A pair of NASA scientists have an unconventional hobby—searching through atmospheric data to find the rare moment when faces swirl up in their data.
By day, NASA scientists Lawrence Coy and Steven Pawson develop complex data assimilation and reanalysis models of Earth’s atmosphere at Goddard Space Flight Center. But when they need a break, they indulge in an unconventional hobby—searching for ghostly faces swirling in their data.
A few of their favorite finds after scouring approximately 40 years of data are shown in the images on this page. They spotted all of these “faces” while examining visualizations of wind and temperature data from a reanalysis called the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2). The reanalysis simulates atmospheric conditions on a global scale by incorporating weather data from satellites and several other sources.
“The best time to find faces is in the winter in the polar regions of the northern hemisphere,” Pawson said. “That’s when the dark conditions of the ‘polar night’ lead to a ring of westerly winds in the stratosphere that atmospheric scientists have long called the stratospheric polar vortex or circumpolar vortex.”
For much of the winter, the polar vortex spins high in the stratosphere. However, about once per winter—usually in January or February—something happens that disrupts this circumpolar flow. This causes those westerly winds to weaken and temperatures to increase over the pole. This can even cause the winds to change direction and the stratospheric polar vortex to split into sections, Coy explained. “It’s often during or after these events—sometimes called sudden stratospheric warming events—that we’ll find a face.”
In all cases, the ephemeral facial features are caused by different arrangements of low and high potential vorticity—a quantity that describes how air masses are spinning. In these plots, areas of high potential vorticity appear orange and have counterclockwise circulation; areas of low potential vorticity appear purple and have clockwise circulation. (German scientist Hans Ertel first published on potential vorticity in 1942).
In the menacing face at the top of the page, a split in the vortex following a sudden stratospheric warming event left two areas of high potential vorticity serving as eyes, while an area of low potential vorticity formed the mouth. The face was seen in potential vorticity data from January 25, 1982, at a height of roughly 30 kilometers (18 miles), roughly the middle of the stratosphere.
In the set of three, the face on the left also had a split vortex, with eyes of high potential vorticity. A more common pattern can be seen in the two smiling images to the right. In both cases, the left eye is made up of a low potential vorticity area, the right eye of a high potential vorticity area, and the mouth by connecting strands of both.
“Steven had the idea of finding at least one of these faces for each winter,” Coy said. “We already have a collection of more than 40 of these, and I bet we will continue looking in future years.”
When Coy and Pawson find faces in atmospheric data, they’re engaging in face pareidolia—the human tendency of seeing faces in everyday objects. Psychological research suggests the habit likely flows from the human brain’s exceptional facial recognition skills, something that even the youngest of babies can do well.
“Finding these never gets old, and it’s a fun way to get people to think about the stratosphere and our data assimilation models,” said Pawson. “It’s also a good way to highlight year-to-year differences in the stratosphere.”
NASA Earth Observatory images by Joshua Stevens, using MERRA-2 data courtesy of Lawrence Coy/GSFC and the Global Modeling and Assimilation Office at NASA GSFC.