IT'S D-DAY MINUS 1, AND BEN DAVIS BEGINS HIS MORNING with a 5:45 am workout. Two days earlier it was an 800 meter run, 30 kettleball swings, and 30 pull-ups, five rounds of each. Today it's ten snatches with maximum weight. He's doing Crossfit training, a popular regimen promoted for "forging elite fitness," especially for first responders.

Fitness matters for Davis. He's a first responder, a crew chief for the lifeguards watching the south end of the 10-mile sand beach at Ocean City, Maryland. A four-year veteran of the Coast Guard and a nine-year veteran of the Ocean City Beach Patrol, Davis has brown, sun-streaked hair and a solid build he has to keep in shape. At 31 years old, he's got to be ready to run fast on wet sand, swim hard through heavy swells, and haul drowning bodies off the bottom of the ocean.

Fitness comes easier for Billy DePaola. He's a wiry, curly-haired first timer, fresh from college and four years of lacrosse and soccer. For DePaola, D-Day minus 1 begins with lifeguard drills on the beach. He carries injured swimmers out of the breaking waves. He practices sprinting, diving, and swimming out through the swells, first with his small red buoy, then with a landline, and finally with a long, yellow rescue surfboard.

It's final exam day for DePaola and 33 other rookies working with the Ocean City Beach Patrol. For them, the beach drills are more fun than work, letting them race from one drill station to the next, then plunge into the surf, compete in teams, pull out practice victims, and celebrate with wisecracks and fist bumps. If they pass, they graduate, go off probation, and get a pay raise.

It feels like any last practice before any big game, but it's also like a war game of sorts. At each station sergeants and lieutenants are giving orders. Wherever they go veterans are walking around, arms folded, yelling "Keep your eyes on the ocean." Tomorrow is the start of the three-day July 4th weekend, D-Day for the Beach Patrol. This holiday always brings a massive invasion of beachgoers down to one of the East Coast's busiest beaches. That means tens of thousands of people will be plunging into the waves, many of them with weak swimming skills, and most with little understanding of what can happen when those waves slide back out to sea.

Beachgoers worry about sharks and lightning, but lifeguards worry about rip currents. Along most American beaches, they are the big killers. On any shore with breaking waves, channels of seaward-flowing water can suddenly open up and sweep swimmers and waders out past the breakers into deeper waters well beyond the beach. For most victims, swimming back against the current proves futile, leaving them exhausted and swept further out to sea. Most drownings on American beaches are rip current drownings. And ninety percent of the rescues by lifeguards at Ocean City are rip current rescues.

"Rotate! Watch your rotation," barks a sandy-haired man who's watching the beach action from beside a white jeep. Butch Arbin, captain of the Beach Patrol and a 37-year veteran, set up these final maneuvers and he wants them done right — all of them. "Station leaders," he yells through a loudspeaker, "Keep your people moving."

He wants all his guards, both veterans and rookies, ready for D-Day. He remembers what happened this time last year.

TONY DALRYMPLE AND VARJOLA NELKO thought they would never find a hotel in Ocean City. This was the fall of 2006 and the two scientists weren't looking for a room. What they wanted was a roof with a view. There they would set up four video cameras that would take pictures of the beach every day of the year. All the hotel managers hesitated. They had to check with ownership and ownership was often a corporation located in another town that saw little to gain from letting two scientists loose on a rooftop. It took dozens of phone calls and treks to 15 hotels before the manager of one hotel, the Grand Stowaway Hotel, said yes. What was the argument that worked? The cameras, the scientists said, could help save lives.

What would save lives, said Dalrymple, would be an accurate forecast for dangerous rip currents. Tony Dalrymple is lean, laconic, and white-haired, a coastal engineer at Johns Hopkins University who's published widely on wave dynamics and their effects on coastal structures like beaches, breakwaters, jetties, and derricks. He's been consulted on tsunamis in Thailand, hurricanes in New Orleans, and giant surfing waves in Hawaii. He's also kept up a long-standing interest in a less-famous subject: rip currents. When a new grad student, Varjola Nelko, arrived from Albania and Turkey in search of a Ph.D. topic, he gave her a tough one. They would work up a new way of predicting when rip currents are likely to show up.

