Skip to main content

The SELES Paradigm

SELES is built on a core idea: landscape change arises from feedback between the current state of a system and the processes that act upon it.

The key propositions are:

  • Landscape change is driven by feedback between system state and definable processes or entities.
  • As agents of landscape change, processes react to and modify the landscape state in spatio-temporal contexts.
  • A spatio-temporal context is the set of information (i.e. state variables) available at a particular time and place.
  • Contexts provide a general hierarchical framework for describing landscape dynamics.

What SELES Is​

SELES is two things at once:

  1. A language for:

    • creating a spatio-temporal state space;
    • defining behaviours to navigate through that state space; and
    • specifying state change along the way.

  2. A simulation engine to run models built using that language.

By managing contexts appropriately, a wide range of model types can be built, including:

  • Natural disturbance models (fire, windthrow, insect outbreak)
  • Habitat supply models
  • Forest estate models
  • Spatial population and metapopulation models
  • Individual-based models

The General Landscape Event Process​

All of these model types are expressed using a single general process called a landscape event (see Landscape Events for full details). One of SELES's key strengths is that it makes a minimal set of assumptions about specific landscape events — other than that they:

  • recur on the landscape (Return Rate),
  • initiate in one or more cells (or even all cells),
  • may or may not occur (i.e. affect the cell enough to trigger potential spreading), and
  • optionally spread to other cells at a specified rate (Spread Rate and Spread Initiation).

Figure 1 — General landscape event process to navigate spatio-temporal contexts. The dark bars represent the passage of simulated time. Arrows represent transitions between process steps: Return Rate → Initiation → Occurrence? → Spread Initiation → Spread Rate.

(Diagram from original documentation — to be recreated)