Kilwa Bonanza

Heritage Futures and Community Archaeology in Tanzania

The Kilwa Kisiwani site off the coast of Tanzania was occupied from the 9^th and was one of the most powerful settlements along the Swahili Coast.  From the 13th to the 16th century, much of the trade in the Indian Ocean was handled by the merchants of Kilwa, who dealt in gold, silver, pearls, perfumes, Arabian crockery, Persian earthenware and Chinese porcelain. In 2004, the ruins were placed on the List of World Heritage in Danger (https://whc.unesco.org/en/list/144).  The Great Mosque of Kilwa Kisiwani is the oldest standing mosque on this coast and the monuments are still in everyday use today by both villagers and as one of the oldest Muslim schools in East Africa. However, the pressure of daily use as well as the impacts of climate change, manifest by sea-level rise and increased weathering and vegetation growth, are threatening to severely degrade the heritage site.

Kilwa Kisiwani World Heritage Site

The interdisciplinary ‘Safeguarding Heritage in Tanzania’ team from the University of St Andrews, comprising Dr Richard Bates, Donald Herd and Dr Ted Pollard (Earth & Environmental Sciences), Elie Graham (History), and Dr CJ Davies (Computer Science), together with colleagues from the University of Dar es Salaam (Dr Elgidius B. Ichumbaki), have recently returned from a highly successful field trip to Kilwa Kisiwani, Southern Tanzania. 

The team are investigating the Kilwa site using remote sensing technologies and initiated a community-led programme for raising awareness to the potential loss of both tangible and intangible heritage in the area.

preparation of dGPS, and drone surveys, Kilwa

 Despite previous international aid programmes for protection of the tangible heritage, often little is done for the local community. In March 2019, the team hosted an education day, “Kilwa Bonanza”, with local schools, bringing to the island over 150 primary and secondary pupils for tours of the site.  Following the tours, the school groups were joined by over 250 people from the local community to enjoy the benefits of their heritage. Raising public awareness and the popularisation of the sites is being aided by the use of contemporary performing arts (https://youtu.be/xaD93UJqfOg).

In addition to the community groups, the “Kilwa Bonanza” was attended by local District Commissioner, representatives of the Department of Antiquities, Dept of Tourism and local business.  The Bonanza was catered for by a new start-up venture, WAUMAKI (The Woman’s Culture, Buildings and Heritage Organisation in Kilwa). This group aims to provide visitors to the site with tour guiding, transport around the site and local region, catering for the visitors as well as offering a range of indigenous, locally-produced  hand-craft.

Kilwa was a site that is intimately associated with the sea, with people relying on the sea for resources and, historically, for trade. The nearshore coast also contains heritage in the form of submerged structures, due to sea-level rise, and shipwrecks.  As part of their preservation training, the team taught the first Nautical Archaeological Society underwater archaeology courses (Parts I & II) in East Africa to a group of local divers from government agencies, university departments and local tour companies.  The trained divers will provide a valuable resource for maritime archaeological investigation in the future. Some stunning new finds from a possible 13^th century wreck near the old Omani port in Kilwa were recovered on the last day of the training.

Pottery destined for curation at the National Museum

Tanzania does not have a country-wide digital database for their cultural heritage. The team met with the Director General of National Museums in Dar es Salaam as the first step to working with them to produce this. A digital record is seen as a vital resource for recording sites (such as those at Olduvai Gorge where the Leakeys accomplished their ground-breaking work on early hominins), managing sites, protecting them and creating a long-term legacy for the country. It is hoped that the approach adopted by the project in Kilwa, especially the inclusion of community outreach, will provide a blue-print for other sites across Tanzania and Africa.

Following the signing of a joint MOU between St Andrews and the University of Dar es Salaam, Tanzanian staff will be visiting the University of St Andrews in May to train on field and processing digital techniques and to work on some joint Scottish-Tanzanian fusion of musical experiences in the promotion of our joint heritage. For more on the project, see www.urithiwetu.org .

Callanish Geophysics

Callanish (Re-visited)

At the end of last year we were invited to do some further recognisance work in the Outer Hebrides at the iconic Callanish (Calanais) standing stone sites.  We had visited these back in 2016 as part of surveys for the Ness Historical Society in northern Lewis and were not only impressed by the stones and their setting but a little surprised about the lack of any other significant historical sites in the surrounding landscape.  On return from this trip I started to read more about the area and it turns out that relatively few excavations have taken place here over the last 50yrs or so.  For those who have not visited the sites then I can only urge you to do so. 

So what was our job here?

We have two primary objectives for studies at Callanish.

• First we need to map the old landscape to enable us to understand where and how people might have used it.
• Second we need to investigate potential occupation sites, in other words, find the people in the landscape!

