Site index determination using a time series of airborne laser scanning data

Abstract

Site index (SI) is determined from the top height development and is a proxy for forest productivity, defined as the expected top height for a given species at a certain index age. In Norway, an index age of 40 years is used. By using bi-temporal airborne laser scanning (ALS) data, SI can be determined using models estimated from SI observed on field plots (the direct approach) or from predicted top heights at two points in time (the height differential approach). Time series of ALS data may enhance SI determination compared to conventional methods used in operational forest inventory by providing more detailed information about the top height development. We used longitudinal data comprising spatially consistent field and ALS data collected from training plots in 1999, 2010, and 2022 to determine SI using the direct and height differential approaches using all combinations of years and performed an external validation. We also evaluated the use of data assimilation. Values of root mean square error obtained from external validation were in the ranges of 16.3%–21.4% and 12.8%–20.6% of the mean field-registered SI for the direct approach and the height differential approach, respectively. There were no statistically significant effects of time series length or the number of points in time on the obtained accuracies. Data assimilation did not result in any substantial improvement in the obtained accuracies. Although a time series of ALS data did not yield greater accuracies compared to using only two points in time, a larger proportion of the study area could be used in ALS-based determination of SI when a time series was available. This was because areas that were unsuitable for SI determination between two points in time could be subject to SI determination based on data from another part of the time series.