Project

Yunhao (Davis) Xu MASc thesis

An active MASc thesis project developing a principal-agent, rolling-horizon optimization model for post-wildfire salvage harvesting and policy design in British Columbia.

Yunhao (Davis) Xu's MASc thesis, working-titled Modeling Salvage Harvesting After Wildfires: A Linear Bi-level Principal-agent Approach, develops decision-support models for post-fire salvage harvesting in British Columbia.

The project frames salvage harvesting as a principal-agent problem: the forest landowner or policy-maker wants to recover value, maintain timber supply, and support forest recovery, while forest companies decide whether salvage operations are economically feasible.

The modelling workflow combines rolling-horizon planning with single-level and bi-level optimization so the system can update as wildfire disturbance, burn severity, timber degradation, and salvage availability change through time.

A Williams Lake Timber Supply Area case study uses controlled semi-synthetic inputs built from public forest inventory, timber-supply-area, fire burn severity, and elevation datasets to test salvage schedules and policy instruments.

The thesis evaluates instruments such as stumpage credits, salvage windows, volume constraints, and harvest-sequence rules, with outputs focused on discounted net present value, product-grade mix, salvage volume over time, retention, regeneration, subsidy allocation, and stand-selection patterns.

The planned public research contribution is an open-source Python modelling framework with preprocessing scripts, optimization model files, scenario data, solver configuration, and documentation for reproducible wildfire-salvage planning experiments.

Status

  • Ongoing MASc thesis

People And Roles

  • Yunhao (Davis) Xu: MASc student
  • Gregory Paradis: supervisor

Outputs

  • MASc thesis in progress
  • Principal-agent optimization model for post-wildfire salvage harvesting
  • Rolling-horizon wildfire salvage planning workflow
  • Williams Lake Timber Supply Area case-study model inputs and scenarios
  • Planned open-source Python modelling framework, documentation, and reproducible scenario materials
  • Thesis, code, and publication links when publicly available

References And Downloads

Xu, Y. Modeling Salvage Harvesting After Wildfires: A Linear Bi-level Principal-agent Approach. MASc thesis in progress, University of British Columbia.