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.
At the mercy of the weather gods: marine surveying in Wales
Planning marine surveys is hard. Surveys have to be organised months in advance, but weather conditions dictate everything on the day. Too windy, and the swell picks up. While the boats we use have equipment (inertial measurement units) that track the motion of the vessel, large waves and the bubbles that come with that disperse the sound waves from the multibeam and leads to “noisy data”. Weather down time is a frequent occurrence.
The way around this is to plan multiple targets that take all weather conditions into account. The predominant wind direction in Ireland and the UK is from the southwest. Choose your primary targets, then choose some more in bays and inlets that face the northeast. Select some targets in shallow water, some in deep. Try to have multiple places to go, so if the wind changes direction and gets stronger, there’s always something to attempt to increase the chances of a successful survey. Thus far, we’ve been lucky!
We have now just completed our third season surveying in Welsh waters. In 2018, the RV Keary operated around Anglesey (Figure 1). Initially targeting wrecks to the south west of the island, the south-westerly winds picked up and the boat was forced to move to the northeast. Sheltered conditions here allowed us to continue data collection, with all bathymetry gridded up to 10 m, 5 m and 2 m resolutions. Finer resolutions can be generated, especially for wrecks which we generally grid at 0.25 m down to 0.10 m. The Puffin Island 2 m data has now been combined with CHERISH LiDAR data to produce one of our seamless onshore-offshore maps (Figure 2).
In 2019, The RV Keary returned to Wales, this time moving south and targeting the Sarn Badrig reef in Cardigan Bay along with other key CHERISH terrestrial sites (Figure 3). The weather this year was kind and allowed us to operate on the exposed western shore. Sarn Badrig is extremely shallow and during low tide the RV Keary shifted locations to St Tudwal’s Islands to continue data collection. The offshore regions of Dinas Dinlle and Rhosneigr were surveyed to tie in with the onshore work of our Welsh colleagues.
This year, the plan was to extend the work started in 2019, with Sarn Badrig and the South Sands around Menai Strait being our primary targets (Figure 4). However, as these are vulnerable to south-westerlies, a number of reserve areas were identified, extending from north of Anglesey right down to Pembrokeshire. The survey was planned around large spring tides, to take full advantage of deeper water over the shallow targets.
We’ve always received a warm welcome whenever we’ve made it over to Welsh waters. With this in mind, on our final day in Wales this year, it was an honour to participate in the sail past in memory of Mark Shackleton, Dock Master at Caernarfon.
The Crumbling Promontory Fort of Dunbeg
The National Monument of Dunbeg, which translates as the little fort, is a popular tourist attraction in County Kerry with its distinctive rows of defences and spectacular views over Dingle Bay to Valentia Island and the World Heritage Site of Skellig Michael. Unfortunately, the site has been periodically closed to the public for repairs and safety measures as the sea continues to erode the cliffs. This erosion is particularly severe during storms, and storms are predicted to become more severe with climate change.
When you approach the promontory from the road, you walk through four banks, five ditches, and an inner drystone rampart. A central causeway crosses the banks to the rampart but people are encouraged to walk on the eastern side of the fort due to erosion at the entrance. An underground stone-built passage, known as a souterrain, extends for over 16m from the rampart to the third bank. A stone-flagged pathway did lead from the rampart entrance to a circular drystone structure known as a clochán in the interior of the fort
We have a relatively good record of changes at this site as the promontory fort attracted the attention of19th century antiquarians and geologists as well as 20th-century tourists. George Du Noyer’s visited and recorded the site in 1856, and the triangular-shape promontory he drew has been indented up to 35m along its western side which sits on the 30m-high cliffs. This has resulted in the fort becoming more crescent-moon shape in plan today.
Human activity in the 19th century also impacted the fort, with hare hunters overturning stones, and stone being taken for building elsewhere. Drystone field walls that once crossed the fort banks and ditches were removed during Office of Public Works (OPW) restoration in 1892. The OPW also repaired the roof of one of the two guard chambers that sat on either side of the rampart wall entrance. The western guardhouse is no longer extant. The OPW repairs also made a curve at the rampart’s terminals and inserted a boundary wall. Previous plans of the site indicated there had been a straight rampart wall.
In 1897, Thomas Westropp said around 3m of land has fallen on the western side in the last 20 years. Professor R.A.S. MacAlister, later of University College Dublin, records that he visited the site in 1896 and again in 1898 and in that time the western end of the stone rampart had eroded into the sea. Another OPW visit in September 1915 sketches the disappearance of 9.5m of the western side of the rampart since 1897 and ground fissures, a sign of impending instability, were also shown.
In 1977, the OPW and National Monuments Service commissioned an excavation to examine the site, its dating and history of occupation before more features were lost. A view from the inside of the fort looking at the rampart wall shows the cliff erosion from the west had reached the western guardhouse to the side of the covered entrance. Excavation led by Professor Terry Barry from Trinity College Dublin revealed post holes, hearths and stake holes within the clochán and suggested wattle shelters supported by wooden posts and stakes. Analysis of occupation debris indicated a diet of pigs, sheep, goats, cattle, deer, birds and fish. Radiocarbon dates suggest it was inhabited in the 10th or 11th centuries AD. Further excavation at the rampart revealed an earlier shallow ditch radiocarbon dated to the 6th century BC. This indicates a long history of use at the site, though it may not have been continuous.
