Darfur Violence Monitoring

For decades Darfur has been at the heart of frequent conflict and civil unrest. In 2003, two armed groups, the Sudan Liberation Army (SLA) and the Justice and Equality Movement (JEM), began fighting against the Sudanese goverment for control of the region. They moved through rural areas, razing towns and villages to the ground by burning buildings and murdering civilians.

In 2019, these factions were ousted from the region, with the Sudanese government signing a peace agreement in the following year. This agreement saw the government accepting responsibility for the well-being of communities and maintaining peace in Darfur.

To the dismay of the people living in the area, international peace keepers withdrew from Darfur. With nobody stepping in to replace them, violence returned.

Many blame the returning conflict on the paramilitary groups the government recruited as rapid support forces to quell the threat of civil unrest. However, the government has failed to comment on this possibility.

This sadly familiar conflict has stalled any aid being delivered to the 1.2 million people who are most in need.

Many families were displaced after their villages and homes were burned to the ground and sought safety in camps. According to some reports, the capital of West Darfur, Algenina, has become a huge IDP camp, with extremely harsh living conditions and inadequate facilities. For the majority at Algenina, this is the second time they have been displaced because of violence. In some cases the camps have become so uninhabitable that families have started to return home, despite the uncertainty.

The withdrawal of the UN and other aid organisations and NGOs has made people fearful for the future, with many more believing that long lasting peace is still very far off.

Remote sensing and satellite imagery analysis have always been valuable tools for monitoring and measuring the extent of violence in fragile conflict affected states. However, there have always been drawbacks and barriers that make it difficult to keep up with the speed of change in these zones.

Imagery availability and processing power have always limited the capability to monitor wide areas and provide continuous data that mirrors the rapidly evolving situations on the ground. New technologies have enabled us to fulfil these demands and carry out rapid imagery processing and analysis which, previously, would have taken months.

One source of data considered for the task of detecting and monitoring burned settlements was NASA FIRMS (Fire Information for Resource Management System). This is a brilliant resource that uses satellite observations from the MODIS and VIIRS sensors to detect thermal anomalies and active fires globally. However, the resolution of this data makes is too large, making it insufficient to monitor the burning of individual homes in rural Sudan and gather accurate estimates for impacted population.

As you can see from the image below, one pixel covers a significant area in relation to the settlement, making it difficult to determine if the fire occured within, on the border, or outside the small town.

Analysing burnt surfaces using remote sensing as a proxy indicator for violence in Darfur, as setting fire to homes is unfortunately an extremely common practice, allows us to monitor the spread of conflict on a weekly basis. Filtering out uninhabited areas using population and settlement data also means that wild or agricultural fires do not distort the results.

With this data we compare one week to the week before, indicating if there has been any significant change in the number of burnt surfaces detected and giving us a temporal picture of events.

Furthermore, recombining the outputs with the population data provides an estimate for the number of people potentially impacted by the event. This information can then be used, in conjuction with aid workers on the ground, to better allocate resources and inform those managing IDP camps with statistics on new arrivals and potentially their places of origin, as well as ensuring people are directed to the most suitable camps.

Below is a map that demonstrates one week of the results in December 2021. It shows the estimated affected population from identified burnt settlements with proportional symbols. The results indicate that communities in West Darfur and Southern Darfur have been especially impacted during this particular week, potentially due to violence.

In Kereneik in West Darfur for example, we can see that an estimated 117 people have been exposed to significant levels of burning which likely constitutes to their homes being destroyed by violent militias in the area. Although we cannot speculate on the wellbeing or conditions of these people from satellite imagery and remote sensing alone, the destruction of these homes will indesputably create pressure and termoil for the community, especially if there was any loss of life.

The next stage of this monitoring process is to use these results to assist better targeted and higher quality investigations, both from above and on the ground. Taskings for higher resolution satellite imagery can be carried out and with greater temporal accuracy. This will result in better images closer to when the event occurred.

With this, more focused and accurate analysis will be carried out, and decisions about where to focus aid workers resources on the ground can be made.

There is huge potential for this project to be a useful resource for aid workers delivering help to people in Darfur who have experienced horrific conflict.

The aims of the project are to ensure, with speed and accuracy, that no affected community, no matter how remote are missed by those organisations trying to help them. The faster the analysis means the faster aid can be designated and delivered.

The success of this methodolgy will hopefully demonstrate its replicability in other places and not be limited to Darfur. Unfortunately there are many conflict affected states around the world that could potentially benefit from this variety of monitoring and analysis.