 Pan-sharpening is a form of spatial enhancement in which we take low-resolution multispectral imagery and integrate it with higher-resolution panchromatic imagery. In this example I'm going to be working with a Landsat 8 scene. You can see displayed right now is the multispectral imagery, there's the panchromatic imagery, then finally you can see the output of my pan-sharpened process. For the multispectral composite I stacked together the first 7 bands of Landsat 8, bands 1 through 7, into a 13 meter multiband composite. The panchromatic image from Landsat 8 band 8 has a resolution of 15 meters. Pan-sharpening works by preserving some of the spectral information in the multispectral imagery with the spatial detail in the panchromatic imagery. Here I'm examining the output of my pan-sharpened process in which I have a new 4 band 15 meter resolution multispectral data set. Now I'll compare the original multispectral imagery and panchromatic data sets to the new pan-sharpened data set. Here we're looking at the difference between the multispectral imagery and the panchromatic imagery. As we can see the multispectral imagery is more spectral information and the panchromatic data set has greater spatial detail. Switching over now to a comparison between the original multispectral image and the output panchromatic image, we can see that the output panchromatic image has improved spatial detail and it has still preserved some of the spectral information in the original multiband image data set. Not all panchromatic routines preserve the original spectral information. Recall that my original multispectral image had 7 bands whereas the output panchromatic image only has 4 bands. Now we'll take a look at how this pan-sharpened data set was created. The geo-processing tool I used is called create pan-sharpened raster data set. As you can see the raster input is the multispectral imagery in which I specified the red, green, blue and infrared channels. And thus the output only consists of these 4 individual bands. It's imperative that you assign the appropriate channel to the appropriate band in your multispectral data set when doing any pan-sharpening. Scrolling down you can see that I've set the output raster data set to be called pan-sharp. I've specified my panchromatic image to be my Landsat 8 panchromatic image and then I've set the pan-sharpening type. The Gram-Schmidt pan-sharpening routine that I've selected here is sensor specific. I've selected the appropriate sensor, Landsat 8, and the routine has automatically populated the weights for all 4 bands. Selecting the appropriate pan-sharpening type requires a bit of understanding about the various capabilities and limitations of each option along with a bit of trial and error. As you can see if I change the pan-sharpening type my options also change. Some of the pan-sharpening types are not sensor specific while others like the Gram-Schmidt that I've used here are. Regardless of the routine you use you may find it necessary to adjust the band weights used in the pan-sharpening process. That being said, if your pan-sharpening type gives you the option to select a sensor please ensure that you choose the appropriate sensor for your input data. Not only do not all pan-sharpening routines return the same number of bands in the output pan-sharpening data set as there were in the original multi-special image data set, but some of those bands may be switched around. Not only does my pan-sharpening data set contain fewer bands, but those bands have been swapped around. For example 432 is a natural color composite in the multi-special data set whereas 123 is now a natural color composite in the pan-sharpening data set. Creating a color infrared composite in the pan-sharpening data set requires that I use bands 4, 1, and 2. As band 4 is near infrared band 1 is red and band 2 is green. To do the same for the multi-special data set I would choose bands 5 which is near infrared, band 4 which is red, and band 3 which is green. This is why it's so important that you visually assess the output of your pan-sharpening process so that you understand the properties of your new pan-sharpening data set. Pan-sharpening can be a useful technique to apply when you want to spatially enhance your multi-special imagery with corresponding higher resolution panchromatic data, but you'd need to pay special attention and understand the inputs and outputs of this routine.