Prediction of forest soil trafficability by topography-based algorithms and in-situ test procedures

Medientyp: E-Book; Hochschulschrift

Titel: Prediction of forest soil trafficability by topography-based algorithms and in-situ test procedures

Beteiligte: Schönauer, Marian [VerfasserIn]; Jaeger, Dirk [AkademischeR BetreuerIn]; Katzensteiner, Klaus [AkademischeR BetreuerIn]

Erschienen: Göttingen, 2022

Umfang: Illustrationen, Diagramme

Sprache: Englisch

Identifikator:

RVK-Notation: ZC 77820 : Forsttechnische Verfahren

Schlagwörter: spatiotemporal modelling ; depth-to-water ; soil moisture ; soil bearing capacity ; penetrologger ; machine learning ; extreme gradient boosting ; forest operations ; trafficability prediction ; Hochschulschrift

Entstehung:

Hochschulschrift: Dissertation, Georg-August-Universität Göttingen, 2022

Anmerkungen:

Beschreibung: Moderne Waldbewirtschaftung bedingt den Einsatz von Forstmaschinen, da diese sichere und effiziente Erntemaßnahmen ermöglichen. Dennoch führen solche Maschinen häufig zu schwerwiegenden Bodenschäden, beispielsweise Verdichtung und Bodenumlagerung. In Zeiten sich ändernder klimatischer Bedingungen stellt die Sicherstellung einer ganzjährigen Nutzung mit minimalen negativen Auswirkungen auf den Waldboden eine anspruchsvolle Aufgabe dar. Eine Lösung für dieses Problem besteht in der Vorhersage der Bodenbefahrbarkeit durch kartographische Indizes. Diese Dissertation zeigt Möglichkeiten zur Vorh...

Modern forest management entails the utilization of harvesting machinery, which enables safe and efficient forest operations. Still, such machines are frequently resulting in severe soil damage, such as compaction and displacement. To maintain or even increase year-round timber mobilization with minimal negative impacts on forest soils is a challenging task, especially in times of changing climatic conditions. One solution to address this issue is the prediction of trafficability, aiming at the reduction of traffic-induced damages. Through multiple investigations, this thesis reports on met...

Modern forest management entails the utilization of harvesting machinery, which enables safe and efficient forest operations. Still, such machines are frequently resulting in severe soil damage, such as compaction and displacement. To maintain or even increase year-round timber mobilization with minimal negative impacts on forest soils is a challenging task, especially in times of changing climatic conditions. One solution to address this issue is the prediction of trafficability, aiming at the reduction of traffic-induced damages.Through multiple investigations, this thesis reports on methods to predict trafficability: (1) values of the depth-to-water (DTW) index and the topographic wetness index (TWI) were related to rut depths observed during a field trial in a broad-leaved forest stand. In addition, different terramechanical test procedures were performed and related to rut depth following the fully mechanized harvesting operation. A modified Cone Index was shown to be successful in the prediction of occurring ruts. Therefore, this parameter was chosen for use in further validations. (2) Time-series data of soil strength, quantified by the modified Cone Index, and soil moisture were captured on six study sites across Europe. The measuring results were validated against DTW-derived predictions, resulting in a prediction accuracy of 76% for soil strength, and 82% for values of soil moisture. Yet, a high share of measurements indicating soft or wet soils deviated from the predictions made. Apparently, the conjectured season-adapted representation of overall levels of soil moisture by DTW map-scenarios could not be confirmed, probably owing to site-specific effects, non-linear behaviour of water accumulation across landscapes and the omission of reliable estimations of current levels of soil moisture. (3) Such effects were considered by machine learning approaches. Tree-based machine learning models were trained on merged data, containing daily retrievals of remotely-sensed soil moisture (Soil Moisture Active Passive mission), values of DTW, TWI and openly available soil maps. This procedure significantly improved the accuracy of predictions and reduced the class error for wet soil states. With this improved trafficability prediction, mitigating measures could sufficiently be implemented in forest management, potentially leading to environmentally sound forest management and lower costs for forest operations. The required in-put data for creating DTW maps is commonly available among governmental institutions of Central and Northern Europe, and in some countries already further processed to have topography-derived trafficability maps and respective enabling technologies at hand. It is hoped that a broader adoption of these information by forest managers throughout Europe will take place to enhance sustainable forest operations. (4) The application of a mitigating technology, namely a traction-assist winch, was surveyed on a flat site where the application of such technology has not yet been investigated.In this dissertation, particular focus was placed on the spatio-temporal patterns of soil moisture and strength on several study sites, indicating the limitation of the basic DTW concept. A method to remedy existing constraints and promote adequate implementation of openly available data, particularly soil moisture retrievals, to further improve predictive tools applicable in forest operations was demonstrated.$yLinzenz

Zugangsstatus: Freier Zugang

Nur in Feld suchen:

Zuletzt gesuchte Begriffe: