Simulation technology to predict refugee crisis

Simulation technology to predict refugee crisis

A new computer simulation of refugees' journeys when they flee major conflicts can correctly predict more than 75% of their destinations, and may become a vital tool for governments and NGOs to contribute to allocate humanitarian resources more effectively and at strategic points.

Researchers at Brunel University London - Diana Suleimenova, Dr David Bell and Dr Derek Groen - from the Department of Computer Science, used publicly available refugee, conflict and geospatial data to construct simulations of refugee movements and their potential destinations for African countries.

The data-driven simulation tool was able to predict at least 75 percent of refugee destinations correctly after the first 12 days for three different recent African conflicts. It also proved to be more accurate than established forecasting techniques (‘naïve predictions') to forecast where, when and how many refugees are likely to arrive, and which camps are likely to become full and need a higher number of resources and assistance. These results were published in Scientific Reports.

The research team created their simulations for Burundian crisis in 2015, which took place after Pierre Nkurunziza attempted to become president for a third term; the Central African Republic (CAR) crisis in 2013, triggered when the Muslim Seleka group overthrew the central government; and the Mali civil war in 2012, which was caused by insurgent groups campaigning for independence of the Azawad region.

The team relied on open data resources to both enable these simulations and validate their accuracy. These sources included refugee registration data from the United Nations High Commissioner for Refugees (UNHCR), conflict data from the Armed Conflict Location and Event Data Project and geographic information from Microsoft Bing Maps.

While not all refugee movements are accurately predicted in these simulations, their approach emulated the key refugee destinations in each of the three conflicts, thus it can be re-applied to simulate other conflict situations reported on by the UNHCR.

For instance, in Burundi, the simulation correctly predicted the largest inflows in Nyarugusu, Mahama and Nakivale throughout the conflict's early stages. Meanwhile, the simulation correctly reproduced the growth pattern in East camp of Cameroon, as well as the stagnation of refugee influx into Chad's camps. In Mali, the simulation accurately predicted trends in the data for both Mbera and Abala, which put together account for around three-quarters of the refugee population.

The researchers used a new-agent based modelling programme named Free, which was revealed to the public with the publication of their paper. Although agent-based modelling has been used more widely to study population movements, and has become a prominent method to explain migration patters, this is the first time it has been used to predict the destinations of refugees fleeing conflicts in the African continent.

Suleimenova, Bell and Groen explain in Scientific Reports that their simulation is not directly tailored to these conflicts, but a ‘generalised simulation development approach' which can forecast the distribution of refugee arrivals across camps, given a particular conflict scenario and a total number of expected refugees.

This simulation development approach allow organisations to quickly develop simulations when a conflict occurs, and enables them to investigate the effect of border closures between countries and forced redirection of refugees across camps. It also serves of assistance to define procedures for collecting data and validating simulation results, aspects which are usually not covered when presenting a simulation model on its own.

According to the authors, Accurate predictions can help save refugees' lives, as they help governments and NGOs to correctly allocate humanitarian resources to refugee camps, before the (often malnourished or injured) refugees themselves have arrived. To our knowledge, we are the first to attempt such predictions across multiple major conflicts using a single simulation approach."

The authors also urge greater investment in the collection of data during conflicts and they explain what this is important and what it's hard to get. "Empirical data collection during these conflicts is very challenging, in part due to the nature of the environment and in part due to the severe and structural funding shortages of UNHCR emergency response missions. Both CAR and Burundi are among the most underfunded UNHCR refugee response operations, with funding shortages of respectively 76 and 62%".

With record levels of 22.5 million refugees on a global scale, "more funding for these operations is bound to save human lives, and will have the side benefit of providing more empirical data – enabling the validation of more detailed prediction models."

The research group aims at collaborating with humanitarian organisations, adapting their technology to help specific humanitarian efforts, and to further reduce the time of development by automating the creation of these simulations.

'A generalized simulation development approach for predicting refugee movements' by Diana Suleimenova, David Bell and Derek Groen (Department of Computer Science, Brunel University London) is published in Scientific Reports.