UFOs spark our imaginations and make us wonder about highly intelligent life beyond Earth. But retired astronaut Scott Kelly cautioned that “so many UFO sightings are probably just playing tricks on us.”

Kelly was part of a panel meeting at NASA headquarters Wednesday to discuss UFOs, which NASA now calls an unidentified weather phenomenon. He described a time while he was flying and the Radar Interception Service (RIO) thought they saw a UFO.

“I remember once flying in a warning zone near the Virginia Beach military district, and my RIO – the man behind the Tomcat – was convinced we were flown by a UFO,” Kelly said. “So I didn’t. We turned. We went to see him. It turned out to be Bart Simpson – a balloon.

In 2016, Kelly broke the record for longest consecutive day in space (340). His record has since been surpassed by NASA astronaut Mark Vande Hey, who spent 353 days aboard the International Space Station.

Recalling his thousands of hours in space, Kelly said Wednesday, “I would often see things in space and think, ‘This is really not behaving like it should.’ And every time I look at it long enough, I realize it’s atmospheric lensing.”

Atmospheric lensing can create optical illusions, such as the sun appearing higher in the sky than it actually is. According to NASA, this could actually stretch daylight along the equator by five minutes. Optical illusions like this are the result of how particles in Earth’s atmosphere bend and manipulate light, which can cause us to see distorted versions of reality.

“What I was looking at was actually flying behind the atmosphere, and the trajectory didn’t look like a straight line because of the changes in the atmosphere,” Kelly said.

Wednesday’s NASA meeting was more of a call to action for better data collection than an opening for people seeking answers to UFO conspiracy theories. Are UFOs real or just a gimmick?

“We have a community of people who firmly believe in the existence of UFOs, and we have a community of people who think it’s ridiculous to pursue this issue,” said David Spergel, chairman of the commission. “I think the scientific way to approach questions is to start by saying ‘We don’t know,’ and then collect the data and try to calibrate your data well.”