
Tree species map Germany / Baumartenkarte Deutschlands Contents / Inhalt Tree species map of Germany with discrete values (see legend below)Trained model as .pth file with hyperparameters, logs and settings.Baumartenkarte Deutschlands mit diskreten Werten (sieh Legende unten)Trainiertes Modell als .pth Datei mit Hyperparametern, logs und Einstellungen. Code availability / Code-Verfuegbarkeit All code relevant to model training is publicly available under / Fuer das Modelltraining relevanter code ist oeffentlich verfuegbar unter: https://github.com/Remote-Sensing-at-FU-Berlin/SITS_class_FUB Legend / Legende Raster value / Rasterwert Species Baumart 0 Spruce Fichte 1 Silver Fir Weißtanne 2 Douglas Douglasie 3 Scots Pine Waldkiefer 4 Oak Eiche 5 Red Oak Roteiche 6 Beech Buche 7 Sycamore Ahorn 8 Others Sonstige Accuracy / Genauigkeit Within the study area an overall accuracy of 77% is achieved using an independent test dataset. For more detailed information on specific accuracies, please consult the linked paper. Innerhalb des Studiengebiets wird eine Gesamtgenauigkeit von 77% mit einem unabhaengigen Testdatensatz erreicht. Fuer genaue Angaben der Genauigkeiten ist das verbundene paper hinzuzuziehen. Additional data availability / Verfügbarkeit zusätzlicher Daten In addition to discrete values, a map with probabilities depicting the model's prediction certainty are available. Due to the large file size, it cannot be shared in this repository. Please contact the authors. Zusätzlich zur den diskreten Werten gibt es eine Karte, die die Wahrscheinlichkeiten jeder Baumart darstellen, die das Modell errechnet hat. Aufgrund der Dateigröße kann diese nicht in diesem repository verfügbar gemacht werden. Bitte kontaktieren Sie die Autoren. CrediT authorship contribution statement Conceptualization JK, FF Methodology JK, BS Project administration FF Funding This work was funded by German Federal Ministry for Environment, Nature Conservation, Nuclear Safety and Consumer Protection in the Future Forest project under grant number 67KI21002C. Thanks, Acknowledgements We deeply appreciate the European Space Agency and the European Commission for their contribution in granting free access to Sentinel-2 data, enabling our research. We are also grateful to the platform EO-Lab that provided us with computing resources and hosts the SITS data cube. In addition, we would like to thank Regierungspräsidium Freiburg and Forst BW for providing reference data. We would also like to thank David Frantz for the development of the FORCE framework and hosting a data cube on EO-lab.
Remote Sensing, Tree species map, Deep Learning, Transformers
Remote Sensing, Tree species map, Deep Learning, Transformers
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 0 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Average | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Average |