The stones are very different to those in the Orkney Isles that we have been working on for the last few years.  The primary rock at Callanish is the beautiful, banded metamorphic Lewisian Gneiss.  At a little over 2.7 billion yrs old these are some of the oldest rocks in the UK – their age is a testament to their surviving power.  Deformed under numerous tectonic events they are hard and resistant to erosion.  Their beauty, with swirling patterns of dark amphibole minerals coalescing into augens (literally eyes), gives an additional sense of mystery to the standing stones.  The rock’s resistance to erosion is another feature of the landscape that the Neolithic people would have had to deal with.  Unlike in the Orkney, where stone is readily prised from the bedrock, Lewisian Gneiss does not break off with the same ease.  Rather, substantial effort is required.  This, along with the soft curves of the standing stones suggests that this material was not quarried in substantial quantities from the bedrock, rather at best it was prised from the surface or picked up from where it had fallen.  (for more on the creation of the Lewis landscape as a function of successive glaciations see Dawson et al. 2016).

Palaeo-environmetal work in Lewis and Harris have shown that prior to the re-peopling of the islands there was open woodland cover but with increasing human activity around 3900yrs before present there was an expansion of heather moor and the spread of grassland.  The latter possibly occurred as a result of fire and grazing. (Birks and Madsen).  The changes in vegetation and the deterioration of weather further encouraged the growth of the peat that is such a dominant part of the landscape today. In fact the peat growth is so prolific that when the main standing stone site was first investigated in the 19^th C over a meter of peat removal was necessary to expose the old landscape. Electromagnetic surveying around site III indicates that this technique could well be used to map peat thickness.

This result of the removal is manifest by the weathering marks seen on old photographs of the site.  The prolific peat growth is likely one of the reasons that we do not know a lot more about this old landscape as it remains deeply buried today.  There is another part of the landscape however and that is the one that has been lost to sealevel change.  As with many places around the Scottish coastline the sealevels have risen significantly since last glacial maximum.  The story in the Orkney Isles is fairly simple with a steady rise to about 4000 years ago when it reached  near enough present levels.  In the Outer Hebrides it is a little more complex with a sharp rise followed by slightly higher sealevel than today and then a fall and rise again.  Such a complex signature is a result of the interplay between global sealevel and local tectonics (that is the compensation for the pushing down affect of the ice on mainland Scotland).

The field work was undertaken over three trips earlier in the year.  Initially I visited with Mel Chocholek and James Killingbeck in order to survey some of the lochs surrounding the main site.  At this time we also undertook more photography for structure from motion 3D models of the upstanding stones sites and tested electromagnetic ground conductivity geophysics across some peat areas near the main stone circle.  In January I returned to the site with Vince Gaffney and Chris Gaffney from Bradford, James and Shiobhan Killingbeck, Tim Raub from St Andrews and Fanny Bessard from Bristol.  During this trip we concentrated on one site, Callanish XI in order to test a range of geophysical methods for intense survey.   I will report on this site later.  On Callanish X we scanned all the stones and this will shortly be rendered as a full 3D reconstruction of the site. 

In late January I then returned with my brother and Mel with the Swordsman to conduct further offshore surveys and to acquire core data through some of the sediment units.

https://youtu.be/LtpAJ2CvVbk

This shows some of the sub-bottom profile data that indicates the area will have sediment that we can target for environmental information. 

Irish Sea Cruise

So if you follow what we have been doing on the Lost Landscapes front over the last couple of years you will know that we have a large programme of work on reconstructing “lost”, submerged and buried landscapes around the world’s continetnatl margins.  The Lost Landscapes project is one of the core elements of this programme and is an EU-funded project based around reconstructions in the North Sea across what has been termed Dogger Land. A recent addition to this project has been an extension into the Irish Sea and specifically in Liverpool Bay and Cardigan Bay off the Welsh coast.  Liverpool Bay was the focus of work a few years ago before the large windfarms were constructed and has been written up in a series of reports and books (see Fitch and Gaffney, 2011 for example). I will leave this for a later discussion.

https://www.youtube.com/watch?v=xl3u7ZymGIY

Cardigan Bay has of course always been of interest to me growing up in Aberystwyth - swimming at Clarach, walking the coastline to Aberaeron and New Quay and playing in the dunes of Ynys Las.  More recently I have also been doing some work with my brother and father on the submerged forests at Borth.  Throughout my childhood I remember walks to see the forest of stumps that are periodically exposed after storms rip in from the Atlantic on this stretch of Welsh coast.  Over the last few years they have been exceptionally well exposed as there have been various engineering works going on along the shore for protection measures.  Back in February 2012 while my parents were taking one of their usual “what’s going on?” walks my father noticed blackened and shattered stones amongst the stumps which my brother later identified as burnt cooking stones.  This led to a series of projects investigating the site that yielded not only some well-preserved animal foot prints but the most impressive set of ancient red deer antlers (dated to 2940+/-30 B.P.)  that I have ever seen!  We did some geophysics (resistivity imaging and electromagnetics) on the beach which was published in 2016 (Bates and Bates, 2016). The forest here is dated to Bronze age but the deposits extend under the beach and offshore.  It is these earlier deposits that we were attempting to target with the current survey.  Specifically we hoped to be able to map the Mesolithic landscape and acquire material from that time period for environmental analysis including DNA sequencing. 

