A Digital Twin Framework For Sustainable Forest Management
Please login to view abstract download link
Increasing natural and anthropogenic pressures make sustainable forest management a critical challenge. Modern and advanced tools that integrate modelling of ecological processes and management actions into a single environment represent a valuable solution to enable adaptive management under increasing disturbance pressures. Hybrid modelling approaches and Digital Twins (DTs) offer a promising framework for simulating management options and forest stand dynamics supporting decision-making and long-term forest planning. Aiming to develop a hybrid forest growth model, we used both differential equation and agent-based approach, coupled with an immersive three-dimensional environment, in which we simulate individual tree growth, spatial competition, and stand structure dynamics in a managed conifer forest, while enabling interactive simulation of management scenarios and their effects. Tree growth is represented through a system of ordinary differential equations, describing biomass accumulation and height increment at the individual level. Tree-to-tree competition is modelled by Zones of Influence (ZOI) where ZOI radius depends on tree biomass and ZOI magnitude depends on tree height. Single tree variables (e.g. biomass, total height) are used to reconstruct architectural attributes (e.g. stem diameter, crown dimensions) via allometric relationships and integrated into an interactive 3D Digital Twin environment, where users can visualize stand dynamics and apply different management actions. Preliminary simulations reproduce realistic spatial patterns and size hierarchies emerging from local competition processes, demonstrating the capability of the model to capture structurally complex stand dynamics, avoiding a full, computationally expensive, architectural description of the tree. The model structure, based on the most relevant number of variables and interaction rules, should reduce parameter uncertainty while retaining the ability to reproduce key emergent patterns in a reasonable computational time. The simple architecture of the model and the 3D environment offer an interpretable tool for management scenarios simulation and their analysis.
