Sinkholes are a natural phenomenon nobody can predict–but researchers at the University of Central Florida are looking to change that.

A research team at UCF created a mini-sinkhole simulator to help understand what causes them.

“We are looking for patterns that may help us predict when a sinkhole will occur,” said Engineering professor Manoj Chopra. “Water plays a very important role in sinkhole formation, and we’re trying to determine if we can find patterns in water levels underground that can help us identify potential sinkholes early.”

A sinkhole occurs when soil collapses into large holes caused by flowing water eroding underground layers of rocks. When the rock wears away, the soil goes through the cracks and creates underground cavities. The cavities grow until they can no longer support the weight above, eventually causing the ground above to collapse.

Mohamed Alrowaimi, one of Chopra’s Ph.D. students, has spent over a year working to scale the experiment correctly to translate to real-life use.

“It’s research in progress,” Chopra said. “But we know there is a connection between water and the sinkhole process. Rainfall affects the levels of water underground in the aquifer so we are examining that too. And temperature may also play a role. Perhaps water levels change or the temperature changes some time before a sinkhole occurs. If we can learn to read those changes, we could have a prediction model.”

Sinkholes have continuously affected Florida, with one of the largest sinkholes being right down the road from UCF in Winter Park at 125 feet deep and 350 feet wide.

“It can happen anywhere, anytime,” Chopra said. “That’s a little scary and we want to help prevent tragedies if we can.”