Stacks, Cliffs and Cauldrons
Back to fieldwork
The Coronavirus epidemic has made 2020 an unusual and difficult year for millions of people. Along with most other organisations the staff of the Royal Commission in Wales, which leads the CHERISH Project, began home working in March 2020. Priority fieldwork recommenced in August, with each task requiring a robust business case and detailed risk assessment.
In mid August the CHERISH Team from the Royal Commission had their monitoring fieldwork on the Castlemartin Firing Range approved. The aim of this new fieldwork was to carry out the first aerial photogrammetric surveys by drone of the four principle coastal promontory forts, Linney Head, Flimston, Crocksydam and Buckspool/The Castle to provide baseline models to monitor future change. New topographic ground surveys were also needed for Buckspool and Crocksydam forts, which were last surveyed in the 1970s, while a possible ‘new’ promontory fort noted during aerial survey on Crickmail Down needed investigation.
Archaeological survey on a live firing range
We can only carry out fieldwork during scheduled non-firing breaks on this very busy military range. While Flimston Bay promontory fort can be accessed on some evenings and weekends, Linney Head promontory fort lies in the live firing area and can only be visited when the entire range is closed, usually during Easter and August.
Although Dan and Toby on the team are qualified drone pilots, our drone surveys also required advanced permission from Cadw, the National Park, Natural Resources Wales and the Defence Infrastructure Organisation. We are grateful to all the staff we worked with to gain permission. The timing of the August survey avoided the sensitivities of cliff-nesting birds but we had to remain aware of early pupping seals on the beaches.
We made our first CHERISH Project visit to the range in March 2018 with colleagues from Aberystwyth University, but the first modern detailed surveys of Linney Head and Flimston forts were made a decade earlier by Louise from the Royal Commission in 2008. These earlier surveys, together with century-old maps and historic aerial photographs, provide excellent baselines against which to judge longer term patterns of erosion in the face of a changing climate.
Travelling in separate cars, and assigned separate sets of equipment, we made our first stop at Castlemartin Range Office early on the first day of fieldwork for the Range Briefing. This was to make sure we could identify and avoid any ordnance that might be lying around in the live firing areas.
Surveying at the edge of the cliffs
We were fortunate to have a week of hot, sunny and largely windless weather to conduct our drone surveys. The drone we fly is a Phantom IV Advanced, using software which allows us to pre-programme a gridded survey flight for photogrammetry, including setting altitude and ground resolution. Before starting the flight a network of control ‘crosses’ are fixed to the ground and surveyed in with GNSS (Global Navigation Satellite System) equipment so that the finished 3D model is accurately located to within a few centimetres or better.
The week started at Linney Head promontory fort in the live firing part of the range, with a condition check on the nearby Bulliber Camp too. Then we moved east to conduct a linked coastal drone survey of both Flimston Bay and Crocksydam promontory forts before finally relocating to near St Govan’s Head for access to Buckspool fort and the newly-identified site at Crickmail Down.
A new promontory fort, and an historic quarry quay
An interesting headland at Crickmail Down had looked like a potential promontory fort from aerial survey in 2018 but it required a ground visit to be sure. It was found that a definite causeway enters the fort between the eroded remains of two ditches. Inside there are traces of low stone walls, possibly from one or two small round houses. The character of the surviving low ramparts and infilled ditches suggests Crickmail Down may be earlier than the other more substantial promontory forts nearby, possibly dating to the Later Bronze Age.
We were also surprised to discover a tall stone wall in the coastal gully below the fort, set between high sea cliffs. This appears to be a loading platform for the historic limestone trade which flourished into the early years of the twentieth century. Although the wall is not marked on historic mapping, and not presently recorded, it would suggest a place where quarried stone was loaded onto waiting ships, similar to a loading quay at the tip of Flimston promontory fort.
Processing the results
These were the first archaeological drone surveys of the eroding cliffs at Castlemartin, leading to an extremely detailed 3D record of the cliff edge & forts and surpassing the old 2004 2-metre resolution airborne laser scanning (LiDAR) survey which can be viewed on the government’s Lle Portal.
During the week we surveyed 2.8km of coastline to 2cm resolution with the drone, flying 87.5 hectares of vertical photogrammetry and gathering around 3200 vertical images plus many obliques, as well as several minutes of aerial video. Processing all this data whilst homeworking has its challenges, but we are starting to produce finished models of the coastal sites.
The new 3D models for Flimston Bay and Crocksydam promontory forts show both sites in extraordinary detail. The design of Crocksydam is very different to the tall, curving ramparts of Flimston and it may be that these neighbouring sites had quite different purposes. These new 3D models provide the basis for new analysis of these interesting prehistoric coastal fortifications.