UBC Forest Ecosystem Simulation Research Group FORCEE Spatially Explicit Individual Tree Model

FORCEE INDIVIDUAL TREE MODEL
FORCEE is a spatially explicit, individual tree, stand-level, ecosystem management (EM) model. Based on the modelling approach developed for the non-spatial, stand level, EM model FORECAST, FORCEE grows trees and understory plants based on simulated photosynthesis, nutrient cycling, nutritional regulation of growth, and competition for light and nutrients. The simulation rules are based on input data that describe past tree and plant growth on sites of differing nutritional quality and the rates of certain processes. From these, estimates of the rates of the key processes are obtained from a back-casting procedure: what rates must have occurred to have resulted in the observed growth of plants and trees, and soil biomass/nutrient accumulation. FORCEE is designed for the analysis of complex stands - complex in vertical and horizontal structure, age and species composition, and spatial variation in biotically - induced forest floor and soil characteristics.
FORCEE 2008 Model Design Update
FORCEE simulates individual trees growing in pure or mixed stands. It uses visualizations to directly illustrate the outcomes of the simulation routines. All representations are pixel-based (with a 10 x 10 cm resolution) except tree stems which are vector based. The FORECAST model was based on a “data first” approach to model construction, whereby development began with available data on ecosystem processes and the model was then built around those data. FORCEE, in contrast, is based on a “structure first” approach in which priority is given to the structure of the computer program and its relationship to the structure of the reality it represents. Its design includes optimization of calculations, data structures which organize numbers into logical groups, and visual displays of the data structures arranged to appear on screen as the real structures that they represent.
Foliage Each foliage cell is held aloft to intercept light for further growth and to cast shade below according to hemispherical view analysis. A foliage-light calculation is used to determine the amount of new growth and the spatial distribution of that growth; foliage tends to grow into areas with the best light conditions. Allocation of growth to foliage itself varies regionally within each foliage grid depending on the light experience of individual foliage cells (see figures below).

Foliage growth within a single grid of cells is directed most to those existing foliage cells which are performing the best but are not yet full, and to new empty cells adjacent to these better performing cells. Branch length limits how far the new cells can be from the stem (see figure below).

Cell performance depends on light intensity, foliage age and cost of supporting stem segments and roots. Poorly performing cells are more likely to die, with the foliage drifting in the wind to land in soil cubes, adding to the forest floor (see figure below). The canopy for each tree becomes hollow as the foliage of inside cells is shaded by that of cells around and above. Less foliage grows in cells on the north side, and lower foliage dies sooner on the north side (see figure above). Shade cast by adjacent trees can be seen on the foliage and on the ground.

Stems Stems are represented by a tapering stack of nested cylinders, with three nested cylinders between each foliage grid to define heartwood, sapwood and bark diameters (see figure). Each cylinder is one meter long except for the top one which terminates at tree top height. Lower cylinders are fatter because they started growing sooner, and because wood and bark cylinder increment depend on the amount and productivity of foliage which must use those cylinders to connect with the roots.
Cylinders represent diameter over bark, so they start very narrow and increase in diameter over time (see figure below). Each cylinder may be offset from the one below it to represent the stem leaning. Coordinates are kept for the top and bottom of each cylinder so that tree stems may lean in any direction. Stems may bend to recover from lean as the stem elongates (see figure below). Stems may divide as in a deciduous tree. Tree stems sprout from the top soil layer. Tree roots spread from the stem bottom through the soil layer cells to access and uptake nutrients.



Topography and soil Soil is represented by a 3D grid of 10 cm cubes. Each cube can have a content of litter, humus and mineral soil, or the cube can be impermeable like bedrock. Bedrock surface and mineral soil surface are mapped as landform grids. When litter falls from one foliage cell and drifts to land on a column of soil cubes, it accumulates in the topmost cube. If the topmost cube is already full then it starts to fill a new cube above. Litter decomposes to humus which mixes with mineral soil. Litter and humus decomposition releases nutrients into soil cubes. When compaction of litter and humus reduces the volume of a cube which is not on top, then material will drop down from the cube above. The resulting piles of litter can be seen at the base of trees (see figure below). Humus moves between cubes through soil mixing, and released nutrients can seep from cube to cube. Roots favour growth into those contiguous cubes with the greatest content of nutrients, providing that the cubes also contain some mineral soil.

The area capable of simulation is necessarily small (10 ha, or less), though the stand has no edges since exterior boundaries join seamlessly i.e., the left side continues on the right side, the top on the bottom, and each corner on the opposite corner. A tree on the boundary will show foliage, cast shade and drop litter on both sides or on all four corners. Edge can be studied by removing trees from part of the plot. Individual trees can be cut, harvested, and new seedlings planted to achieve desired scenarios.

FORCEE Model Development
The model is in the final stages of development. Currently, trees can be established by natural regeneration or by planting (in various spatial patterns), harvested individually, in groups or strips, or by clearcutting. Planned capabilities include mechanical and fire site preparation, herb and shrub control, spot or broadcast fertilization, thinning, underplanting, pruning, and seed production and dispersal. Soil is represented by user-defined depth layers and forest floor in 10x10cm grids. Each tree creates a soil, forest floor, and light "footprint". The soil and forest floor footprint affects the growth of the next generation of trees and plants unless it is removed by simulated site preparation.
Figure: Example of the user interface and visualization capabilities of FORCEE.

Figure: Example of visual output illustrating simulated shading between trees causing asymmetrical canopy growth.

Figure: Example of visual output illustrating simulated shading within a tree causing asymmetrical canopy hollowing.

Figure: Example of visual output illustrating simulated piles of litterfall on the forest floor.

Figure: Example of visual output illustrating a top-down view of the simulated tree canopy and forest floor.