Hi

Generally, satellite hyperspectral data suffer from low spatial resolution. In a dense  and homogeneous vegetation community (e.g., dense forest), we can freely classify tree species with a low/medium resolution (e.g., 30 m) images. But in sparse community with heterogeneous spices, using of aerial hyperspectral images with high spatial resolution can be more useful. Another words, your study area and your budget determine your methods! Meanwhile, aerial images have some advantages in aspect of acquiring time, scale and etc.

Good luck

Hello Partha.

You can also use the satellite sentinel 2, new in EarthExplorer (http://earthexplorer.usgs.gov/), with a spatial resolution minimun of 10 m. The spatial resolution is dependent on the particular spectral band:

:1.- 4 bands at 10 meter: blue (490 nm), green (560 nm), red (665 nm), and near-infrared (842 nm).

2.- 6 bands at 20 meter: 4 narrow bands for vegetation characterization (705 nm, 740 nm, 783 nm, and 865 nm) and 2 larger SWIR bands (1,610 nm and 2,190 nm) for applications such as snow/ice/cloud detection or vegetation moisture stress assessment.

3.- 3 bands at 60 meter: mainly for cloud screening and atmospheric corrections (443 nm for aerosols, 945 nm for water vapor, and 1375 nm for cirrus detection)..

With a supervised classification you can resolve your question.

Best regard.

Emilio Ramírez Juidías.

If you have reference spectra, i would suggest to use Spectral information divergence method.

Good luck