does SEADAS Tool use GDAL for re-projection? or what algorithm it uses to re-project the image from one projection to another?

I am trying to re-project MODIS IGBP Products (Sinusoidal projection) to UTM WGS 84, I tried to re-project it both ways using GDAL & SEADAS, I choosed resampling method as nearest neighbor. But while counting the pixel, both gives different results.

Here's a help page showing the algorithm methodology used in SeaDAS for reprojection

https://seadas.gsfc.nasa.gov/help-8.2.0/general/overview/ResamplingMethods.html

@rkshah ,

"But while counting the pixel, both gives different results."

If you are seeing a different number for rows and columns or total pixels after reprojecting from sinusoidal, that is to be expected. If you are seeing data values that we not in the source file, that suggests that nearest neighbor resampling was not used since nearest neighbor should only return data values that are present in the source file.

Thank you for the reply.

I am counting the pixel of each IGBP classes and it's different from the original sinusoidal projection than re-projected UTM projection.

I used nearest neighbor resampling method in both SEADAS and GDAL.

Is there a other way to re-project MODIS IGBP Land Cover classes (1-17) into UTM WGS 84 projection?

@rkshah,

From what I can tell, what you are seeing is what is expected. The rows and columns of the grid will change, as well as the pixel count, when you transform from the Sinusoidal grid to your UTM projection. Any tool or utility that reprojects will exhibit this. Is there anything else that leads you to believe that your results are incorrect?