Bolt in the Bay – P47-D Crash Site


A few years back Grahame Knott of Deeper Dorset found the dispersed remains of an aircraft in Weymouth Bay. Over the years pieces of the jigsaw have come together, with pieces of information as diverse as component part numbers and an entry in the Portland War Diary (Held at the National Archives)  all adding up to tell us:

  • The aircraft is a Republic Thunderbolt P47-D.
  • Time and date of crash: 12:25pm 7th May 1944.
  • Reported crash position very close to present site.
  • Pilot bailed out and was rescued by HM Trawler Northern Gift.

At this point the trail of evidence runs out. Had the loss been an accident we would have the report. If the crash was combat loss and the pilot was missing for more than 48 hours a combat report would exist. The USAF have confirmed no paperwork exists – indicating the pilot was quickly back at base after a brief dip in the Channel.


The site has been on the list to scan for a long time. With the weather holding us from the planned site the team (Grahame, Nathan, Ruth, Mike and Rob plus me) thought the P47-D site was an excellent fallback. Two dives on the site, which sits in 25.2m of water, has now delivered a scaled plan of the known site. It took 1126 images to cover the area, and its always pleasing to see 100% alignment from Photoscan Pro.

Firstly, heres a peek at just the engine:

Typically for aircraft crash sites, the heavy components like engines, landing gear and weapons survive much better than airframes.

However, the site is quite dispersed, so the plan was to scan as much as possible. We added four control points (white squares in the model) and two of the team were tasked with measuring distance and depth.

The full model ran overnight. The original sonar was fine to dive the site, but we needed better data to embed into the model. A few days later Grahame went to sea to improve the sonar side scan view of the crash site. Multiple passes gave us the data we need to geo reference.

Sonar scan of the remains of the P47-D as they lie on the seabed, with boat track on the right.Both images © Grahame Knott.

Here’s the finished model, complete with GPS data:

The model has been annotated (numbered circles) pointing out recognisable parts, and parts yet to be identified.

With a scaled and geo referenced model we can now interpret and understand far more than a simple model:

  • The P47-D engine lies in a west-east orientation.
  • The main landing gear lies 16.1m west from the engine.
  • The turbo supercharger, originally mounted behind the cockpit, now lies in front of the engine.
  • Typically of sites like this, we have snagged fishing gear including a lobster pot.
  • To the north of the engine we have inlet/exhaust pipes plus lots of unidentified components.
  • One of the engine cylinder liners is 9.17m from its original location.

DEM and Ortho Photo

Regular visitors will know just how much we love DEMs. With the model complete the DEM quickly followed:

Digital Elevation Model of the Republic P-47D crash site.

There is quite a bit of the aircraft missing. We only have one main landing gear, no propellor and what we think is part of one wing. The tail wheel was found around 400m away, so its remains could be very dispersed. Or was it buried in the seabed? The DEM does not dig for us, but does reveal some lumps and bumps that might be natural, or hide parts of an aircraft.

Area of the DEM showing lumps and bumps that do not look like natural features.

The engine stands out, dominating the site. The lumps and bumps are quite subtle, but present.

The remains of the engine dominate the site.

As ever, the ortho photo quickly followed. The ortho photo yields such detail we can count the slipper limpets on the seabed, if we want:

Orthophoto of the entire crash site, extents of which are known and identified today.

As with all military aircraft, this site comes under the Protection of Military Remains Act (1988) but with a level of detail the ortho photo gives us we can learn much more, sitting in the comfort of home.

Interpretation & More Questions

We will carry on investigations – we really seek to name the pilot – but much more is now possible without getting wet. Discovery, investigation, recording and subsequent interpretation is what gets the Deeper Dorset team excited. The model, DEM and ortho photo can now be examined by a non-diving aircraft engineer who maintains a flying example (if you know one, do get in touch). No one needs to get wet or even go near scuba kit or a boat. Everyone can contribute, not just divers.

Here’s just a few questions Deeper Dorset now seeks answers to:

  • How did it crash?
  • Was it flying east-west when it crashed?
  • Has the site been trawled and dispersed by dragging?
  • Where is the rest of the aircraft?
  • Who was the pilot?
  • Did he survive the war?

Photogrammetry won’t directly answer these, but for speed and accuracy of recording the process is very efficient indeed.

Watch this space.