Empirical validation of an agent-based model of wood markets in Switzerland

Authored by Stefan Holm, Oliver Thees, Renato Lemm, Lorenz M Hilty

Date Published: 2018

DOI: 10.1371/journal.pone.0190605

Sponsors: Swiss National Science Foundation (SNSF)

Platforms: Java

Model Documentation: ODD Flow charts Pseudocode

Model Code URLs: Model code not found

Abstract

We present an agent-based model of wood markets and show our efforts to validate this model using empirical data from different sources, including interviews, workshops, experiments, and official statistics. Own surveys closed gaps where data was not available. Our approach to model validation used a variety of techniques, including the replication of historical production amounts, prices, and survey results, as well as a historical case study of a large sawmill entering the market and becoming insolvent only a few years later. Validating the model using this case provided additional insights, showing how the model can be used to simulate scenarios of resource availability and resource allocation. We conclude that the outcome of the rigorous validation qualifies the model to simulate scenarios concerning resource availability and allocation in our study region.

Tags

Simulation Segregation explanation systems Prediction Protocol