Abstract

Forest information at the level of the stand is crucial for deriving reliable estimates of the capacity of the future carbon sink, yet the data are restricted to a small subset of the forest. Collection of the required datasets can be labour-intensive and costly and thus data availability is restricted. Our aim is to develop a model that allows forest stand structure to be derived from LiDAR remote sensing data; this therefore offers a tool for predicting inventory information across large scales, from which current carbon stocks can be measured and future stocks predicted.

We have developed a model for predicting the distribution of LiDAR first returns retrieved from stand-level data, demonstrating that exposed, accumulated and overlapping crown area at a given height are all critical factors in determining where a return is recorded. This model predicts to a high degree of accuracy which has allowed it to be effectively implemented in the extraction of stem diameter distributions from LiDAR return patterns across whole landscapes.