How could they sharpen rip current forecasts using cameras? By capturing actual photos of rip currents as they formed. Then by correlating those rips in the photos with weather data about wind and waves, the natural forces that drive these killer currents. That, at least, was the approach he wanted to try.

Observation, analysis, and predictive modeling, that's a classic progression, but it's not the way scientists usually forecast rip currents. Ever since the late 1980s scientists have focused on rip current rescues, not on actual rip currents observed in action. With their cameras, Dalrymple and Nelko were trying to change the ground rules of the forecasting game.

HISTORY'S MOST FAMOUS D-DAY WAS the large beach invasion that happened 65 years ago this summer. On June 6, 1944, more than 150,000 American, British, and Canadian troops waded ashore through a three-foot surf, launching the Normandy Invasion that began the liberation of Europe and helped end World War II. One little-known key to the success of that long-ago assault was a top-secret technique, newly invented, that helped forecast the waves and surf that would be hitting the target beaches that day. On one of the bloodiest days of the war, that forecast helped save lives.

That technique, created by an Austrian named Walter Munk and a Norwegian named Harald Sverdrup, is no longer secret. Its basic claims about wind-created waves form the conceptual starting points for contemporary wave forecasting — and now for rip current forecasting. A wave, they said, starts with a wind somewhere in the world scraping along a stretch of ocean. They called that expanse of wind-stroked water a "fetch," and they theorized that the size and speed and direction of a wave depends on how long and wide the fetch is, how long the wind blows across it, and how strong the wind is. Six decades after the Normandy Invasion, any surfer can now go to websites like Surfline.com or WaveWatch.com and find an up-to-date forecast for the wave and surf conditions they'll probably see that day when they wade out with their boards from their local beach.

At Ocean City the beach patrol gets its wind and wave forecasts from the National Weather Service, specifically from the Wakefield, Virginia office where forecasters work up predictions using data from offshore buoys and output from the powerful numerical models that form the heart of contemporary wave forecasting. Since the buoys and the models don't always match, the forecasters turn to other statistical tools and, finally, to their own judgments. "It's not just: ‘Here's the model, it's saying this.' And we just go with it," says John Billet, science operations officer at the Center. "We make adjustments." The final forecast is always made by people.

Standing on the shore at Ocean City, lifeguard captain Butch Arbin can now find out what kind of swells are heading towards him, even if they're all the way from a "fetch" near Europe or Africa. Like most lifeguards, Arbin watches these wind and wave forecasts closely, but warily. What he really needs for his D-Day weekend is a way to turn those reliable wave forecasts into useable rip current forecasts.

IN THE 1980S, A METEOROLOGIST in Miami began asking medical examiners and beach patrols for reports about drowning deaths and rip current rescues along the beaches of southeast Florida. Jim Lushine, a forecaster with the National Weather Service, collected these reports and then looked for correlations between rip current rescues and weather conditions. What combinations of wind and wave conditions matched up with high rescue days, with moderate rescue days, with low rescue days? When lifeguards told him to focus on wind events, he pulled ten years of wind records for Miami Beach and found a robust correlation between strong onshore winds and high numbers of rip current rescues. Ebbing tides also correlated strongly. Focusing on winds and tides, he built the country's first system for forecasting rip current dangers. The Miami office began issuing daily warnings about high risk, moderate risk, and low risk days.

Long-distance ocean swells, surprisingly, got little play in his early predictive model for rip currents, apparently for geographic reasons. Long-running swells that crossed the ocean from Africa or Europe and headed towards Florida were largely blocked by the shallow waters of the Bahama Islands. Surfers who track big swells call this effect the island shadow. Later scientists who adapted the Lushine scale to other beaches gave greater numerical weight to wave heights and to long-period swells originating in distant locations. The game of mixing and matching weather factors with rip rescues was on.

And the game continues today. To work up a rip current prediction for Ocean City, John Billet and the other forecasters at the Wakefield Center use a later version of the system Lushine pioneered in Florida. They call their tool MALURCS, short for the Mid-Atlantic Lushine Rip Current Scale, and in their version of Lushine's scale ocean swells — their size, timing, and direction — are hugely important. "They are definitely the biggest input," says Billet. There are, after all, no nearby islands standing between Ocean City and the other side of the Atlantic.

