Boston University
Forecasting forest futures: integrating mechanism-based modeling and remote sensing for climate-adaptive restoration
Abstract
dc:description.abstractDesigning effective, climate-adaptive forest restoration strategies requires a predictive understanding of both the fundamental biophysical limits of tree growth and the dynamic ecological disturbances that continuously reshape them. Traditional, empirically derived growth models, which rely heavily on historical observations, are fundamentally ill-equipped to forecast forest structure and function under novel, non-analog environmental conditions. To address this, this dissertation develops a multi-scale, mechanism-based modeling framework designed to support carbon-oriented silviculture and ecosystem restoration. To establish a theoretical baseline, this research first develops a static, equilibrium water-balance model on Google Earth Engine. This lightweight model predicts the potential canopy top height (pCTH) of northeastern United States Beech-Maple-Birch forests by solving for the equilibrium between environmentally driven water availability and canopy transpirational demand. Comparisons with existing canopy height products demonstrate model fidelity in mature, undisturbed stands. Because disturbances prevent landscapes from maintaining this theoretical equilibrium, the research transitions to temporal dynamics using an individual-based forest landscape model coupled with an agent-based management engine. Simulations of post-disturbance recovery show that synergistic wind and bark beetle disturbances suppress landscape-level standing carbon far below theoretical limits. Crucially, this dynamic modeling identifies a silvicultural trade-off: management cannot simultaneously maximize absolute carbon storage and long-term ecosystem resilience. Passive recovery maximizes carbon at the cost of demographic stagnation, Salvage+Planting accelerates functional recovery but creates a homogenized structural trap, and Plenter Thinning ensures structural complexity and regenerative capacity at the expense of standing woody biomass. To resolve the parameter uncertainty that hinders broad spatial application of theoretical models, the final phase bridges mechanism-based modeling and observational big data. Utilizing a Markov Chain Monte Carlo (MCMC) Bayesian inference framework, the static equilibrium model is calibrated using high-resolution canopy height observations from the GEDI LiDAR mission. This model-data fusion mathematically constrains critical plant functional traits across diverse forest type groups, yielding a 1-kilometer resolution mapping of potential canopy height across the Eastern United States under current climatic scenario.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- doctoral
- Discipline thesis:degree_discipline
- Earth & Environment
- Grantor
- Boston University
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Zuo, Zhenpeng
- Advisor dc:contributor.advisor
-
- Myneni, Ranga B.
- ORCID dc:identifier.orcid
Subjects
dc:subject × 8Rights
dc:rights- Statement dc:rights
-
- Attribution 4.0 International
- Licence dc:rights.uri
- Language dc:language.iso
- en_US
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- https://hdl.handle.net/2144/53570