Features
The Patchworks model is a sustainable forest management optimization model that is unique in several respects:
Patchworks provides a convenient visual mechanism to analyze trade-offs between competing sustainability goals.
Patchworks integrates operational-scale decision-making within a strategic-analysis environment: spatially explicit harvest allocations can be developed over long-term planning horizons. Patchworks has been designed to be useful to strategic planners and to operational planners, in the development of FMP's, and in the design of compatible 20-year and 5-year operating plans.
Patchworks implements a generic capability to model a variety of spatial feature distribution objectives. Spatial targets are defined in the input datasets, and can be adjusted and customized on the fly.
Patchworks can evaluate intensive forest management investment options within the same trade-off framework as physical wood supply, habitat and old-growth retention, and patch distribution objectives.
Patchworks can simultaneously develop efficient long-term harvest allocations and haul networks. This allows analyses to generate more realistic allocations, and accurately depict delivered wood costs at a variety of supply levels. Patchworks provides a convenient framework for the development of roadless area options, including the determination of the economic cost of such options.
Patchworks can simultaneously evaluate forest operations and log transportation problems using a multiple-product to multiple-destination formulation. The model can identify in precise detail how wood will flow to mills over a complex set of road construction and transportation alternatives.
In operations research terms, Patchworks can be described as a multiple-objective goal-programming model. A variety of constraints and objectives are supplied either in the data sets or through the user interface. The model proceeds to seek a solution that minimizes the value of the objective function while not violating the constraints.
The main Patchworks window integrates a full function map viewer with a suite of spatial and aspatial simulation controls.
Patchworks is a complete off-the-shelf product, and is based on a flexible toolkit architecture that can easily be adapted to a variety of regulatory and corporate situations. Stand dynamics and silvicultural regimes can be expressed by strata or on an individual stand level. The model can account for economic terms including harvest costs, stumpage, silvicultural costs, road construction, hauling costs, and multiple product values. A variety of realistic long-term spatial allocation criteria can be applied simultaneously, such as patch size targets, adjacency constraints, sub-regional targets, zonal constraints (e.g. landscape management, visual quality objectives).
Patchworks is highly interactive. Weighting factors on the constraints and objective function terms may be adjusted as the model is running, and the model rapidly reacts to seek a new equilibrium value. The model does not terminate upon finding a solution. Instead, it remains active and ready to find alternate solutions to different input value combinations. Planners can work with the model to seek solutions that address a range of issues.
The terms of the objective function may represent different features, possibly measured in different units. The terms are combined using weighting factors, which rank the importance and contribution of each factor. This formulation allows planners to explore the interactions between physical wood supply, harvesting economics and biodiversity values.
The model is controlled through a graphical user interface. Constraints and objective function terms are adjusted using control panels. Objective values and other indicators are shown in graphical displays that update as the model operates. The current harvest allocation is displayed as a continuously updated map. Patchworks provides standard map manipulation techniques (such as zoom, pan, symbolize and reclassify) as well as an ability to interactively step through time and visualize changes in the allocation and retained forest.
Visual controls provide a simple and easy-to-understand view of the elements of the objective function.
Patchworks is similar to a strategic wood supply model. Patchworks can model long-term sustainable forest management. It is able to determine the spatial allocation of harvest blocks that best suit operational and strategic goals and constraints, including spatial allocation objectives such as opening sizes and adjacency delays. Patchworks can effectively examine the policy space issues, and at the same time develop realistic and implementable operating plans.
The Patchworks model provides visual interactive controls to influence the harvest scheduling and the retained forest landscape pattern. It is perhaps best to think of the Patchworks model as two complementary components:
a GIS point and click interface to create manual allocations; and
a scheduler that is available to suggest optimal harvest patterns.
Both of these components operate simultaneously and cooperatively.
The point and click GIS interface consists of a map viewer with complete desktop GIS viewer functions. Planners may pan, zoom to move around the map and point and click to assign treatments and timing options to individual stands or compartments. As assignments are made the system computes the impact on the indicators and the objective function, and updates the appropriate visual indicators. This provides instant feedback on the consequence of manual allocation decisions.
The scheduler module runs continuously in the background, constantly searching for adjustments to the long-term allocation that improve the value of the objective function. As the scheduler finds improvements, it makes the changes and updates the display (both the map of the current allocation, and the histograms showing indicator values).
Planners may interact with the scheduler at any time by changing the weightings of objective function terms (including disabling terms altogether), or by changing target values. As changes are made to the model parameters, the scheduler reacts and begins to change the allocation.