People are constantly migrating around the globe. But scientists have long had trouble quantifying how many people are moving and where they are coming from and going to. Part of the problem is that countries vary widely in the amount and quality of data they collect on incoming immigrants; globally, these data are often difficult to compare directly. A report last year by the United Nations aimed to fix that problem by combining all available data on immigrant populations into one comprehensive, harmonized dataset. Now, a new paper just published in Science has taken that dataset and gone a step further by generating more data and visualizing the global flow of people in a new way.

The United Nations dataset included information for 1990, 2000 and 2010. However, the authors of the new study wanted to see how global migration changed at finer timescales. Using similar techniques to those the U.N. used when filling in data gaps, the researchers generated data for 1995 and 2005 as well, giving them four five-year periods.

The new dataset revealed some expected patterns and some surprising ones, says Nikola Sander, a researcher at the Wittgenstein Centre for Demography and Global Human Capital in Vienna and a co-author of the new study. “What we see is that sudden events—for example, the fall of the Iron Curtain in the early nineties, the violent conflicts in Rwanda and in Afghanistan in the early nineties… triggered a large number of moves,” she says. However, the data does not show an overall increase in the number or percentage of immigrants worldwide, despite the widespread idea that immigration has been increasing over the past 20 years.

Sander also wanted to show these new data in a way that would be easy to understand and grasp. “The typical visualization of flow data has been a world map and then ten or 15 black arrows printed on top of it,” she says. “It has a very low visual appeal, and it can only go to a certain level of complexity.” Frustrated, she realized that she had to borrow data visualization ideas from “out of the discipline,” as she puts it, to better represent the findings.

While searching online she came across Circos, a software tool that uses a circular layout to visualize various types of data such as genomes and cancer mutations. Sander realized that a similar plot would also show the intricacies of the migration data. She published the plot above in the Science paper and teamed up with another company, Null2, to code an interactive version, below.

Sander expects to continue analyzing the data. “This is just the very first set of estimates” of the global movement of people stemming from the United Nations dataset, she says. She hopes others will join in the effort to improve the estimates as well; she and her co-author Guy J. Abel are publishing the code they used to generate the 1995 and 2005 datasets. As gaps in the U.N. data are filled and methods for harmonizing the data improve, Sander says, estimates will become increasingly accurate.

Interactive visualization credits: Nikola Sander, Guy J. Abel, Ramon Bauer at the Wittgenstein Centre for Demography and Global Human Capital. Created using Mike Bostock's D3.js library. Larger version here.