The Ottawa County Health Department is developing a beach water equation to predict when E. coli levels are dangerously high.
Click
to enlarge
In past summers, the county took water samples to a lab in Grand Rapids
for testing, and didn't get results until the following day. By then, levels
of E. coli can drastically change, causing officials to close beaches even
with low E. coli counts and keep them open with high levels of the bacterium.
"We want to look at the real-time environmental data," said Adam London, the Health Department's environmental health manager, "which would of course be more helpful than looking at yesterday's information."
A model based on statistics would allow the county to react to changes on a daily basis rather than the typical once-a-week sampling.
In order to develop a predictive model, London said the county is compiling a record of how factors like rainfall, water temperature and the presence of seagulls affect the E. coli count. The Health Department does some of its own data collection, but is also pulling together statistics kept by other sources, such as local airports and the National Weather Service.
London isn't sure when he'll have sufficient data to build a usable equation.
"It's possible that we could have something as soon as this summer, and certainly within the next couple summers," he said. "It may turn out that we just don't have enough data yet."
Several Great Lakes areas are now either using predictive modeling or looking into it, but it's still a new trend.
Mark Phister, the associate director of environmental health in Lake County, Ill., has been using a predictive model at four beaches with positive results.
He said at one of the Lake Michigan beaches in Lake County, the predictive modeling system resulted in 95 percent accuracy — where as the regular sampling would have been correct only 77 percent of the time. At another beach, regular sampling would have been wrong 100 percent of the time over the course of the summer, because the E. coli levels were continually changing.
Even though Lake County has improved accuracy on four beaches, Phister said predictive modeling wouldn't work well on all 13 area beaches. He said a predictive model wouldn't be as effective for beaches with either very little E. coli or a lot of the bacterium.
"You need to have both low and high days through the season to create a (predictive) model," Phister said.
High levels of E. coli indicates dangerous levels of pathogens, which can lead to human infections for swimmers, Phister said.
Holland State Park beach has only been closed once in the last six years, according to London, and less than 3 percent of all the beach sampling in the county come back with dangerous E. coli levels, or more than 300 E. coli per 100 milliliters of water.
During the first weekend of the 2007 Coast Guard Festival in late July, when the local beach is usually bustling, the Grand Haven State Park and City Beach were closed on a Friday and Saturday because of high E. coli levels. The only other 2007 beach closing in the county was for two days at Tunnel Park in Park Township.
"Our challenge is that our
water quality has been so good — if you want to call it a challenge,"
London said. "Even though we're limited in that we have a moderate
level, we still have seen some trends."