 This video will show you how to compute spectral indices from imagery using ArcGIS Pro. Spectral indices are computed from multiband imagery such as the Landsat image that you see here. This Landsat 8 multispectral composite has seven bands sensitive to the visible, near-infrared, and shortwave-infrared portions of the electromagnetic spectrum, making it ideal for computing many different types of spectral indices. Some spectral indices are available as raster functions. From the analysis menu, choose raster functions, and in this example, I'm going to compute NDVI or the Normalize Difference Vegetation Index. The NDVI function works by first specifying the input image dataset, and then setting the parameters for the visible and infrared bands, which are 4 and 5 respectively. I'm also checking the option for scientific output, which will mean my NDVI values will range from negative 1 to 1. A raster function generates a new layer, meaning that it's not actually writing a new raster file, it's simply creating a virtual layer computing NDVI based on my source image dataset. We can confirm this by going into the layer properties of the NDVI layer, where we see that it points to the original Landsat raster dataset. If you have the image analysis extension, additional indices can be computed by going to the imagery tab, selecting the image dataset of your choice, and then choosing the appropriate indices from the drop-down indices menu. Once again, I'll compute NDVI, only this time using the image analysis tools. These tools work similarly to raster functions, in that they generate virtual layers rather than new output raster datasets. Just like I did with the NDVI raster function, I have to specify the appropriate near-infrared and red bands to compute NDVI. It's imperative that you understand what bands correspond to what wavelengths when running any type of spectral indices. NDVI is great for analyzing vegetation. Now let's compute NDVI, the normalized difference built-up index, which is excellent for highlighting man-made built-up features on the landscape. Just as I did with NDVI, I'm going to have to specify the appropriate bands for use in the indices calculation. Finally, I'll compute NDMI, the normalized difference moisture index, which is excellent for drought monitoring, and also giving an idea of possible fire risk in fire prone areas. The indices layers I generated here are virtual ones, and not actual raster datasets. So, if I want to use them in another software package, I can right-click, choose data, and export raster, and export the raster to an external file format, such as GeoTiff or Imagine. Spectral indices are useful for highlighting certain features on the landscape, aiding in feature extraction, and even making scientific measurements. It's imperative that you understand the properties of your source data, each band and what wavelength it's sensitive to, so that you can apply the appropriate band for when you're computing the spectral indices.