The Wakefield forecasters are also using another tool: the eyeballs of Ocean City lifeguards. As Ben Davis, the veteran lifeguard, takes his place atop his chair tower, he scans the ocean in front of him, looking for signs of rip currents down near the Ocean City pier and inlet. Up the beach at 120th Street, Billy DePaola, the rookie, does the same. By 10:00 am more than 90 pairs of eyeballs are reading the ocean, looking for rips. Sergeants in charge of each section report their reads back to Beach Patrol headquarters, and the dispatcher on duty faxes them to the National Weather Service. The rip current threat is forecast as low or moderate or high as calculated by lifeguards — not by a Lushine predictive scale. The fax goes out three times every day, and the last one includes another number: the day's total for rip current rescues.

Eyes in the skies, these cameras watch the beach from the rooftop at the Grand Stowaway Hotel at 21st street in Ocean City. They send their photographs over the internet to desktop computers at Johns Hopkins University, where Robert "Tony" Dalrymple and Varjola Nelko analyze the images for evidence of how dangerous rip currents form. To see what these cameras see, go to their website: http://www.ce.jhu.edu/oceancity/. Can you pick out the rip currents? Credit: Michael W. Fincham.

These rip current reads and rescue stats will play a key role, not for today's forecasts or tomorrow's, but for next year's. In the off-season the Wakefield forecasters will verify their forecasts from the past year by comparing them against all these reports from the field. With this kind of groundtruthing, they can adjust their weighting values and sharpen their predictive power for the next year. It's a kind of off-season tune-up for their forecasting engine.

The result, in theory, should be rip current forecasts that grow more precise year by year. The official forecasts, in practice, seem to be more helpful for occasional beachgoers as more media outlets every year carry the forecasts. If you're headed for the beach you can turn on the radio or go online and get the weather forecast. If you're going in the water, you can also get the rip current forecast. If it says "moderate risk" of rip currents, you might want to stay on the beach. If it says "high risk," you might want to stay home.

For now the official forecasts are largely ignored by Captain Burch Arbin and his lifeguards. When it comes to rip currents, they'd rather trust the readings they make themselves, watching the waves from chair towers eight feet above the beach.

WHEN TONY DALRYMPLE AND VARJOLA Nelko finally found an Ocean City rooftop for their cameras, they began reading the ocean from 14 stories above the boardwalk.

Their high-angle perch in the sky lets them test a radically different approach to rip current forecasting. Instead of collecting records of rip current rescues — as Lushine and his followers have been doing for two decades — they're collecting visual records of the rip currents themselves.

Their field work was simple in concept: put video cameras on a high rooftop and then photograph the beach and surf zone where rip currents form. The cameras take photographs at 3 frames a second for 10 minutes, then average them together and shoot them straight to Dalrymple and Nelko. From the rooftop cameras to a hotel computer to the internet to a desktop computer, the beach pictures fly from Ocean City to two university offices in Baltimore. Much like lifeguards sitting on their beach chairs, the scientists can sit in their office, without sunglasses or sunblock, and try reading the waves for rip currents.

From behind a desk strewn with books and assorted spiral-bound reports, Dalrymple swivels his chair and pulls up to a large-screen computer against the wall. He taps the keyboard. Here in this long office lined with bookshelves, with a guitar case sitting on the floor, with a window looking out onto the green, quiet campus of Johns Hopkins University, we are suddenly back in Ocean City. We're looking down on a beach busy with sunbathers sprawled on blankets and sitting under umbrellas. Perhaps they're having a good time. Beyond is the surf zone: shallow water, then lines of breaking waves, then deeper, dangerous waters. Only a few waders are actually in the water. Perhaps they know what they're doing. One wishes them well. It's a godlike view from up here in the sky.

Two ways to read the rips: from on high, from on the beach. Scientists Tony Dalrymple (above) and Varjola Nelko (below) work at Johns Hopkins University using high-angle images from their rooftop cameras. With funding from Maryland Sea Grant, they are analyzing how rip currents form and are devising a formula for accurately forecasting their arrival. A lifeguard (right) works from a tower chair along the Ocean City beach using his training and experience to spot rip currents and rescue swimmers from their clutches. Credit: Michael W. Fincham.

