By Rob Rohde-Szudy - Madison, Wisconsin - USA


Shallow Water Salvage

So I finally joined the rest of the 21st century and got a GPS receiver. The first time I took it out was on the schooner on a day we’ll call “breezy”. I had only the reefed main up and a gust promptly snatched the sheet out of my hand and popped the reef tie off the tack. I turned downwind to get out of the channel to the launch ramp and noticed we were going pretty fast under reefed main alone. My speed freak wife asked if she should put up the foresail. After guffawing, I picked up the GPS to tell her exactly how close to hull speed we already were.

Another gust forced me to drop the GPS. I aimed for the chine and saw it heading that direction, but I had to look away before it landed. Well, this gust was strong enough I had to let the sheet out all the way – the boom thwacking against the foremast as I dumped wind. After a few moments the worst of it passed, and I headed up to get the sheet back and fix the reef tie, but the next gust pulled the steering cable out of its clamp. I had Colleen drop the sail while I had a word with Mr. Evinrude.

It was about this time I realized the GPS was no longer in the boat. I had cleverly thought of putting it in a freezer bag for additional waterproofing. I had even more cleverly thought to add foam flotation…but I didn’t actually do it! So it sank. But it sank in only 5-6 feet of water, and close to the launch ramp.

I think we both know that I couldn’t leave that alone.

It’s never that simple

Unfortunately, this is harder than it sounds. In the Florida Keys or Wisconsin’s cold northern lakes, you could just paddle over the area in question and look down into the water. When you see it on the bottom sand, well, there it is! This lake is nothing like that. The water is full of algae and visibility is maybe two feet. Worse, the bottom is soft muck at least 3” deep before it will support any weight. This means anything you drop is immediately enveloped in muck and invisible.

I jumped in the water and felt around a bit with my feet right after dropping it. But there’s too much ground to cover without a better plan.

Retracing steps

The single biggest question is how accurately you know the position where you dropped the item in question. Any doubt about the position increases the area to be searched exponentially.

I knew I went straight out past the buoy, then turned downwind, then came up as I was adjacent to a bit of a peninsula.

Here is a map of the area with my presumed track.

But this doesn’t seem quite right. The distance didn’t seem noticeably wrong, but most people are pretty bad at judging time and speed in such a situation, so I didn’t put much stock in my impression of the length of that downwind run. More importantly, the peninsula didn’t look that pronounced – more like a bulge. And I don’t remember any inlet before the peninsula.

I figured out why. The map was made over 25 years ago and the launch ramp has been moved since then!

Here’s where it is now.

This makes that downwind run a lot shorter! This is a good thing. Let’s look at why.


I figured I would probably need a rake of some kind to comb through the muck. Let’s assume I’ll build something 3.5 feet wide with a 16-foot handle. The 16 foot handle means that standing in one place I can rake a circle that is 32 feet in diameter. But to cover a large area, I need squares that can butt against one another. This is easy math.

Square Calculation

The diameter forms the hypotenuse of a right triangle, which is also half of the square we care about. Since the legs are equal, we square our 32-foot diameter to get 1024. Divide that in half because the two legs are equal. That’s 512. The square root of 512 is the length of any side of the square – 22.6 feet. Let’s round that down to 22 feet.

Let’s also figure out how to cover that area. The circumference of a 32 foot circle is 100.5 feet. To allow a little overlap I’m calling the rake 3.5 feet wide. Dividing the circumference by the width tells us we need to make 28 strokes to cover the circle. In case we need to use a compass to line up the strokes, dividing 360 degrees by 28 tells us the strokes will be a little under 13 degrees apart. Assuming I can make one stroke per 15 seconds, a square would take 7 minutes. Realistically it would probably be at least 15 minutes per square to get any kind of accuracy.

Back to the map

So let’s divide up the area we need to search. The downwind run is about 400 feet long. Divided by 22 feet, we need 18 squares to cover it. We probably need a wider swath than 22 feet, since the sheet might have flung it some distance overboard. So let’s assume two parallel runs of squares, or 36 squares. This would come to 9 hours based on my estimates above.

That is within the realm of possibility, so it’s time to design a rake. Time to bring in the big guns.

“Trust us, we’re experts…”

Jim Michalak once described me as having “boundless energy and enthusiasm.” I guess he never met Dave “Shorty” Routh. You might remember that I mentioned him in the Welding 101 article. He has never met a project he didn’t like, and his infectious enthusiasm has spawned a global fleet of PDRacers. Need I say more? Anyway, Short also springs to mind for shallow water salvage. Check this out: Surely a GPS would be easy if we put our heads together.

Shorty’s concurred that a rake was only way to recover anything from muck. Shorty pointed me toward special rakes used for recovering golf balls from water hazards. I had a vague memory that clam rakes were fairly similar, and so they are. (

After I got this together, Shorty pointed out that a heavy rake might want to sink into the muck. He had done some Koi rescues where he stood on a full sheet of plywood laid across the muck at the bottom of a pond. By the time he was done the sheet had sunk 20” deep! His helper stepped off and sunk to the shoulders and had to be hauled out with a rope from shore. Because of this, Shorty felt that skids were a necessity. He was also worried that a stick handle would cause the tine angle to change as it was hauled in, and suggested making it rope pulled. It was too heavy to lift on the end of a pole anyway.

