Prediction of Migration Paths Using Agent-Based Simulation Modeling: the Case of Syria

Abstract

The Syrian civil war, which started in 2011, has caused a great wave of forced migration in the Middle East. One of the most popular destination points for Syrian refugees has been Turkey. The purpose of this study is to predict the routes of refugees who leave the conflict areas in Syria to reach the refugee camps located in Turkey during a crisis. The study proposes an agent-based model to simulate the decision mechanisms of refugees in a highly uncertain environment. The model employs the A* algorithm to calculate the cost of each available destination point (refugee camp) for each agent, based on their risk preferences and starting locations, and allows agents to choose the camp with the minimum cost as the destination point. By use of the model, we simulate a moment of crisis namely the South Idlib bombardment (from December 2019 to January 2020) under four different scenarios that are generated considering the real-life data gathered from the newspapers of December 2019 and various other sources. The simulation results show the main pathways of Syrian refugees and give insights on the required camp capacities. The results are compared with the gathered secondary data to validate the proposed model.