Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Abdi, Abdulhakim M., Wang, Fan
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2509.18228
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866914176795934720
author Abdi, Abdulhakim M.
Wang, Fan
author_facet Abdi, Abdulhakim M.
Wang, Fan
contents We present a new 10-meter map of dominant tree species in Swedish forests accompanied by pixel-level uncertainty estimates. The tree species classification is based on spatiotemporal metrics derived from Sentinel-1 and Sentinel-2 satellite data, combined with field observations from the Swedish National Forest Inventory. We apply an extreme gradient boosting model with Bayesian optimization to relate field observations to satellite-derived features and generate the final species map. Classification uncertainty is quantified using Shannon's entropy of the predicted class probabilities, which provide a spatially explicit measure of model confidence. The final model achieved an overall accuracy of 85% (F1 score = 0.82, Matthews correlation coefficient = 0.81), and mapped species distributions showed strong agreement with official forest statistics (Spearman's rho = 0.94).
format Preprint
id arxiv_https___arxiv_org_abs_2509_18228
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Forest tree species classification and entropy-derived uncertainty mapping using extreme gradient boosting and Sentinel-1/2 satellite data
Abdi, Abdulhakim M.
Wang, Fan
Quantitative Methods
Machine Learning
We present a new 10-meter map of dominant tree species in Swedish forests accompanied by pixel-level uncertainty estimates. The tree species classification is based on spatiotemporal metrics derived from Sentinel-1 and Sentinel-2 satellite data, combined with field observations from the Swedish National Forest Inventory. We apply an extreme gradient boosting model with Bayesian optimization to relate field observations to satellite-derived features and generate the final species map. Classification uncertainty is quantified using Shannon's entropy of the predicted class probabilities, which provide a spatially explicit measure of model confidence. The final model achieved an overall accuracy of 85% (F1 score = 0.82, Matthews correlation coefficient = 0.81), and mapped species distributions showed strong agreement with official forest statistics (Spearman's rho = 0.94).
title Forest tree species classification and entropy-derived uncertainty mapping using extreme gradient boosting and Sentinel-1/2 satellite data
topic Quantitative Methods
Machine Learning
url https://arxiv.org/abs/2509.18228