Tree Species Classification in a Temperate Mixed Mountain Forest Landscape Using Random Forest and Multiple Datasets

For monitoring protected forest landscapes over time it is essential to follow changes in tree species composition and forest dynamics. Data driven remote sensing methods provide valuable options if terrestrial approaches for forest inventories and monitoring activities cannot be applied efficiently due to restrictions or the size of the study area. We demonstrate how species can be detected at a single tree level utilizing a Random Forest (RF) model using the Black Forest National Park as an example of a Central European forest landscape with complex relief. The classes were European silver fir (Abies alba, AA), Norway spruce (Picea abies, PA), Scots pine (Pinus sylvestris, PS), European larch (Larix decidua including Larix kampferii, LD), Douglas fir (Pseudotsuga menziesii, PM), deciduous broadleaved species (DB) and standing dead trees (snags, WD). Based on a multi-temporal (leaf-on and leaf-off phenophase) and multi-spectral mosaic (R-G-B-NIR) with 10 cm spatial resolution, digital elevation models (DTM, DSM, CHM) with 40 cm spatial resolution and a LiDAR dataset with 25 pulses per m2, 126 variables were derived and used to train the RF algorithm with 1130 individual trees. The main objective was to determine a subset of meaningful variables for the RF model classification on four heterogeneous test sites. Using feature selection techniques, mainly passive optical variables from the leaf-off phenophase were considered due to their ability to differentiate between conifers and the two broader classes. An examination of the two phenological phases (using the difference of the respective NDVIs) is important to clearly distinguish deciduous trees from other classes including snags (WD). We also found that the variables of the first derivation of NIR and the tree metrics play a crucial role in discriminating PA und PS. With this unique set of variables some classes can be differentiated more reliably, especially LD and DB but also AA, PA and WD, whereas difficulties exist in identifying PM and PS. Overall, the non-parametric object-based approach has proved to be highly suitable for accurately detecting (OA: 89.5%) of the analyzed classes. Finally, the successful classification of complex 265 km2 study area substantiates our findings.

Download Full-text

Tree Species Classification with Random Forest Using Very High Spatial Resolution 8-Band WorldView-2 Satellite Data

Remote Sensing ◽

10.3390/rs4092661 ◽

2012 ◽

Vol 4 (9) ◽

pp. 2661-2693 ◽

Cited By ~ 322

Author(s):

Markus Immitzer ◽

Clement Atzberger ◽

Tatjana Koukal

Keyword(s):

Random Forest ◽

Spatial Resolution ◽

Satellite Data ◽

Tree Species ◽

High Spatial Resolution ◽

Species Classification ◽

Tree Species Classification ◽

Very High Spatial Resolution ◽

Very High

Download Full-text

Influence of tree species on epiphytic macrolichens in temperate mixed forests of northern Italy

Canadian Journal of Forest Research ◽

10.1139/x09-013 ◽

2009 ◽

Vol 39 (4) ◽

pp. 785-791 ◽

Cited By ~ 28

Author(s):

Juri Nascimbene ◽

Lorenzo Marini ◽

Pier Luigi Nimis

Keyword(s):

Tree Species ◽

Mixed Forest ◽

Abies Alba ◽

Northern Italy ◽

Tree Level ◽

Sensitive Species ◽

Tree Species Composition ◽

Epiphytic Lichen ◽

Lichen Diversity ◽

Temperate Mixed Forests

Tree species is a key factor in shaping epiphytic lichen communities. In managed forests, tree species composition is mainly controlled by forest management, with important consequences on lichen diversity. The main aim of this work was to evaluate the differences at tree level in macrolichen richness and composition between Abies alba Mill. and Fagus sylvatica L. in a temperate mixed forest in northern Italy, in addition to evaluating two different proportions of the two species at the stand level. Abies alba and F. sylvatica host lichen communities including several rare and sensitive species. Our findings indicate that both tree species were important for lichen diversity, since they hosted different communities. However, F. sylvatica proved to be a more favourable hosting tree for several rare and sensitive species. Species associated with A. alba were mainly acidophytic lichens, while those associated with F. sylvatica were foliose hygrophytic lichens, mainly establishing over bryophytes. The frequency of the flagship species Lobaria pulmonaria (L.) Hoffm. was a valuable predictor of cyanolichen richness and was useful in identifying sites hosting lichen communities that are potentially more sensitive to thinning and human disturbance. The results support the relevance of mixed A. alba – F. sylvatica formations among European habitats worthy of conservation.