A couple more taps and Dalrymple brings up another image, a freeze-frame that averages together 10 minutes of pictures. Think of a double exposure multiplied 1800 times. The image is slightly blurry like an X-ray, and like a doctor advising a patient, Dalrymple begins diagnosing the big picture. "All the breaking is occurring right here," he says, pointing to a smudged line that compresses together hundreds of surf breaks. Waves break in shallow waters, and that smudged line of breaks tells him that's where the sandbars are.

"Then the water gets deeper again," he says, pointing to a dark strip of water trapped in a trough between the sandbar and the shore, water that has to run out to sea again somewhere. "Probably a low spot right there," he says, his finger on a dark gap in the breaker line. "Not necessarily a rip current there, but it's more likely to be there than anywhere else."

None of those tiny humans wading down there in the surf can hear Dalrymple's diagnosis of danger delivered from an office in Baltimore. "A well-trained lifeguard would know that there are likely rips there," says Dalrymple, "but it is really obvious from here." One hopes a lifeguard is alert. The view from the 14th floor may be godlike, but it's the vision of a distant, powerless deity.

Accurate forecasts may save lives in some future summers — that's the hope at least — but there's a whiff of scientific hubris around any project trying to predict natural forces as sporadic as rip currents. Following on the success of wave forecasting, however, scientists at more than a dozen universities are now working with wave-basin studies, current meters, pressure gauges, and time-lapse photography, all in hopes of tracking rip current behavior. Much of that field and lab data are then fed into numerical models.

Ocean City is a long way from home for Varjola Nelko. A native of Albania, Nelko moved to Istanbul, Turkey where she studied engineering at Bogazici University along the shores of the Bosporus, the famous strait leading south from the Black Sea. In 2004 she arrived in Baltimore to start graduate studies with Tony Dalrymple, a widely published professor of coastal engineering at Johns Hopkins University. Awarded a Maryland Sea Grant Fellowship, Nelko is close to completing her Ph.D. dissertation­ based on their research on the rip currents of Ocean City. Credit: Michael W. Fincham.

With their cameras and freeze-frame X-rays, Dalrymple and Nelko are among the first to directly observe rip currents in action. By turning their observations into usable data, they are working around the drawback found in all the earlier forecasts. The Lushine predictive scales all focused not on rip currents but on rescues. Rescue totals can go up and down for reasons that have nothing to do with rip currents. Rescues can rise when sunny days or holidays bring out large crowds, and they can fall when cloudy skies keep people away. When nobody's at the beach, rip currents are still there, churning in the surf.

And Dalrymple's cameras are still there watching, capturing rips on cloudy days and all through the cold and windy off-season days when nobody's on the beach but a few walkers, bird watchers, and fishermen casting lines into the boiling surf.

Why try for a new way to forecast rip currents? Because Dalrymple and Nelko tested the forecasting tool now in use and came up with a failing grade. They took MALURCS, the Mid-Atlantic Lushine Rip Current Scale used by the National Weather Service, and they asked it a simple test question: How many of the actual rip currents they caught on camera could be predicted by the Lushine scale? Their answer: For every 100 actual rip currents, MALURCS predicted only 40, well below a passing grade in most schools. A score of 40 may be better than nothing, perhaps, but it means the forecasts now used for the Mid-Atlantic region are probably failing to predict 60 percent of the rip currents along the Ocean City shore.

IT'S NOT GODLIKE, BUT THE VIEW from his 8-foot lifeguard tower is high enough for Ben Davis to easily spot a flash rip and quickly whistle at a small boy who's being tugged gently seawards. The boy looks up, then wades slowly sideways out of the rip channel. That doesn't count as a rescue stat, but it's probably a lifeguard's most common catch, spotting a risk before it becomes a rescue event.

He drops the whistle and swivels his head slowly from north to south, surveying his kingdom, the slice of beach that stretches from the long wooden town pier down to the rockpile jetty at the south end of Ocean City. As crew chief, he has dominion here. With several hundred people under the watch of his crew, he has to catch more than 40 percent of the rips.

