Modelling attributes of rubberwood (Hevea brasiliensis) stands using spectral radiance recorded by Landsat Thematic Mapper in Malaysia

Investigates the relationship between Landsat TM data and rubberwood stand parameters, and establishes and evaluates models for estimating stand volume and predicting area of rubber plantations in Malaysia. The total sample set of stands was divided into two independent groups: model-building and validation data sets. Regression analyses were used to explore relationships between volume and Landsat TM bands and ratio-based indices. Selected TM bands were found to be inversely related to stand volume (P<0.0001). Relations between TM data and measured stand volume were found to be significant (P<0.01), with an I² (correlation index square) of 0.79 and root mean squared error of 49.5 m³/ha. A logistic regression model produced classifications with an accuracy of 95% for predicting the area of rubber plantations. Thus, Landsat TM provides an acceptable data source for estimating wood volume and predicting area of rubber plantations.