Download Full-text

Qualitative and value production of tree species in mixed spruce-fir-beech stands under the conditions of the Western Carpathians

Central European Forestry Journal ◽

10.2478/forj-2021-0004 ◽

2021 ◽

Vol 67 (3) ◽

pp. 155-165

Author(s):

Igor Štefančík ◽

Rudolf Petráš ◽

Julián Mecko ◽

Jiří Novák

Keyword(s):

Tree Species ◽

Management Strategies ◽

Abies Alba ◽

European Beech ◽

Silver Fir ◽

Individual Tree ◽

Tree Species Composition ◽

Self Development ◽

Management Method ◽

Value Production

Abstract Value production is one of the most important information for comparing different tree species composition and management strategies in forestry. Although the value production of forest stands is affected by various factors thinning can be considered as one of the most important one. This paper aims at the evaluation of qualitative and value production in mixed Norway spruce (Picea abies [L.] Karst.), silver fir (Abies alba Mill.) and European beech (Fagus sylvatica L.) stands, which were managed by crown thinning for a period of 44 to 50 years and/or left to self-development. More than 1,500 individual trees aged from 61 to 132 years from 15 subplots established in western part of the Low Tatras Mts. and the Great Fatra Mts. in Slovakia were assessed. The proportion of stems in the highest quality A (stem quality classes) reached a low percentage, i.e. 12% in beech, 28% in spruce and 13% in fir out of the number of evaluated trees. The percentage of the highest quality log classes (assortments I + II) of beech ranged from 0 to 23% and of coniferous ones from 2 to 12%. Regarding the management method used, this percentage accounted for 0.1 to 23% for plot with self-development, whereas in plots with tending it was from 1 to 23%. Value production of coniferous tree species was always higher compared to beech, regardless of the management method. Regarding individual tree species, we found the highest value production in fir (81.4 € m−3) and the lowest in beech (46.5 € m−3).

Download Full-text

Site classification in a mixed hardwood forest (Hallerbos, Belgium) with a homogeneous ground vegetation dominated by Hyacinthoides non-scripta (L.) Chouard ex. Rothm.

Silva Gandavensis ◽

10.21825/sg.v59i0.868 ◽

1994 ◽

Vol 59 ◽

Cited By ~ 2

Author(s):

D. Maddelein ◽

B. Muys ◽

J. Neirynck ◽

G. Sioen

Keyword(s):

Tree Species ◽

Ph Value ◽

Soil Analysis ◽

Base Cation ◽

Site Quality ◽

Ground Vegetation ◽

Site Classification ◽

Species Classification ◽

Tree Species Composition ◽

Understorey Vegetation

The forest of Halle (560 ha), situated 20 km south of Brussels is covered by a beech (Fagus sylvatica) forest, locally mixed with secundary species (Tilia, Fraxinus, Acer, Quercus,... ). In almost all stands, herbal vegetation is dominated by bluebell (Hyacinthoides non-scripta). The research intended to classify 36 plots of different tree species composition according to their site quality. Three classification methods were compared: the first one based on the indicator value of the understorey vegetation, a second one on the humus morphology and a last one on some quantitative soil characteristics. According to the plant sociological site classification, the plots have the same site quality. However, humus forms differ apparently and significant differences were found in pH value and base cation saturation of the soil, abundance and biomass of earthworms and biomass of the ectorganic horizon. Tree species proved to be the main cause of these differences. The results illustrate that the herbal vegetation is not always a reliable indicator of site quality. In the case of a homogeneous vegetation dominated by one or more indifferent species, classification on humus morphology or soil analysis are more appropriate. In the forest of Halle, the tree species is probably the main cause of the observed differences in site quality.

