Development of models for forest variable estimation from airborne laser scanning data using an area-based approach at a plot level

Abstract

Airborne laser scanning (ALS) is increasingly used in the forestry over time, especially in a forest inventory process. A great potential of ALS lies in providing quick high precision data acquisition for purposes such as measurements of stand attributes over large forested areas. Models were developed using an area-based approach to predict forest variables such as wood volume and basal area. The solution was performed through developing an object-oriented script using Python programming language, Python Data Analysis Library (Pandas), which represents a very flexible and powerful data analysis tool in conjunction with interactive computational environment the IPython Notebook. Several regression models for estimation of forest inventory attributes were developed at a plot level.