TOPEKA — The data suggest restricting saltwater injection in southwest Kansas might have reduced the number of earthquakes, but it isn’t clear how long it could take for quakes to taper off — or whether Kansans need to prepare for the possibility of a more dangerous incident.

Justin Rubinstein, a seismologist with the U.S. Geological Survey, spoke to representatives from the Kansas Geological Survey and the Kansas Corporation Commission on Wednesday afternoon in Lawrence.

The number of earthquakes has fallen since the KCC issued an order in March restricting the amount of saltwater that could be injected in Sumner and Harper counties, Rubinstein said, but there isn’t yet enough data to pinpoint the order as the cause. Researchers recorded 372 earthquakes in the 164 days before the order, and 203 earthquakes in the 164 days since, he said. The day count was just the number of days since the order, compared to the equivalent period before it.

Multiple earthquakes often happen in a short period of time, so it could be that natural variation has caused a period with less shaking, Rubinstein said. On the other hand, the timing of the decrease suggests the order could have been a factor, he said.

“There’s a pretty significant decrease,” he said. “We all hope the regulation is reducing the earthquake rate, but I don’t think we can hang our hats on that.”

The USGS is working on understanding the earthquakes in southern Kansas and northern Oklahoma and coming up with a risk forecast, Rubinstein said. Even if both states decided to stop allowing injection wells, it could take anywhere from a few weeks to more than 10 years for the induced earthquakes to stop, he said.

In Kansas, research suggests human-induced earthquakes are coming from injecting large amounts of saltwater into formations where it hadn’t been done before, rather than from hydraulic fracturing or other methods of recovering oil and gas. Drilling typically brings up large amounts of highly saline water, and companies have typically found it more economical to inject the water back into the ground instead of attempting to treat for another form or disposal or reuse.

So far, most earthquakes caused by injection in Kansas have been small. The largest earthquake linked to injection was a magnitude 5.6 quake in Prague, Okla., in 2011. That might not be an upper bound, however, because other human activities have caused larger earthquakes, Rubinstein said. A magnitude 7.3 earthquake was linked to natural gas drilling in Uzbekistan, and the magnitude 7.9 earthquake that killed approximately 80,000 people in China’s Sichuan province might have been caused by building a dam in the area.

The ratio of small earthquakes to larger ones in Kansas hasn’t changed much in the last six months, Rubinstein said, because their numbers have gone down at the same rate. One as-yet-unexplained change is earthquakes are increasingly happening outside of five small areas where the shaking was concentrated before, he said.

“There’s no clear smoking gun as to why these earthquakes are happening” outside the areas of concern, he said.

Some geologic formations can handle high levels of injection without earthquakes, Rubinstein said, and it isn’t clear whether there is a certain threshold where injection could continue in Kansas without raising the risk of earthquakes. Companies have injected large amounts of water into the ground for years in some parts of the country, he said, but what happened over one formation isn’t a good indicator of what will happen over another.