Augmenting Forest Stand Parameters using Landsat TM Spectral Images

Abstract:

Forest stand parameters, such as diameter at breast height (DBH), tree height (H), or volume per hectare (V), are imperative for forest resources assessment. Traditional inventory of forest stand parameters, usually based on fieldwork, is often difficult, time-consuming, and expensive, to conduct in large areas. Therefore, estimating forest stand parameters in large areas using traditional inventory approach augmented by satellites data has a significant implication for sustainable forest management and natural resources efficiency. However, obtaining suitable satellite image data for such purpose is a challenging task mainly because of insignificant knowledge between the forest stand parameters and satellite spectral response relationships. Here, we present the use of Landsat Thematic Mapper (TM) spectral responses data for augmenting forest stand parameter obtained from fieldwork at the Lahav Forest, in the Israeli Northern Negev. A new algorithm was developed in order to use all eight TM band when calculating the linear combination which correlates the most to each one of the forest stand parameters. Each linear combination is obtained first for local area inside the entire studied grid and is then fitted using a simple linear polynomial curve to the known forest stand parameter. Once the relationship between the two is characterized by a linear polynomial equation, the TM linear combination local area data is translated to the same equivalent area of the chosen forest stand parameter. At last, we interpolate the entire TM grid using a higher order polynomial fit applied to all the augmented local area combined together to attain full coverage of the desired forest stand parameter.