Investigation on leaf spectral reflectance of most important species of Caspian forests using field Spectroradiometry

Remote sensing systems especially hyperspectral remote sensing for recognizing tree species and mapping forest type requires fundamental knowledge of spectral reflectance of desired tree species. A field of remote sensing named field spectrometry dealing with determination of these spectral characteristics tries to provide the spectral libraries for different objects. The main objective of this study was to prepare the spectral signature of the most important forest tree species of the northern forests of Iran. With this respect for the first time, field Spectroradiometric measurements were carried out using a full range spectrometer with Wageningen University of the Netherlands corporation during the summer 2007. The spectral fingerprint of five tree species namely Fagus orientalis, Quercus castaneifolia, Carpinus betulus, Alnus subcordata and Parrotia persica located in the Experimental Forest of Tehran University was prepared. A total of 321 leaf pile spectral curves were acquired of 107 trees of the mentioned species in altitude gradients between 400m and 2100m under clear and cloudless sky. A total of 100 scan for each leaf pile sample have been performed and averaged. After quality control and noises remove, the spectral fingerprint of the species was prepared along 350-2500nm. In order to investigate the spectral reflectance differences, five important vegetation indices related to leaf chlorophyll content were calculated and statistically analyzed. We conclude that the Vogelmann index and Simple Ratio is more sensitive to chlorophyll content in comparison to the other indices. It shows that hornbeam is significantly different in spectral signatures compared to beech, oak, and alder as well as ironwood with alder and oak being statistically different (p<0.0001, α= 0.01). For further investigation of spectral reflectance differences, we propose to make a correlation with extracted chlorophyll content and investigate the suitable vegetation indices related to other biochemical constituents and water content.