DYNA-PLAN Model for Simulating & Optimizing Landscape Level Forest Management
Purpose: Project forest harvesting, growth, and natural disturbances at the landscape level in order to predict impacts from climate change on timber supply, wildlife habitat, biodiversity, etc.
DYNA-PLAN is a raster based, spatially explicit forest estate model that is capable of implementing natural disturbances over time while also seeking optimal harvest schedules. The basic unit used in the model is a stand (raster cell), where each stand belongs to an ecosystem group that has similar attributes (volumes, snags, coarse woody debris, etc.) for a given age. Stands can have multiple eligible treatment regimes that the model can select from. Once a specific treatment regime is initiated by the model, it implements a set of activities at predetermined times (i.e. clearcut, plant, fertilize).
The model’s job is to select harvest timing and treatment regimes for each stand over time that will best meet user defined objectives. Common objectives include maximizing net present value from the forest, maintaining specific levels of old growth forest, limiting the amount of active road in each period, or limiting the amount of disturbance within a landscape unit. DYNA-PLAN makes its harvest decisions using a unique approach that is decentralized or ‘bottom up.’ Optimal solutions are first determined for each stand while considering the state of it’s neighbours (stands try to be similar to neighbours). Where necessary to meet forest level objectives, local decisions are adjusted by the model.
The way in which natural disturbances are implemented into the solution process is similar to real world processes. As disturbances occur on the landbase, future plans are constantly being changed to react to the changed state of the landscape. In the model, disturbances are simulated for the first period only and then an optimal harvest schedule is found for all future periods. The schedule (or plan) is then implemented for the first period. Then disturbance is simulated in the second period, and the landscape management plan is reoptimized from the second period onward. The cycle of simulation/reoptimization continues until disturbance has been implemented in each period and harvest reoptimized in all future periods.
The model requires the following list of inputs:
- Inventory data (e.g. age and species composition)
- Spatial data (roads, spatial restrictions)
- Growth and yield functions
- Assumptions about disturbances: fire, drought, disease, etc.
- Management strategies: type of harvesting, silviculture treatments, reserves/retention
- Prices and cost information to inform the economic component of the model
- Only one climate condition can be modelled in a given DYNA-PLAN scenario. For example, all assumptions around landbase size and productivity would be based on a single climate condition (i.e. Hadley global emissions scenario A1F1 in 2050) and it would be assumed to be constant throughout the DYNA-PLAN scenario.
This page has the following sub pages.