Download Full-text

An Evaluation of Pixel- and Object-Based Tree Species Classification in Mixed Deciduous Forests Using Pansharpened Very High Spatial Resolution Satellite Imagery

Remote Sensing ◽

10.3390/rs13101868 ◽

2021 ◽

Vol 13 (10) ◽

pp. 1868

Author(s):

Martina Deur ◽

Mateo Gašparović ◽

Ivan Balenović

Keyword(s):

Spatial Resolution ◽

Satellite Imagery ◽

Tree Species ◽

Classification Accuracy ◽

New Technologies ◽

Deciduous Forest ◽

Species Classification ◽

Tree Species Classification ◽

Object Based ◽

Very High

Quality tree species information gathering is the basis for making proper decisions in forest management. By applying new technologies and remote sensing methods, very high resolution (VHR) satellite imagery can give sufficient spatial detail to achieve accurate species-level classification. In this study, the influence of pansharpening of the WorldView-3 (WV-3) satellite imagery on classification results of three main tree species (Quercus robur L., Carpinus betulus L., and Alnus glutinosa (L.) Geartn.) has been evaluated. In order to increase tree species classification accuracy, three different pansharpening algorithms (Bayes, RCS, and LMVM) have been conducted. The LMVM algorithm proved the most effective pansharpening technique. The pixel- and object-based classification were applied to three pansharpened imageries using a random forest (RF) algorithm. The results showed a very high overall accuracy (OA) for LMVM pansharpened imagery: 92% and 96% for tree species classification based on pixel- and object-based approach, respectively. As expected, the object-based exceeded the pixel-based approach (OA increased by 4%). The influence of fusion on classification results was analyzed as well. Overall classification accuracy was improved by the spatial resolution of pansharpened images (OA increased by 7% for pixel-based approach). Also, regardless of pixel- or object-based classification approaches, the influence of the use of pansharpening is highly beneficial to classifying complex, natural, and mixed deciduous forest areas.

Download Full-text

Comparison of support vector machine, random forest and neural network classifiers for tree species classification on airborne hyperspectral APEX images

European Journal of Remote Sensing ◽

10.1080/22797254.2017.1299557 ◽

2017 ◽

Vol 50 (1) ◽

pp. 144-154 ◽

Cited By ~ 95

Author(s):

Edwin Raczko ◽

Bogdan Zagajewski

Keyword(s):

Neural Network ◽

Support Vector Machine ◽

Random Forest ◽

Tree Species ◽

Support Vector ◽

Species Classification ◽

Tree Species Classification ◽

Neural Network Classifiers

Download Full-text

Tree Species Classification Based on Hybrid Ensembles of a Convolutional Neural Network (CNN) and Random Forest Classifiers

Remote Sensing ◽

10.3390/rs11232788 ◽

2019 ◽

Vol 11 (23) ◽

pp. 2788 ◽

Cited By ~ 4

Author(s):

Uwe Knauer ◽

Cornelius Styp von Rekowski ◽

Marianne Stecklina ◽

Tilman Krokotsch ◽

Tuan Pham Minh ◽

...

Keyword(s):

Neural Network ◽

Random Forest ◽

Convolutional Neural Network ◽

Tree Species ◽

Classification Performance ◽

Majority Voting ◽

Imaging Data ◽

Species Classification ◽

Weighted Voting ◽

Linear Discriminant

In this paper, we evaluate different popular voting strategies for fusion of classifier results. A convolutional neural network (CNN) and different variants of random forest (RF) classifiers were trained to discriminate between 15 tree species based on airborne hyperspectral imaging data. The spectral data was preprocessed with a multi-class linear discriminant analysis (MCLDA) as a means to reduce dimensionality and to obtain spatial–spectral features. The best individual classifier was a CNN with a classification accuracy of 0.73 +/− 0.086. The classification performance increased to an accuracy of 0.78 +/− 0.053 by using precision weighted voting for a hybrid ensemble of the CNN and two RF classifiers. This voting strategy clearly outperformed majority voting (0.74), accuracy weighted voting (0.75), and presidential voting (0.75).

