Estimating secondary forest structure through the integration of remote sensing and modelling
MetadataShow full item record
The growth of secondary forest (SF) on former pastures has recently become an important topic of research in Neo-tropical regions as their role in carbon sequestration is proving to be important. In spite of the importance of secondary forests, monitoring and estimation of their growth has proven to be difficult as there is a lack of permanent sampling plots. As a result, most published forest growth models have been developed for primary and managed forests. The advancement of remote sensing technology, such as high spatial resolution space-borne images, has greatly aided the field of forest growth modelling and monitoring where ground-based data are rare. Remote sensing techniques are increasingly being integrated with forest growth models in an effort to monitor and project future tropical forest yields and carbon stocks. These studies have tended to focus on the effects of changing land-use on biomass accumulation instead of forest structure and have used airborne sensors rather than satellite imagery. It is the aim of this paper to integrate a secondary forest succession model which incorporates the effects of pasture management on tree size distribution of secondary forest in central Amazonas, with remotely sensed high spatial resolution images of that area. The algorithm, combined with the allometric equations, correctly estimated tree size distribution (α 0.01); however it under-estimated mean biomass and tree density by almost 70%. Although the algorithm and allometric equations accurately estimated biomass, the algorithm estimated mean tree crown width to be 3% smaller than field data estimates. The secondary forest growth model’s simplification of plant species present in secondary forest led to the over-estimation of very small trees and thus a significant difference between the model and the algorithm. Future work should involve creating a biomass function that is more stratified to represent trees with different growth rates.