[gdal-dev] Need to resample climate data to very fine resolution

Ivan Lucena lucena_ivan at hotmail.com
Mon Jul 13 22:31:46 PDT 2015


Bailey,

It seems like you don't need geometric transformation of your images. We can see that your input is called reprojected. You just need to re-scale it, interpolating/filling it up with pixels values based on cubic method. 

You can do that with gdal_translate -tr or -outsize parameters. 

It should be faster since it requires less floating point computation on the warping/transformation.

But it could be that gdalwarp is smart enough to figure out that it doesn't need to do any transformation.

You can give it a try.

Regards,

Ivan

From: bdmorri2 at illinois.edu
To: gdal-dev at lists.osgeo.org
Date: Mon, 13 Jul 2015 17:12:02 +0000
Subject: [gdal-dev] Need to resample climate data to very fine resolution









Hello,
 
I am having issues using gdalwarp to resample climate data to a finer resolution. Each climate raster has been clipped to the extent of Alaska, and contain >300 layers (1 for every month from 1979-2010). The current resolution of the rasters
 is >1000 km and they need to be resample to around 33 meters, making the new files go from ~8 MB to 33 GB. I have been testing out gdalwarp on these files but so far I cannot get them to resample in a reasonable amount of time. I ran one for over 15 hours
 on my universities campus cluster and it never passed the 0% complete mark.  I need to be able to get each file done in less then 48 hours since that is the time frame limit on the cluster.

 
My current code is: gdalwarp /project/oa/akpaleo/climate/ccsm4/modern/reprojected/pr_Amon_CCSM4_amip_r1i1p1_197901-201012.tif
 –tr 33.945415050894333 33.945415050894333 –r “cubic” /projects/oa/akpaleo/climate/ccsm4/modern/resample/pr.tif

 
 
I don’t know if I have written this out incorrectly to get the resample to work since I am new to gdal, but I believe this is correct.
 
Does anyone know a reasonable way to speed this process up (or get around the time limit issue) correctly without causing incorrect calculations
 to the output files? 
 
I’ve attached a link to one of the climate rasters I am using.

 
https://uofi.box.com/s/4cv7tvpnf29d70n9wb0a8t9y4oxmudql

 
 
Thank you,
 
Bailey Morrison
PhD Student
NASA Earth and Space Science Fellow
University of Illinois at Urbana-Champaign
Program in Ecology, Evolution, and Conservation Biology
Global Environmental Analysis and Remote Sensing Lab (GEARS)




_______________________________________________
gdal-dev mailing list
gdal-dev at lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/gdal-dev 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/gdal-dev/attachments/20150714/9073f8ba/attachment.html>


More information about the gdal-dev mailing list