Day Three part 1

Day 3 Objectives:
Facility in building and using SELES models to address ecological and socio-economic analysis issues

Outline

 * Building, debugging and verifying SELES models
 * Working with time series data
 * Tutorial #8 - Set up the habitat model to run on a time series
 * Introduction to Landscape Agents for spatial population modelling
 * Tutorial #9 - Create a random walker individual based model
 * Review modelling process
 * Collaborative development and communicating results
 * Scenario analysis for decision support (risk assessment)
 * Tutorial #11: Morice Case Study
 * Demonstration #3: Hydrology model

Building SELES models
select File/New
 * Start from an existing model
 * Use the Landscape event templates available in LSEditor
 * Check the “model garden” on www.SELES.info
 * Share models!
 * KISS: Keep it simple, stupid
 * Small models are beautiful (Ludwig)
 * Don’t create a model you can’t understand (Lertzman)
 * Do not build a complicated model when a simple one will suffice (Smith)
 * Start simple and increase model behaviour
 * Verify and test the model with each incremental addition

Debugging and Verifying

 * Isolate bugs and syntax errors in Scenario (*.scn) and Model definition (*.sel) files
 * try commenting out sections, retry the model, and then adding back sections one at a time once you’ve ruled them out
 * Isolate bugs in landscape events
 * try turning off events in the .sel file to narrow down which event the bug is in
 * try commenting out properties to determine which property is problematic
 * Use the simulation probe to explore variable states for each landscape event/property
 * Use DISPLAY commands to explore variable states
 * Beware of display commands when running from a SimStart command!
 * The model will hang…
 * Verification, validation and sensitivity analysis are not optional!
 * Attempt to reject your model, not support it (Lertzman)
 * Use output variables and display commands and check the results
 * Do they make sense?
 * If not, then why?
 * Never trust the results of a model unless you can explain them

Working with time series data

 * Tutorial #8 - Set up the habitat model to run on a time series


 * Tutorial #9 - Create a random walker individual based model