Like a lot of lifeguards, Davis is also an on-site oceanographer, reading the waves and the wind so he can make his own instant forecast of rip current dangers. Watching swells roll past the end of the pier, he notes their direction, estimates their heights, and counts off the seconds between swells.

Reading the rip, like reading music, is a learned skill. The best sign of a rip, for Davis, is color, color that's different from the rest of the water. A rip current can be darker because the water is deeper where it flows through a channel. Or it can be lighter, especially with "flash rips," because they pick up sand from the bottom and carry it seaward. "It can be very deceptive," warns Davis. A rip current, ironically enough, can look like a safe patch of water. As a current surges out, it can knock down the surf break, creating calm-looking water that draws in timid waders who don't want to battle breaking waves.

Lifeguards up on their tower chairs are not gods, but they're not powerless either. When a rip opens up and it's too late to whistle people away, they have options, primarily speed, teamwork, and training. A guard signals the next tower, then hits the sand running. "You just head there," says Davis. Next comes in-the-water triage. "You start getting people out of the shallows. If they are already in over their heads, you tell them how to swim out." If they can't swim, then the guards go after them.

That could mean battling through the breakers with a rescue buoy to reach swimmers in panic. Or bounding into the breakers with a surfer-style rescue board, perhaps the fastest way to reach a failing swimmer. Or swimming out a lifeline so lifeguards on the beach can haul exhausted swimmers back through the surf.

Last year about this time, they had to use all their tools. As the annual July 4th invasion hit Ocean City beaches, a tropical storm began forming some 3,000 miles away, starting wave trains heading this way from the west hump of Africa. By July 7, the storm, now a hurricane called Bertha, was hanging well east and south of Bermuda — and still sending wave trains headed west. The storm never came near Ocean City, hanging out past Bermuda for nearly a week. But its long-period waves began coming ashore by July 9. For the next week, a week of mostly sunny days, rip channels were opening up all along the beach, and lifeguards were scrambling to pull out all their lifesaving gear. The Ocean City Beach Patrol, according to Captain Butch Arbin, set its all-time record. In one seven-day period, lifeguards rescued over 2,000 people from killer-size rip currents. While three people drowned along the New Jersey coast, nobody drowned on the beaches at Ocean City.

Final exams for rookie lifeguards at Ocean City, directed by Captain Butch Arbin (top right), include working with lifelines (top left) for hauling swimmers out of rip currents and with rescue boards (middle left and right) for reaching swimmers fast. If they pass all their tests, rookies become certified Surf Rescue Technicians and receive a pay raise from $13.25 an hour to $14.42 an hour. The grand finale (bottom left) is graduation for the 34 rookies in the class of 2009. To kick off the ceremony, Arbin, a 37-year veteran, congratulates the first graduate of this year’s class: his son Michael Arbin. Credit: Michael W. Fincham.

HUGE STORMS LIKE BERTHA, IT TURNS out, can play a surprising role in the dynamics of rip currents at Ocean City. With freeze-frame images from their cameras in the sky, Dalrymple and Nelko are able to watch how sandbars are born and track where they go to die. And it's the life cycles of sandbars that largely control the setup and spacing of rip current channels.

The world of rips, according to Dalrymple, begins with a flat beach and a big storm that tears sand off the shore and carries it seaward. The sand never gets very far because the outgoing water runs into other incoming waves and simply drops its load. And voila! New sandbars are born.

The beach is now "set up" to form rip channels. When backwashing seawater can no longer run out over the sandbar, on ebbing tides for example, it then goes looking for another exit. Wherever there's a notch in a sandbar or a low spot along the beach, backwashing water will begin wedging its way through. A notch becomes a channel, and a rip current runs through it, pulsing most strongly after the arrival of large wave sets.

In the weeks after the storm, those sandbars, as seen in their photos, start creeping shorewards, pushed back towards the beach by incoming waves. During the pushback, sandbars move at different rates, and new rip channels appear among them, flashing open in unexpected places. "The beach gets very wriggly," says Dalrymple. Barring another big storm, the sandbars will eventually reattach to the beach. The beach will flatten out again and stabilize. Rip currents will dwindle.

