Jornal de Sensoriamento Remoto e GIS

Abstrato

Modeling Forest Stand Volume and Live Aboveground Woody Biomass Using Landsat 5 TM Satellite Imagery Spectral and Textural Features

Tariku Geda

Even though, field-based forest surveying provides highly accurate measurements, it has limitations with regard to incurring high cost, being time consuming and having low spatial coverage and frequency. Taking this challenge into account, this study presents the utility of Landsat 5 TM satellite imagery spectral and textural features for the estimation of forest stand level stem volume and live aboveground woody biomass (AGB) for Eucalyptus globulus plantation forest. The research was conducted to improve accuracy and decrease uncertainties in the modern approach in general, and replace the classical approach in the study site in particular by developing a function that estimate both attributes (dependent variables) as a function of spectral and textural features. The modeling of the stem volume and AGB equations as a function of spectral and textural independent variables were developed using ordinary least square regression method.