To better understand the complexity of migration processes, there is a need for more and better data and knowledge on the flows, the drivers, and the effects of migration. When it comes to drivers and effects of migration that have a visible impact on the earth’s surface, satellite-based earth observation is increasingly able to provide amazing insights at large scales. However, not all migration drivers and effects can be observed from space. Migration is also a social phenomenon, and in this domain, it depends on social and economic processes and relationships, which are increasingly being conducted and discussed in the virtual space.

To incorporate this aspect into migration research, MIGRAWARE makes use of additional data sources, which can provide valuable insights into migrants’ motivations, interests, and opinions. Among social media, Twitter is a particularly powerful source of information. While not the most widely used social media platform in Western Africa, it stands out due to its public nature and its geotagging feature. It is this geotagging which allows the linking of virtual spaces with the geographical space and supports a wide range of applications in migration research.


Broadly speaking, there are two major ways in which social media is being used within MIGRAWARE.

Firstly, the geotagging feature allows us to perform mobility detection: Users which regularly choose to geotag their tweets leave traces in space and time, which enables us to identify migration pathways at a large scale. By aggregating traces of thousands of users, we can answer questions like: Which cities are national migration hubs? Which cities have the strongest connections abroad, and to which countries? This aggregation at the level of cities or regions further ensures the anonymity of the users.

Figure 1: Migration Flows in West Africa. Caption: “Migration flows in West Africa from 2015 to 2019 as detected in Twitter users‘ geolocations 2015-2019 (Data: Twitter Inc.; Basemap: World Settlement Footprint, DLR)

Secondly, within the texts produced by users, topics can be detected. These might be general topics such as politics, traffic, or religion. They might be recurrent events such as elections, the UEFA Champions League, or Big Brother Naija. And they might also be specific one-time events, such as natural disasters. In this context, these data complement earth observation. While satellites can observe and measure the impact of large-scale events, such as floods, in a quantitative manner, social media allows us to understand how they are received by the people.

Figure 2: Flood Tweets, Caption: “The 2017 Benue state flooding through the lenses of Twitter and remote sensing (Data: Twitter Inc.; Basemap: Sentinel-2 Mosaic; Watermask: Sentinel-1; Urban Areas: World Settlement Footprint, DLR)”

Benefits and Limitations

In the face of the massive transformations caused by urbanization and climate change, it might seem odd to seek answers from individuals’ views on Twitter. But because migration decisions are ultimately decisions by individuals, the aggregation of these subjective perspectives might be the most valuable information that social media can bring to the migration research table. To give some examples: Peoples’ sentiment about disaster relief, the social status of a neighborhood, or the impact of ethnic conflict on people’s sense of security are all deeply subjective. In understanding migration, information on such subjective views can be just as important as the objective numbers.

If we compare social media to interview-based surveys, which might provide similar information, social media excels in availability and timeliness. MIGRAWARE is ambitious in its scale, covering long-time processes in large areas like countries. Social media can keep pace with other big-data sources in providing information at international scale.

However, unlike surveys, social media gives us analysts no influence over what participants (users) choose to talk about. We also cannot choose who participates in the generation of social media data. And that is probably its greatest limitation. Although statistics are hard to come by, it is very likely that social media are predominantly being used by young and urban citizens which are rarely from the lowest stratum of society. Refugees, our data suggests, are less likely to use Twitter extensively. Therefore, statistics generated from social media should not be expected to generalize to the whole population. Instead, they allow us to understand particular demographics: Young Africans, urban people, tourists, students, and expats. Demographics which play important roles in West African economies. By using social media, we can understand what interests them, what connects them to their places of origin, what makes them migrate, and what makes them stay.


For providing such insights at large scale, social media is a valuable complement to traditional surveys and the more regular sources of geodata such as earth observation. Where surveys give deep migration-specific insights and remote sensing informs about changes on the land surface, social media gives a large scale understanding of societal views and perspectives. By using various data sources in conjunction, we can assess the complex spectrum of migration patterns, their drivers, and their effects.

We presented our approach at the ESA Living Planet Symposium 2022. You can find the poster at the following link:


Johannes Mast

Research Scientist/DLR, Germany

Dr. Martha Sapena-Moll

Research Scientist/DLR, Germany

Prof. Hannes Taubenböck

Senior Scientist/DLR, Germany

Leave a Reply

Your email address will not be published.