Introduction to Agent Based-Models Using NetLogo

Monday

• Morning. Lecture: Introduction to agent-based modelling.

In the first lecture, we will give a general overview of model applications in ecology and socio-ecological systems. We will then focus on our main course topic: Process-based, spatially explicit agent-based models. This model type is suitable in the context of many ecological and socio-ecological problems, as it allows incorporation of many sources of natural variability and heterogeneity (e.g. spatial, temporal or individual variability).

• Afternoon. Lecture: First Steps in NetLogo.

In the afternoon session, we will introduce the participants to a development environment and a domain specific computer language for agent-based modelling, NetLogo. We will present basic functionality and example models during the first NetLogo session.

Tuesday

• Morning. Classroom exercise: Example Model in NetLogo.

Our main intention is that after the course the participants are able to implement and analyse simple models in NetLogo. Therefore, a substantial amount of time is dedicated to hands-on exercises and project work. Thus, in the morning, we will demonstrate the development and implementation of an ecological model in NetLogo.

• Afternoon. Project work: Presentation of instructive ecological and socio-ecological research questions that will be solved by participants in groups of two during the week applying modelling.

In the afternoon, we will introduce participants to the instructive modelling projects they will develop during the course. In essence, we will suggest a selection of ecological and socio-ecological research questions. Participants will select and investigate one of them by developing, implementing and analysing their own agent-based model in NetLogo. The suggested modelling problems will cover fields of population ecology, behavioural ecology and resource management. The modelling projects are chosen specifically to familiarise participants with strategies to develop solutions to typical practical problems in the field of ecological and socio-ecological modelling.

Wednesday

• Morning. Project work: Development and implementation of models to solve the instructive ecological and socio-ecological research question.

In the morning participants will work on their projects assisted by the course instructors.

• Afternoon. Lecture: Analysing and testing models.

In the afternoon lecture, we will discuss the issues of model analysis and testing. In particular, we will address the analysis of complex and stochastic model outcomes. We present techniques of pattern-oriented modelling (sensu Grimm et al. 2004, Science) as a means for multi-dimensional model calibration. Sensitivity analysis will be introduced to identify how simulation results depend on the chosen parameters and their values. Therefore, we will introduce the NetLogo tool BehaviorSpace and present examples from case studies.

Thursday

• Morning. Project work: Analysis of self-developed models.

In the morning participants will finish their model projects assisted by the course instructors.

• Afternoon. Lecture: Model documentation.

In the afternoon lecture, we will discuss documentation and communication of models. We will introduce the ODD-protocol (Grimm et al. 2006, Ecol. Model.). ODD is becoming a standard for the description of complex individual-based models; it facilitates model comparison and communication. To practice model description and communication, participants will be given time to write an ODD description for their own model project. In this process, they will make use of the Info tab of NetLogo.

Friday

• Morning. Project presentation, wrap up and feedback.

The last (half) day is reserved for the presentation of the participants’ self-developed projects. Each group will present its project and discuss their experience using NetLogo and the ABM approach. Finally, there will be time for synthesising comments by the instructors and feedback by the participants.