Mapping Age Stages of Forest Vegetation Based on an Analysis of Landsat Multiseasonal Satellite Images

Abstract

The possibility of increasing the efficiency of forest vegetation mapping is studied using Landsat multitemporal satellite images of medium spatial resolution. To simplify the process of supervised satellite-image classification, the method of automated generation of reference samples is applied based on forest inventory materials. Using the test forest areas of the southern part of Yenisei Siberia, it is demonstrated that native and secondary stands of different ages, from which regeneration series of dark coniferous and light coniferous forests are developed in different forest growing conditions, are mapped as a result of the classification of multiseasonal images with a sufficient level of significance (Kappa coefficient more than 0.7).