Publikasjoner
NIBIOs ansatte publiserer flere hundre vitenskapelige artikler og forskningsrapporter hvert år. Her finner du referanser og lenker til publikasjoner og andre forsknings- og formidlingsaktiviteter. Samlingen oppdateres løpende med både nytt og historisk materiale. For mer informasjon om NIBIOs publikasjoner, besøk NIBIOs bibliotek.
2023
Forfattere
Ana Maria De Lera Garrido Terje Gobakken Marius Hauglin Erik Næsset Ole Martin BollandsåsSammendrag
The aim of this study was to analyze the accuracy of predictions of dominant height, mean height, basal area, and volume from the nationwide forest attribute map (SR16). The analysis took advantage of field observations from 33 different forest inventory projects across Norway used for validation. Forest attributes for more than 5000 plots were predicted using non-stratified and stratified models of SR16 and the predictions were compared against corresponding ground reference values. Finally, the effect of different factors that might have influenced the prediction errors were analyzed using partial least squared regression (PLSR) to determine under which conditions the SR16 is less apt. The overall results across all plots were adequate (RMSE of 10%, MD of 2% for dominant and mean height; RMSE of 28%, MD of 4% for basal area; RMSE of 31%, MD of 5% for volume). However, when the accuracy was assessed locally for each inventory project, large differences in accuracy were observed. The MD% values for some inventory projects were substantial (>30% for basal area and volume). The results showed that stratification did not necessarily improve the results and that factors related to the forest structure had the greatest impact on the PLSR analysis.
Forfattere
Marius HauglinSammendrag
Det er ikke registrert sammendrag
Forfattere
Marius HauglinSammendrag
Det er ikke registrert sammendrag
Forfattere
Alexander Becker Stefania Russo Stefano Puliti Nico Lang Konrad Schindler Jan Dirk WegnerSammendrag
Monitoring and managing Earth’s forests in an informed manner is an important requirement for addressing challenges like biodiversity loss and climate change. While traditional in situ or aerial campaigns for forest assessments provide accurate data for analysis at regional level, scaling them to entire countries and beyond with high temporal resolution is hardly possible. In this work, we propose a method based on deep ensembles that densely estimates forest structure variables at country-scale with 10-m resolution, using freely available satellite imagery as input. Our method jointly transforms Sentinel-2 optical images and Sentinel-1 syntheticaperture radar images into maps of five different forest structure variables: 95th height percentile, mean height, density, Gini coefficient, and fractional cover. We train and test our model on reference data from 41 airborne laser scanning missions across Norway and demonstrate that it is able to generalize to unseen test regions, achieving normalized mean absolute errors between 11% and 15%, depending on the variable. Our work is also the first to propose a variant of so-called Bayesian deep learning to densely predict multiple forest structure variables with well-calibrated uncertainty estimates from satellite imagery. The uncertainty information increases the trustworthiness of the model and its suitability for downstream tasks that require reliable confidence estimates as a basis for decision making. We present an extensive set of experiments to validate the accuracy of the predicted maps as well as the quality of the predicted uncertainties. To demonstrate scalability, we provide Norway-wide maps for the five forest structure variables.
Forfattere
Stefano PulitiSammendrag
Det er ikke registrert sammendrag
Forfattere
Stefano PulitiSammendrag
Det er ikke registrert sammendrag
Forfattere
Jaime Candelas Bielza Lennart Noordermeer Erik Næsset Terje Gobakken Johannes Breidenbach Hans Ole ØrkaSammendrag
Det er ikke registrert sammendrag
Sammendrag
Taper models, which describe the shape of tree stems, are central to estimating stem volume. Literature provides both taper- and volume models for the three main species in Norway, Norway spruce, Scots pine, and birch. These models, however, were mainly developed using approaches established over 50 years ago, and without consistency between taper and volume. We tested eleven equations for taper and six equations for bark thickness. The models were fitted and evaluated using a large dataset covering all forested regions in Norway. The selected models were converted into volume functions using numerical integration, providing both with- and without-bark volumes and compared to the volume functions in operational use. Taper models resulted in root mean squared error (RMSE) of 7.2, 7.9, and 9.0 mm for spruce, pine, and birch respectively. Bark thickness models resulted in RMSE of 2.5, 6.1, and 4.1 mm, for spruce, pine, and birch respectively. Validation of volume models with bark resulted in RMSE of 12.7%, 13.0%, and 19.7% for spruce, pine, and birch respectively. Additional variables, tree age, site index, elevation, and live crown proportion, were tested without resulting in any strong increase in predictive power.
Sammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag