A dragonfly in August. (Joel Saget/AFP)

A glance at weather radar in the Midwest on Tuesday might have had you reaching for the umbrella. The only problem? It was a clear, sunny day.

The likely culprit couldn’t have been rain. It probably wasn’t even meteorological. There’s a strong chance that these signatures came from dragonflies.

Social media video emerged showing swarms of dragonflies descending across the eastern third of the nation. From Ohio and Indiana to New Jersey, Maryland and Virginia, witnesses have reported the migrating dragonflies. And Doppler weather radar probably has detected the bugs, as well.

Geeky radar analysis cracks the buggy code

Take a look at this plot from the National Weather Service radar in Pittsburgh. There’s “ground clutter” but also a few patches of darker blue to the north. Ordinarily, that might suggest drizzle or light rain, but it was a virtually clear day.


It looks like rain, but it actually might be a cloud of dragonflies that was detected by radar over Pennsylvania Tuesday. (NOAA/GR2 Analyst/Matthew Cappucci)

When we switch on over another radar product known as correlation coefficient, which helps us differentiate shapes in the sky of different sizes, we see more jagged, nonuniform shapes — highly unusual if we were talking raindrops.

There is yet a third radar view that helps gather information about what’s lurking in the air known as “differential reflectivity.” This allows us to evaluate the shapes of the objects intercepted by the radar beam.

Differential reflectivity reports data back as a number between negative-7.9 and +7.9. If an object is as tall as it is wide, it will reflect a zero signal. But for objects wider than they are tall, that number skews positive. Within the cloud of radar returns over Ohio Tuesday evening, differential reflectivity values ranged between positive 5 and 6. That means we’re likely talking something with a decent width compared with its height. Like, say, a swarm of dragonflies.


Differential reflectivity sees a lot of high values, plotted in pink. That represents a slurry of objects wider than they are tall. (NOAA/GR2 Analyst/Matthew Cappucci)

Robert LaPlante is the science and operations officer at the National Weather Service in Cleveland. He says there is a good chance that dragonflies contributed to the radar returns but weren’t solely responsible. “Record high temperatures likely allowed them to fly higher,” he said. That makes it easier for radar to detect them.

Did dragonflies make it to Washington?

Over the next day, more dragonflies appeared, as did more wonky radar returns. By the time Thursday rolled around, the critters were meandering over the Appalachians. Or were they?

Thursday’s radar in the District showed bizarre high-reflectivity strips before any afternoon showers developed. These were horizontal convective rolls.

Convective rolls form parallel to the wind direction in the lowest levels of the atmosphere. They’re bands of alternating rising and sinking air. As the ground heats up, the vertical depth of those rolls grows, as does their width.

We saw these on radar Thursday afternoon in the District before storms fired. See if you can find a couple of them:


Horizontal convective rolls turn over the DMV on Thursday. (NOAA/GR2 Analyst/Matthew Cappucci)

There weren’t storms around yet. Sure, a few clouds formed along the enhanced convergence between rolls, but clouds are just air and moisture.


Satellite shows that, while we did have "cloud streets," there were no strong storms at the time of the convective rolls. (NOAA/RAMMB)

Now air and moisture itself isn’t something we can “see” on National Weather Service radar. To resolve convective rolls, something has to be embedded within them. That could be where the dragonflies come in.

We know there were dragonflies nearby. Reports from Maryland and Virginia showed they were just north and west of the nation’s capital Wednesday night and Thursday:

Now the radar Thursday afternoon was busy searching for thunderstorms. But it did show lots of stuff elsewhere with similar characteristics to how the dragonflies showed up in the Midwest. The thunderstorms can also help us to probe for possible dragonfly signatures.

Around this time, we saw some pretty impressive aberrations associated with outflow boundaries — the leading edges of thunderstorm exhaust — propagating through the D.C. region. Something was caught up along the outflow. Dragonflies or insects might have been responsible. But we are unaware of dragonfly sightings in D.C.


A highly resolved outflow boundary plunges south across D.C. on Thursday. What was it carrying that made it so visible to the radar? (NOAA/GR2 Analyst/Matthew Cappucci)

It’s entirely plausible that we did get a bunch of dragonflies in the Washington area as radar was suggestive of bugs, but we haven’t seen reports to confirm. They may have just been suspended a little too high in the air for people to see.

If you happened to be on a flight departing National Airport or Dulles between 3 and 6 p.m. Thursday and you spotted anything, let us know!