Possibly a better classification scheme

The 1987 article on moss vegetation identification using Landsat TM proved not very useful as no Sphagnum-related information was revealed.  However, in the afternoon I combined RGB aerial photographs and the infrared (Band 1) image of color IR image using layer stack, and found the new image can be used to clearly identify different landcover in combination of ESA vegetation map.  A new spectral signature is also derived.  The classification using this signature shows very realistic distribution of Sphagnum, which can be verified in next week's field survey.