Peat-fire is a complex phenomenon with various possible causal factors, involving numbers of processes from hydrological, extreme weather and ecological conditions, as well as social-economic and cultural factors. Thus, management decision for managing peat-fire should be equipped with understanding all of those factors and processes. Unfortunately, the existing body of knowledge does not provide any decision support tool for peat fire management, involving those complex systems. Individual / Agent – Based Model (IBM/ABM) is an emerging scientific and management approach which able to reveal complex phenomena using bottom-up mechanistic processes and capable of providing various management options for dealing with peat-fire in Indonesia. Peatfire is a computer model to simulate the ignition and spreading of fire throughout tropical peatland ecosystem. The purpose of PeatFire is to simulate existing knowledge on the mechanisms of peat fire ignition and spreading which emerges from the interactions of household behavior, peat hydrological conditions, and the environmental dynamics. It allows various factors causing fire ignition and spreading to be represented, to understand their influence to landscape transformation. Peatfire can emerge from the interactions of human-triggered land conversion, peat hydrological conditions, and environmental dynamics.
The objective of this project is to develop individual based model to simulate fire occurrences in tropical peatland ecosystem in Sumatra Indonesia. This ongoing study develops the PeatFire model using NetLogo involving three main agents, i.e., patches, households, and fire entities. Each agent behaves uniquely according to their state variable, environment, and interaction between agents. Our first version of PeatFire is a promising individual-based model to predict the occurrence of fire in a peatland ecosystem in tropical regions. The model is sensitive to several parameters which can lead to different output of the model. Further insight into those parameters through literature study will be used to improve the model. PeatFire was developed as a part of ToFEWSI project funded by RCUK-DIPI.