Project

Back to overview

Embracing structural uncertainty in models of forest dynamics

English title Embracing structural uncertainty in models of forest dynamics
Applicant Bugmann Harald
Number 188882
Funding scheme Project funding (Div. I-III)
Research institution Departement Umweltsystemwissenschaften ETH Zürich IBP / D-USYS
Institution of higher education ETH Zurich - ETHZ
Main discipline Other disciplines of Environmental Sciences
Start/End 01.07.2020 - 30.06.2024
Approved amount 959'036.00
Show all

All Disciplines (2)

Discipline
Other disciplines of Environmental Sciences
Pedology

Keywords (8)

Systems analysis; Pattern-oriented modeling; Long-term forest dynamics; Model assessment; Carbon turnover; Models of forest dynamics; Pattern-oriented modeling; Uncertainty analysis

Lay Summary (German)

Lead
Wälder haben eine grosse Bedeutung für das Wohlergehen der Menschheit. Deshalb sind Abschätzungen der zukünftigen Walddynamik unter sich veränderndem Klima und sich verändernder Bewirtschaftung sehr wichtig. Dazu werden dynamische Modelle benötigt. Diese fokussieren in aller Regel auf die Bäume als die bestimmende Lebensform im Wald; Bodenprozesse werden oft nur rudimentär modelliert, und die Unsicherheiten in den Modellformulierungen sind noch nie umfassend analysiert worden. Wir haben in den letzten Jahren eine Methodik entwickelt, um diese Unsicherheiten zu untersuchen.
Lay summary

Das Gesamtziel dieses Projektes ist die systematische Analyse der Unsicherheit in den mathematischen Formulierungen dynamischer Waldmodelle und ihre Reduktion auf ein Mass, das robuste Aussagen über die zukünftige Dynamik erlaubt. Dabei werden sowohl ober- als auch unterirdische Prozesse intensive studiert.

Als Fallstudien verwenden wir ein Bestandesmodell (ForClim) und ein Landschaftsmodell (LandClim). Beide Modelle sind in der Professur Waldökologie an der ETH entwickelt worden. Für den Abgleich des Modellverhaltens mit Messdaten erstellen wir eine grosse Datenbank mit verschiedensten Daten, die für den Test verschiedener Teile der Modelle dienen.

Das Projekt besteht aus fünf Arbeitspaketen (AP). Im AP1 fokussieren wir auf die bioklimatischen Einflüsse auf die Walddynamik. Im AP2 geht es um verbesserte Formulierungen der Bodenwasser-Dynamik, ganz besonders der Auswirkungen von Trockenperioden. Im AP3 geht es um die Herleitung eines neuen und umfassenden Modells der Kohlenstoff- und Nährstoffdynamik im Boden inkl. Verwitterungs- und Erosionsprozesse. Im AP4 wird die Unsicherheitsanalyse, die wir für das Bestandesmodell bereits angefangen haben, vertieft und auf das Landschafts-Modell ausgeweitet. Und schliesslich werden im AP5 die Erkenntnisse aus den anderen Teilprojekten zusammengeführt und ein neues, robusteres Bestandes- resp. Landschaftsmodell definiert und angewendet unter Szenarien des sich ändernden Klimas und sich ändernder Waldbewirtschaftung.

Direct link to Lay Summary Last update: 18.05.2020

Responsible applicant and co-applicants

Employees

Project partner

Associated projects

Number Title Start Funding scheme
163250 Advanced Tree MOrtality MOdeling (ATMO^2) 01.04.2016 Project funding (Div. I-III)

Abstract

Forests provide a wide range of benefits to human societies and will have an important role in the transition to a “bio-economy”, i.e. an economy that is based on renewable biological resources. To assess the future trajectories of forests, observations and experimentation need to be complemented by modeling activities. A wide range of models of forest dynamics (MFDs) is available, but most current models under-emphasize the belowground component. Furthermore, MFDs have developed into complex structures (“monoliths”) whose underlying assumptions and behavior are difficult to assess.Multiple approaches have been used to handle this problem: parameter sensitivity analysis, ensemble modeling and model comparison exercises have helped to shed light on some uncertainties and the underlying reasons, with substantial contributions by the Forest Ecology group at ETH (FE-ETH). However, the uncertainty underlying the choice of the formulations for the representation of ecological processes in MFDs has largely been neglected to date. We have developed a methodology to systematically investigate the structural uncertainty in MFDs regarding tree demography, with promising results that clearly demonstrate the potential of this novel approach.Thus, the overall objective of this proposal is to quantify the structural uncertainty in MFDs, and to reduce it to levels that enable the elucidation of the nature of the underlying ecological processes as well as robust projections of future dynamics, taking into account structural uncertainty.The research will focus on a stand-scale model (ForClim) and a landscape-scale model (LandClim), both of which were developed in FE-ETH and applied successfully to study a wide range of fundamental and applied research questions in forest ecology. For the benchmarking of the models, a comprehensive database will be built that comprises large amounts of data with highly varying grain and extent in time and space (from annual to millennial and from individual-tree to entire landscapes), so as to be able to test the models for emergent properties (so-called pattern-oriented modeling).The project will be structured into five Work Packages (WPs). In WP1, the formulation of bioclimatic influences on forest dynamics will be investigated. WP2 will focus on alternative formulations of soil water dynamics and particularly the representation of drought effects on tree establishment, growth and competition; also, the possible direct effects of soil moisture and drought trajectories on mortality will be investigated. In WP3, a soil carbon and nutrient turnover model will be adapted based on existing models from the literature that will be complemented by novel components to rectify deficiencies of existing approaches. WP4 will deepen the work that we have begun on structural uncertainty in the formulation of tree demography, and it will, as a novel component, emphasize the uncertainty underlying disturbances, with a focus on bark beetles. Lastly, WP5 will serve to integrate the findings from WPs1-4, ultimately providing recommendations for a suite of model formulations to be used in ensemble mode at the stand and landscape scale, respectively. This suite of models will be applied for fundamental (e.g., relationship between biodiversity and ecosystem function) and applied research questions (e.g., adaptive forest management in a global change context).The project breaks novel ground by finally and systematically assessing the structural uncertainty in MFDs, which is an important component of the overall uncertainty that is underlying (1) conclusions drawn from such models e.g. regarding the importance of certain ecological processes for long-term forest dynamics, and (2) projections of the future state and dynamics of forests and the associated multiple benefits to human societies.
-