Kolloquium: Henrik Hartmann
Institutskolloquium
- Datum: 23.01.2025
- Uhrzeit: 14:00
- Vortragende(r): Henrik Hartmann
- Julius Kühn Institute for Forest Protection
- Raum: Hörsaal (C0.001)
- Gastgeber: Susan Trumbore
Digital forest twins – breaking the glass dome of forest vegetation models
Global
warming and climate extremes can be precursors and triggers for tree
vitality loss and mortality, often amassing to forest decline or dieback
over large spatial scales. In many instances, even tree species that
were considered drought and heat tolerant have experienced severe damage
and dieback levels, suggesting a lack of process understanding of tree
stress physiology. Research during the last decades has focused on
improving our understanding of tree physiological responses to drought
and heat and substantial progress has been achieved, like in unveiling
the role of carbon reserves during periods of resource limitation. This
knowledge may provide process realism for predictive vegetation models,
which notoriously fail in realistically forecasting mortality, but
progress is still in its infancy. In addition, vegetation modelling
often ignores important drivers of forest dynamics and dieback, like
insects or disease, making trees live under a glass dome.In
this talk, I will present innovative approaches of the recently founded
Julius Kühn-Institute for Forest Protection. In a vegetation model, we
establish forest digital twins that are continuously fed by
physiological data streams from trees in the forest, and implement
mechanisms of tree vulnerability to biotic agents using entomological
and pathological expertise. The large-scale application of this twin
requires the extensive and near real-time integration of remote sensing
data on forest condition and forest damage. Digital twins learn from
real forest responses to climate extremes and incorporate interactions
between trees and biotic agents, making the digital forest twin more
than just trees under a glass dome. They learn from instantaneous
behavior of trees to new conditions and thus constantly calibrate
themselves - away from fixed model parameters and towards dynamic,
nature-informed models. Digital twins may eventually be used as time
machines for policy decision-making, allowing embracing uncertainty
across management and climate scenarios.