Underwater Search and Recovery: How Search Patterns Actually Work

By Gus Marchetti, Salvage master. Reviewed by Dana Whitcombe, technical reviewer.

Marine Salvage & Recovery | Hull & Bottom Cleaning Last spring, a shrimper lost his net and an anchor in 45 feet of water off Port Aransas. He gave me a waypoint he'd saved from his plotter. It was off by two hundred yards, easily. So we ran a creeping line parallel to the current, not an expanding square, because the drift was steady and the area was a long, narrow strip. Found the anchor on the third leg, hung up on a cable. I've been running wreck removal and marine casualty response for 22 years. Search patterns are the unglamorous core of this job—the part nobody films for TV. Anyone can buy a side-scan sonar. Not everyone knows how to drive the boat so the sonar actually sees what's down there. Patterns are what separate a recovery from a wild goose chase. They're the difference between spending eight hours on a job and spending three days.

Why Patterns Beat Gut Instinct

When you're staring at murky water, your gut lies to you. It tells you, 'It's over there, I can feel it.' That's how you burn fuel and daylight. Search patterns exist to impose math on chaos. They guarantee coverage—provided you follow them like a pilot follows a checklist. No one gets to wing it. Not even me. And the first rule is: you don't know where the target is. You know where you think it might be. So you build a grid. You cover the area systematically, with a set spacing between lines. That's the only way to ensure you don't leave a hole the size of a boat in your search area. Call it paranoia, but I've found too many targets just outside the previous guy's search line.

Expanding Square: The Old Reliable

The expanding square is the first pattern I learned, and it's still the one I use when the target could be anywhere within a mile. You start at a datum—the last known position—and run a square that gets larger with each leg. You steer 90 degrees left or right, lengthening the legs by a fixed amount. It's elegant. But it has a flaw: if there's a current pushing the target while you're searching, your square drifts, and you're searching the wrong water. I remember a job off Galveston where we were looking for a 40-foot trawler that went down in a squall. No mayday position, just a last AIS hit. We ran an expanding square with a Klein 3900 towed side-scan. Took us four hours and two false targets—one was a car engine, the other a rock formation. Good thing we kept the lanes tight, because the boat was sitting right on the edge of the second-to-last leg.

Creep Line and Current: The Real World

If you know the target is in an area with a current, you don't fight it—you use it. A creep line, or search lane, runs the boat parallel to the current axis. You run one leg down-current, turn 180, and come back up-current, slightly offset. The current moves you through the water, so on the down-current leg you're covering ground fast; on the up-current leg you're pushing against it, which gives your sonar a better look. That's how we found that anchor off Port Aransas. We used a 100-foot lane spacing in 45 feet of water, towing an EdgeTech 4200 side-scan. It's a pain to set up, but it's honest work. And here's the digression: once we were looking for a container off the Mississippi birdfoot. The nav data said it went overboard at a certain bend. We ran five miles of creep lines east of the ship channel because that's what the plotter showed. Third day, we got a tip from a shrimp boat that he'd seen a patch of sheen to the west. Turns out the current had swung after the incident, and the container had settled a full mile west. We lost two days to bad data. That's the job. Sometimes you're just wrong.

Lane Spacing: The Math That Saves Time

Lane spacing is the most important number you'll decide. It comes from the sonar's range. If your side-scan can see 300 feet to each side, a 600-foot swath is your theoretical max. But to ensure overlap and to catch weak returns, you don't run the edges. You set lanes at 50 to 70 percent of the range. So for a 300-foot range, that's 150 to 200-foot spacing. Call it three-quarters of the distance, give or take. Get it wrong, and you're either double-covering—wasting time—or leaving gaps that hide the target. I've seen guys run lanes too wide to save time, and then they wonder why they 'swept' the area and found nothing. They found something, but they didn't look close enough. One time, we had a diver lose a $5,000 camera in 30 feet of water. We ran a magnetometer (a JW Fishers pulse-8x) with a 100-foot lane spacing in a lake. It took us 45 minutes. The camera was in the first gap we would have had if we'd done 200-foot spacing.

Tools: Side Scan, Magnetometer, and the ROV

The pattern doesn't matter if your tool can't see the target. Side-scan sonar is the workhorse. It paints a picture of the bottom using sound shadows. Good side-scan, like a Klein 3900 or EdgeTech 4200, can spot a car tire in 50 feet of water if you're driving straight and steady. But it needs speed. You've got to maintain 3 to 5 knots, over the ground, not through the water. Too slow, and you get silt clouds. Too fast, and the image smears. Magnetometers are for ferrous metal—anchors, chains, submerged cars. They don't need as tight a lane, but they can't tell you shape, only that something metal is below. Usually, I run side-scan first, then fly the magnetometer over an anomaly to confirm it's not a rock. If we find something that looks like a human body or a critical piece of evidence, we'll send down an ROV—a VideoRay or BlueROV2—to get eyes on it. The ROV is not for searching; it's for inspecting. You always search with sonar, then inspect with video. That's a rule.

Frequently Asked Questions

What's the best underwater search pattern for a small area? For a small area under half a mile, an expanding square is quickest—you start at the last known position and spiral out. Keep lane spacing tight, at 50% of your sonar's range, to avoid missing the target. How do you account for current when planning a search? Run your lanes parallel to the current, not across it. That way, the current helps you cover ground on one leg and slows you on the other, but your GPS track stays true. If the current is strong, adjust your pattern by starting up-current so drifting target stays in your search box. What equipment do you need to find a sunken object? At minimum, a side-scan sonar or a magnetometer, depending on whether the target is ferrous metal. A boat with GPS and a track plotter is essential. For confirming what you'll find, an ROV with video is a good add, but you search with sonar first. Why do search patterns fail in real life? Mostly because people don't follow them. They see a likely spot and break the pattern to 'check it out,' creating a gap. Also, bad coordinates or current drift can put you in the wrong area entirely. That's why you always record your search track and verify your datum.