I am using the incredible MODISDownload function from the rts package in R to download NDVI data on vegetation index in Africa.

I have two questions.

I am using the latest version of the package available on github, and I download and reproject the images with the following code (which works beautifully):

ModisDownload(x=x,h=c(19),v=c(5), version = '006', dates=c('2000.01.01', '2000.02.28'),
              MRTpath=path,mosaic=F,proj=T, bands_subset="1 0 0 0 0 0", 
              proj_type="GEO",proj_params="0 0 0 0 0 0 0 0 0 0 0 0 0 0 0",

where x = "MOD13Q1" and path is the path to my MRT/bin folder. I choose only band 1 because I am interested in the NDVI index, which is the first band. I choose proj_type = "GEO" as I want the projection on longitude/latitude format.

I am not very sure about my choice of the proj_params and pixel_size parameter. For the former, I mimic what is automatically set if I do the reprojection using the MRT GUI interface, but I am not sure whether that is correct.

For the pixel_size, though, I am completely lost. It is clear that the smallest its values, the higher the resolution of the outcome tif file. However, the higher will the its size on disk. Thus, I was wondering whether someone would know how can I calculate an 'optimal' pixel size for the 250m resolution.


Always check reference manual. See Apendix C to know about projection parameters. They are used for datum conversions:

enter image description here

If you want a 250m pixel size, use a meter projected CRS like WGS 84 UTM zone X n/s and set pixel size = 250. Working with a lat/long projection isn't possible to get a 250m pixel size for all locations.

| improve this answer | |
  • That is exactly what I was looking for. Thank you so much! – Bruno Conte Leite Dec 12 '17 at 10:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.