Within the last 7 years, the cliff has been experiencing another period of instability. In January 2014, a storm resulted in the southern side of the entrance through the rampart collapsing causing a section to fall away close to the passageway through the stone rampart. The CHERISH Project began early in 2017 and has been recording the latest changes with regular drone and laser scanner surveys. In December 2017, the site had to be closed again after flash flooding down Mount Eagle caused stream erosion within the fort causeway, banks and ditches. Then during Storm Eleanor on 3rd January 2018, most of the covered entrance through the rampart and the ground below collapsed into the sea. The last covered area of this entrance had collapsed by our next visit in April 2019.
Pembrokeshire’s Prehistoric Promontory Forts
It is hard to deny that Pembrokeshire’s coastline is amongst the finest in the British Isles in terms of its beauty and diversity. Most people who walk along the famed Pembrokeshire coastal path will come away stunned by what they see, and it easy to see why. However, when asked if they had noticed any of the 50+ prehistoric promontory fort sites on their journey many will remark at how they encountered some ‘strange humps and bumps’ but nothing out of the ordinary. This is a far cry from the Prehistoric Period where many of these sites would have been focal points in a busy prehistoric Pembrokeshire landscape, and extremely visible both from the land and the sea. Modern perceptions of these sites and their associated land and seascapes have moved far beyond those of their builders, with climate change and coastal erosion playing a key part in this perceptual shift.
Promontory forts are one of the most abundant classes of archaeological monument in the county but so little is known about them. Their density along the coastline suggests that they were important components of the prehistoric landscape, but their functions are not clear. Excavations have been extremely limited, hindering the reliability of interpretations. Porth y Rhaw to the south east of St Davids is one of the few sites to have benefitted from intrusive investigations where extensive evidence of several phases of settlement from the early-mid Iron age period to the 4th century AD were uncovered. Based on this work many promontory forts in the county have been interpreted as places of settlement for coastal communities to live and work while maintaining easy access to the sea. Whilst this interpretation may be true for most, vast differences in the size and layout of defences, interior space, likely functions, and access to the sea means that the catchall term ‘promontory fort’ is actually used to describe a hugely diverse class of monument. These sites are not only present in Pembrokeshire but along many coastlines that face out to the Irish Sea, English Channel and the Atlantic, extending their significance beyond the borders of Wales.
Whilst plentiful, promontory forts face a stark lack of understanding, made worse by the fact that they have been suffering from coastal erosion for centuries, traceable through analysing historic maps and aerial photographs. Not only has erosion led to the loss of potential archaeology key to understanding the site’s history but has also influenced the way they are perceived and managed by people today.
Work by CHERISH has aimed to improve the understanding of these sites and monitor them for coastal erosion through the use of several traditional and modern surveying techniques. For a selection of sites this has involved carrying out analytical earthwork surveys using GNSS equipment to record all archaeological features visible on the surface, which in the case of coastal promontory forts largely comprises the defensive earthworks. Not only have these surveys helped guide how sites are managed but they have provided an opportunity to think about them in a more critical way than before. At Porth y Rhaw a more meaningful interpretation of the large defences augments the findings of the numerous excavations. Several phases of defence construction along with numerous possible hut scoops were identified, adding another dimension to this site. At Caerfai Camp, a site which exhibits a unique series of four large banks and ditches, at least two phases of defence construction were identified. Earthworks relating to the route into the fort also provoked thoughts surrounding the function of the defences. Compared to the huge bank and ditch defences the entranceway appears to have lacked any form of defensive structure which begs the question whether this was a fort built for defence or to merely impress visitors. Geophysical survey of Caerfai’s interior, although partially obscured by geological interference, also interestingly revealed a stark lack of buried archaeological remains.
UAVs have also been used to good effect where they have not only helped to provide centimetre accurate baseline data for monitoring coastal erosion but objective 3D data that can be used for further archaeological interpretation and dissemination. Numerous sites have been recorded in both Wales and Ireland using similar equipment and techniques, resulting in some lovely data and products. In Pembrokeshire, both Caerfai Camp and Porth y Rhaw have been recorded using this method, with more surveys planned for sites situated within the Castlemartin firing range such as Linney Head and Flimston Bay promontory forts. 3D models of sites in Wales and Ireland have been created using UAV derived data and can be viewed here.
Conducting this sort of work on Pembrokeshire’s promontory forts is important for not only furthering the archaeological understanding of the monuments but raising awareness of them to the public, both in terms of their cultural value and the threats they face from climate change and coastal erosion. Archaeological survey and research work to provide the stories behind the monuments that now seemingly stand isolated along our coastlines, whilst erosion monitoring helps to make sense of how they are physically changing and the physical impacts upon the archaeology. Raising awareness of the archaeological stories and climate change issues is important as society moves forward and the preservation of archaeological sites slip down the agenda. This is especially pertinent to Pembrokeshire’s promontory forts which are a testament to the rich history of the region. Losing them to the unforgiving claws of climate change before we understand them as best we can would be a tragedy locally, national and internationally. CHERISH will continue with this objective over the coming years.