Since the first time I learned remote sensing, I am only focus on multipectral images. However, recently hyperspectral imageries are more widely used than before.
One of the hyperspectral imagery that available in public domain is EO-1 Hyperion. The spatial resolution is 30m -same as Landsat- which is comparable for time series analysis with Landsat OLI-8. However, Hyperion has 242 spectral!
I tried myself to produce some color composite RGB combination, and the following bands combination as the result:
VNIR Visible RGB:29-23-16 >> it will shows natural color
VNIR Vegetation RGB: 45-33-20 >> it will highlighted the healthy vegetation in red color
SWIR RGB: 204-150-93 >> it will shows false false color image
For hyperspectral analysis we need to conduct more steps than multispectral. For instance you could learn from the following url http://www.exelisvis.com/docs/HyperionVegetationAnalysisTutorial.html, for vegetation health analysis