Machine-Learning Techniques Applied to Biomass Estimation Using LiDAR Data

Abstract

With the development of artificial intelligence, alternative advanced machine learning approaches have allowed the training of increasingly sophisticated models via the available data. The light detection and ranging (LiDAR) remote sensing technique is being increasingly applied to obtain informative terrain maps, due to its ability to collect large amounts of data with satisfactory accuracy. Forest ecosystem management needs a multi-faceted approach, combining forest mapping and inventory in order to provide comprehensive knowledge on the current state and future trends of forest resources. Estimation of forestry aboveground biomass (AGB) by means of LiDAR data uses high-density point sampling data obtained in dedicated flights, which are often too costly for available research budgets. In this paper, we exploit already existing public low-density LiDAR data obtained for other purposes, such as cartography. This paper focuses on the application of machine-learning-based predictive systems for the extraction of biomass information from low-density LiDAR data (0.5 points/m2) taking into account the Pinus radiata species in the Arratia-Nervión region (Spain).