I have remote sensing data of (Landsat data) 1989,1999, 2009 with 30m resolution and 1973 of 60m resolution. Which is the best possible way to increase the accuracy of this process?
Down-sampling 1973 to 30m or up-sampling all 3 data sets to 60m?
I will also go for downsampling the coarsest layer for the simple reason that it is always a shame to loose information from the finest layers. Besides, it gives you a good reason to argue for downsampling one coarse-grain layer rather than upsampling three fine-grained layers.