Creatures of winds and waves, rip currents are also, it's now clear, creatures of bathymetry. For scientists with their freeze-frames, reading the rips now means reading the bottom also: seeing the setup, charting the rip channels, tracking the slow, wriggly trek towards shore. For lifeguards without X-rays of the bottom, reading the rips just got more complicated. Now they need to keep a weather eye out for how rip channels can change during a storm cycle.

The work day ends just the way it began, with Ben Davis hauling his 300-pound lifeguard chair across the beach. At day’s end he lays the chair on its side. Tomorrow morning, he’ll pick it up again and haul it back to the edge of the ocean where he’ll climb aboard and begin watching the waves for rip currents. Credit: Michael W. Fincham.

AT 5:30 BEN DAVIS STANDS UP on his chair, blows his whistle, and begins waving swimmers out of the water. And so do 91 other lifeguards along the 10-mile beach. It's closing time and Davis wraps up his American flag, climbs down, and begins packing up his gear.

For his last chore he tilts his tall chair tower over his back, all 300 pounds of it, and drags it thirty yards back through the soft sand and lays it on its side. It's pure grunt work and every guard does it every day.

The ocean empties, but the beach doesn't. And ten minutes later several swimmers begin wading back into the surf.

SO WHAT'S THE FORECAST FROM THE 14TH FLOOR?

When Dalrymple and Nelko built their new predictive system, they took their real-life rip currents as captured by their cameras and looked at each of the weather forces in play that day. In university research like this most of the grunt work falls to grad students like Nelko. First she plotted rip currents against wave heights, the best measure of how energy is hitting the beach. Then she did the same thing with wave period, which gives an estimate of speed. Then with wave direction. Then with wind speed and wind direction. That's an ocean of data she had to swim through.

When they tested their new system, they got a nice number. Their forecast predicted 72 out of 100 actual rip currents. Seventy-two percent is a big jump over 40 percent, the best the official forecasts could do with the same rips. That sounds like a passing grade in most schools, but it's not good enough at Hopkins, not yet. Nelko still has more data to wade through.

The scientists want to raise their POD, their Probability of Detection, and lower their FAR, their False Alarm Ratio. Their forecast unfortunately also predicted rip currents when there were none, a prospect that might keep people at home on perfectly safe beach days. "If people don't go to the beach," says Nelko "then you have a lot of merchants who are not selling saltwater taffy. It has economic consequences." If everybody stays home, then nobody's selling much of anything, not T-shirts or hot dogs or hotel rooms with a high-angle view of the ocean.

IT COULD HAVE BEEN AN EVENING like this. The chair towers are down, the guards are gone, and the slanting sun is lighting the ocean with a brilliant, celestial blue. The two boys who went swimming that evening stayed down at the south end of the beach where their parents could watch them from the shore. From there the father was able to watch as the rip current carried both the boys, now shouting, out past the jetty. And the mother was able to watch as the father, now swimming, went out in the glowing sea to save his sons.

It was a passing boat that pulled the boys to safety. And it was an off-duty lifeguard who found the father's body. They call the work "search and recovery," and lifeguards practice this drill also. It was Butch Arbin, captain of the Beach Patrol, who sat on the beach with a sobbing mother holding a two-year-old, now fatherless baby.

"Our guards don't remember every person they've saved. I don't remember how many people I've rescued," he says. "I just don't remember." But they clearly remember the ones they couldn't save. Around the country, most rip current drownings happen much like this one at Ocean City. No lifeguards in their chairs. No cameras in the sky.

And here's where new forecasts could save lives, even when lifeguards can't. For the foreseeable future, lifeguards will probably remain their own rip current forecasters, relying on their well-earned skills at reading the waves in front of them.

But for the rest of us who may be driving to the beach or standing on the sand trying to read the waves, good forecasts count. We heard the weather forecast before we left the house and we trusted it enough to get in the car and head out.

It's here perhaps, on beaches empty of lifeguards, that good rip current forecasts would matter the most, helping us decide whether to go home, rest on the beach, or go for a plunge in the unpredictable waves.