Download Full-text

Tree Species Classification of the UNESCO Man and the Biosphere Karkonoski National Park (Poland) Using Artificial Neural Networks and APEX Hyperspectral Images

Remote Sensing ◽

10.3390/rs10071111 ◽

2018 ◽

Vol 10 (7) ◽

pp. 1111 ◽

Cited By ~ 4

Author(s):

Edwin Raczko ◽

Bogdan Zagajewski

Keyword(s):

Tree Species ◽

National Park ◽

Accuracy Assessment ◽

Research Area ◽

Hyperspectral Data ◽

Hyperspectral Images ◽

Species Classification ◽

Tree Species Composition ◽

Tree Species Classification ◽

R Programming

Knowledge of tree species composition is obligatory in forest management. Accurate tree species maps allow for detailed analysis of a forest ecosystem and its interactions with the environment. The research presented here focused on developing methods of tree species identification using aerial hyperspectral data. The research area is located in Southwestern Poland and covers the Karkonoski National Park (KNP), which was significantly damaged by acid rain and pest infestation in the 1980s. High-resolution (3.35 m) Airborne Prism Experiment (APEX) hyperspectral images (288 spectral bands in the range of 413 to 2440 nm) were used as a basis for tree species classification. Beech (Fagus sylvatica), birch (Betula pendula), alder (Alnus incana), larch (Larix decidua), pine (Pinus sylvestris), and spruce (Picea abies) were classified. The classification algorithm used was feed-forward multilayered perceptron (MLP) with a single hidden layer. To simulate such a network, we used the R programming environment and the nnet package. To provide more accurate measurement of accuracy, iterative accuracy assessment was performed. The final tree species maps cover the whole area of KNP; a median overall accuracy (OA) of 87% was achieved, with median producer accuracy (PA) for all classes exceeding 68%. The best-classified classes were spruce, beech, and birch, with median producer accuracy of 93%, 88% and 83%, respectively. The pine class achieved the lowest median producer and user accuracies (68% and 75%, respectively). The results show great potential for the use of hyperspectral data as a tool for identifying tree species locations in diverse mountainous forest.

Download Full-text

Muster und treibende Kräfte der Samenproduktion bei Waldbäumen

Schweizerische Zeitschrift fur Forstwesen ◽

10.3188/szf.2016.0316 ◽

2016 ◽

Vol 167 (6) ◽

pp. 316-324 ◽

Cited By ~ 1

Author(s):

Thomas Wohlgemuth ◽

Anita Nussbaumer ◽

Anton Burkart ◽

Martin Moritzi ◽

Ulrich Wasem ◽

...

Keyword(s):

Tree Species ◽

Driving Forces ◽

Abies Alba ◽

Forest Tree ◽

Fruit Production ◽

Mast Seeding ◽

Silver Fir ◽

Fine Scale ◽

Central Plateau ◽

Forest Tree Species

Patterns and driving forces for seed production in forest tree species Why is the annual fruit production in forest tree species not constant, and which factors cause massive fruit production (seed mast year)? These and other related questions were already posed more than 100 years ago when tree breeding was economically beneficial. The questions have not been fully answered, yet. Rather, the same questions are studied again today as the climate is changing and the uncertainty about the continuation of forests at their current locations is growing. A 25 year long observation series on the variation of fruit production in Switzerland revealed a mean frequency of three years for mast seeding (full and medium mast) at low elevation on the Central Plateau in European beech (Fagus sylvatica), oak (Quercus petraea, Q. robur) and silver fir (Abies alba). In contrast, mast seed years of Norway spruce (Picea abies) occurred, on average, every sixth year. In 1992 and 2011, all four species synchronously showed mast seeding. The results are discussed in the light of different theories and new research findings. From the state of the current know ledge, we derive the need for long-term and fine-scale baseline data and present the new reporting and information webpage «mast web.ch». Here, volunteers can report observations on the fruit production of main tree species following a few simple criteria (citizen science approach). With this data, distribution maps on mass fructification levels will be made available and will serve for spatio-temporal fine-scale studies on mast seeding phenomena.

Download Full-text