A multibeam echo sounder makes a picture of the seabed out of a million soundings. Ours renders a fifty two metre hull, upright, with a scour pit round it. Two dive teams have been down and found nothing but sand.
I am a hydrographic surveyor. Fifteen years, mostly nearshore charting and pre-dredge surveys for port authorities.
The instrument does one thing very well and it is worth understanding.
A multibeam echo sounder sits under the boat and fires a fan of sound across the track, a few hundred beams at once, hundreds of times a second. Each beam's return gives a depth at an angle, and the boat's position and attitude come from a GPS and a motion sensor accurate to a fraction of a degree.
Put that together over a survey and you have a grid of the seabed. Ours is half a metre.
A wreck at that resolution is not a smudge. You see the hull, the deck line, the shape of the superstructure if there is one, and the scour pit the current digs round anything that stands proud.
At a position in about thirty one metres of water, four miles off, we render a wreck.
Fifty two metres long, upright, sitting a little down by the stern, with a scour pit around it about two metres deep, which is exactly what a hull that size develops on that bottom in that current.
It is a good target. If you showed the surface to any surveyor in this country they would say wreck without hesitating.
It is not on the chart.
That is not automatically strange. The wreck database has thousands of entries and there are things on the seabed nobody has ever recorded, and finding one is a good day.
So we did the correct thing, which is report it. Position, dimensions, orientation, least depth, and the surface itself.
The Hydrographic Office logged it as an unidentified contact and asked for a diver examination, which is standard for anything with a least depth that could matter to shipping.
The first dive was in 2019.
A commercial team, four men, two dives, on a slack tide, with a shot line dropped on our position by our own boat with our own GPS.
There is nothing there.
Flat sand with a slight ripple. No hull, no debris field, no scour pit, no magnetic anomaly on the magnetometer they towed on the second day.
They swam a hundred metre circular search from the shot and found one lobster pot and a length of trawl warp, which is what four miles off this coast is full of.
We surveyed it again the following week, because obviously.
It is there. Same position to within about two metres, same dimensions, same orientation, same scour.
Since then it has been surveyed eleven times, by us five times and by two other contractors, on three different vessels, with three different manufacturers' multibeam heads.
Every one of them renders the wreck.
The second dive was in 2022 and it was arranged properly.
The dive boat and the survey boat were on site together. We ran a line over the position with the multibeam live on the screen, called the mark, and the divers went down a shot that was placed while the target was on our display.
Two divers. Twenty two minutes on the bottom. Video.
I have watched the video. It is sand. You can see the shot weight sitting on it and the divers' fins putting up silt and about six metres of visibility, and they swim a square out from the shot and there is nothing but sand and ripple and one small crab.
At the same time, on our screen, forty metres away and eleven metres to starboard of where the shot was, we were painting a hull.
Things a surveyor tries when this happens, and it does happen, because sound in water is a difficult business.
A multiple. Sound bounces off the surface and the bottom and comes back late, and a strong return can render a false target at twice the depth. This is at thirty one metres in thirty one metres of water, not sixty two, and multiples do not have scour pits.
A ray bending error. If the sound speed profile is wrong, the outer beams curve and the seabed renders as a smile or a frown. We take a sound speed cast every two hours because of exactly this, and the artefact it makes is symmetrical about the track, and this thing does not care which direction we approach it from.
Gas. A plume of gas from the seabed does render as a target, and it can be persistent, and this was our best answer for two years.
Gas plumes go up. They render as a column in the water column data, which we record. There is nothing in the water column at that position at all, and gas does not have a deck line.
The wreck database.
I did the thing everybody does eventually and looked for what was lost near that position.
There are four entries within two miles and all four are accounted for and dived and known.
There is a fifth entry with a position given as approximate, which is normal for anything lost before radio position fixing. It is a coaster, fifty one metres, lost with all hands in the winter of 1936, and the position on the record is a dead reckoning from a report by another ship.
It is one and a half miles north east of our contact, which in 1936 dead reckoning is the same as saying nowhere in particular.
What happens now.
The contact is in the record as unidentified, with a note that two diver examinations found no obstruction, and its least depth is entered as the surrounding seabed, which is the correct entry for something that is not there.
It does not affect navigation. Thirty one metres is thirty one metres and nothing on this coast draws that.
We survey that block every three years under the port's rolling programme and I will be on the boat for the next one.
The one thing I have changed is that I no longer put the live display on the big screen in the wheelhouse when we run that line, and I take it on the smaller one at the survey desk instead.
I told the skipper it was so the glare did not bother him at night. He said fine, and he has not asked again, and he takes us over that block at a course and speed that puts the line at the end of the watch, which he did not use to.