So I spent about five hours I together this rake from scrap and $6 worth of 3/8” concrete reinforcing rod. (I used about 15 feet – it was on sale.) I assembled the wood first and welded the metal parts in place. Just be ready to smother some minor fires as you weld.

Note how much of my 5x7 foot cockpit this implement takes up!

Rake in cockpit


Shorty also felt that the angle of the tines should be adjustable. He had a mechanical adjustment in mind, but I simply bored some extra holes. This requires more disassembly to adjust, but it was faster to make. I didn’t even know if it would work at all, and I had invested enough time without any kind of test.

So let’s find out if it does work.

Rake Testing

Guess what? That big, heavy rake floated like a cork! I had to take off all the wood except the vertical plywood on the sides to get the tines to sink into the bottom. I could pull it a little by hand, but the 18 horse Johnson pulled it just fine, leaving a giant plume of mud in the water and a large mass of Asian Milfoil weed on the rake.

But that wasn’t the big discovery at the lake.

Shorty and I independently had the same idea to make a fake wooden GPS and weight it with lead to be the same as the original. I guess it must be a good idea. If the rake can grab that out of the mud, it should grab the real thing.

Dummy GPS

I got in the water to test how quickly this would sink. Guess what? It also floated like a cork! Garmin says it doesn’t float, but I guess they are just covering their butt so people don’t expect it to float after the battery compartment floods. In a plastic bag it most certainly floats, and with the bag sticking out of the water.

Shorty tested his with the same results and my new one does the same thing. In the sink it looks like it’s touching the bottom, but it’s not.

GPS floating

I should have made this test first, because this completely changes the game. Rather than sinking where I dropped it, the GPS would have floated downwind, and rather quickly in the winds we had that day.

Let’s look back at the map.

Lake Map 4

The thin lines show the probable direction it would have been blown. It would have floated until some time after the bag was holed, so it could have easily gone to the far end of the lake (not pictured), creating a huge search area. But chances are it became the catch of the day for one of the fishing boats downwind. Oh well. I guess I’m done looking for something that probably isn’t there.

Rake Re-Engineering

The case might be closed on the GPS, but Shorty still thought there was something worth pursuing. He liked the big rake for dragging for lost anchors, but I couldn’t see spending hours digging up the bottom of the lake for anchors I don’t need.

A note from the Department of Seriously, Don’t Bother

I know my buddy Don is going to read this and start up again plotting to recover an old Evinrude from the bottom of a lake up north. Everybody say it with me, now: “Seriously, don’t bother.” I’ll bet $10 that motor swallowed water and broke a connecting rod or two when it fell off the transom 20-some years ago. If you want to find it you can borrow my boat…if you keep the rake! I take no responsibility for what your wife says or does to you if you take that bet.

Shorty also brought up the idea of fishing around near piers for all the stuff people drop – tools, fishing tackle, cell phones, etc. Or next to bridges for stuff they dispose of – engagement rings, murder weapons, mobsters. I thought this was more interesting, but now we’re looking at a whole different tool. It has to be small and light enough to reasonably stow in a boat, and it has to work in most bottom conditions. It doesn’t need to reach that deep. Probably it would make most sense to have it mount on the pushpole you already have.

Shorty had the idea to attach a net to the open space above the tines of a common garden rake - Brilliant! But I think we want a bigger net, and I think the tines would get in the way in any bottom but soft muck. It is almost never soft muck at a boat ramp, since the propellers tend to blow away anything soft.

The simplest approach might be a common fish net – the kind with the metal handle that looks like an oversized tennis racket – lashed to the pole. Maybe we’d like to flatten it a little, and maybe we’d like the net cells a bit smaller. And I think we’d probably like the rim to be thinner so it can scoop up small stuff rather than shoving it along the bottom.

But I will have to leave that experiment for spring or to someone further south.

Was this worth it?

Good question. I think the experience was interesting. Even if I had recovered the GPS, it is only economical if you don’t value your time very highly. A GPS like this one only costs about $100. After deducting materials and spreading it over the about 20 hours I spent, my labor comes in at under minimum wage. Way under when you consider that I paid only $51 for a used GPS to replace it. (I gave up the color screen, but that doesn’t matter much to me.)

On the other hand, it started me thinking on whether I’d like a net that can fasten to a pushpole, which might have been worth something.

That’s Shorty’s take on it, and why it is never a good idea to talk to him about any project you don’t want getting out of hand. Don’t believe me? Look at the number of PDRacers all over the world!

Rob Rohde-Szudy
Madison, Wisconsin, USA

Click Here for a List of Articles and Columns by Rob Rohde-Szudy

To comment on Duckworks articles, please visit our forum