Borth fossil forest - photo D. Bates

To survey kilometres offshore and get core material through potentially metres of modern marine muds requires large survey vessels.  Thankfully a collaboration with colleagues at UCC (Dr Ben Geary) and IT Sligo (Dr James Bonsall) gave us access to a week on the Celtic Explorer.  This vessel is operated by the Marine Institute, Ireland and is equipped with all the tools of the trade including multibeam sonar, sub-bottom sonar, sidescan sonar, and both vibro-corer and gravity corer!  Lots of TOYS!!!  (For a link to these jump to the end of this blog)

After some initial set-backs with logistics the Voyager crossed the Irish Sea up to Liverpool Bay.  A night of geophysics yielded a number of target spots for the coring.  Here the previous work greatly aided us in picking target sites but the overlying sand proved too much of a challenge for the 3m vibrocorer.  So on to Cardigan Bay.

Background work for Cardigan Bay is a lot more sketchy than up north with most of the previous surveys collected back in the late 70s for speculative oil and gas exploration.  Our first task was once again to gather information from the geophysics.  The subbottom profiler on the Voyager is a two by two pot pinger.  Usually I would be a little sceptical of these for giving any great penetration of the seafloor however this setup has completely made me re-evaluate that.  From the first we started to see small channels buried a few metres beneath the present day seafloor that cross cut what appeared to be a glacial till horizon. Till is the material that is left as a result of ice retreat and so tells us about conditions during these cold periods of history.  The last cold (glaciation) period had its maximum about 27,000yrs ago (Clark et al., 2009) with ice reaching far down the Irish Sea and covering the Welsh landscape.  Draining the ice sheets would have been rivers that flowed across a barren plain in front of the sheets.  Because of the ice global sea level was much lower than today by up to 120m and so this part of Cardigan Bay would have been land following glacial retreat.  Where the rivers coalesced large channels formed which cut deep into the till as the climate became warmer.  By the Mesolithic, it was these landscapes criss-crossed by the rivers that people would have migrated back into before they became covered by the rising sea. 

Some modelling work that I had done a few years ago was used to test where the rivers had once flowed and it proved remarkably insightful.  The extent of the deep erosion of rivers however was something that totally blew us away!  Channels 20-30m deep contain multiple fill events suggesting a complex drainage history to the Welsh landscape.  There is clearly a stunning story to be pulled out of this area that will need a lot more geophysics to understand.  We will also need a lot more coring as our attempts at this proved not as successful as the geophysics due to drift of the boat in the currents and wind.  The cores that were taken will now go back with my brother to his lab for analysis and we hope to return to the site next summer for more geophysics and coring.  Ultimately these core will be examined for a whole range of environmental signatures including some DNA analysis that could potentially reveal what the landscape was like and what or who might have inhabited it!

Channel offshore Aberaeron

The Celtic Voyager and willing crew made the cruise a real pleasure, though I am going to have to go on a serious diet – how the crew do this month after month is beyond me or would be beyond my will power when it came to meal times!

This project is part of the Lost Frontiers EU research grant led by Prof Vince Gaffney at Bradford University.

Multibeam sonar -  Konsberg EM2040 
Pinger - SES500 Geopulse
Sidescan sonar - Edgetech 4200

Bates, C. R. and Bates, M. R. 2016. Palaeogeographic Reconstruction in the Transition Zone: The Role of Geophysical Forward Modelling in Ground Investigation Surveys. Archaeological Prospection, v. 23, pp. 311-323

Clark, et al., 2009. The Last Glacial Maximum. Science, v325. 710-4. 10.1126/science.1172873

Fitch, S and Gaffney, S. 2011. West Coast Palaeolandscape Survey

The Mysteries of Gilmerton Cove

The Mysteries of Gilmerton Cove

Back in the mid-80s I lived in Edinburgh while studying geology at the University of Edinburgh (happy days!).  I would like to think that during that time I got to know the city quite well and I certainly made sure investigate many dimly lit, subterranean hangouts (Bannermans comes to mind as a place frequently visited but full of hazy memories), however I never heard of the Mystery of Gilmerton Cove. 

Gilmerton is a suburb on the south side of the city and is home to a series of caves or tunnels that are not on the main tourist route.  The tunnels were for long assumed to have been the work of an 18^th century blacksmith, George Paterson.  After having been caught selling liqueur from them on a Sabbath (he blamed this on his wife) he claimed that he had dug the labyrinth over a five year period as an underground dwelling for his family.  There are a number of “rooms” in the tunnels with stone tables, benches and even some that look like stone beds. In the walls there are markings suggesting that some of the rooms were separated by, presumably wooden, doors and curiously there are various skylights at strategic places.

http://www.bbc.co.uk/news/uk-scotland-edinburgh-east-fife-38991458

The Cove was well known in the late 1700s with local historian Rev. Thomas Whyte noting their unusual construction. Various recordings have been made since with the first investigation undertaken by F. R. Coles, Assistant Keeper of the Museum in Edinburgh in 1867.  He described the construction of the caves in an article in the Scotsman in 1906 as having been dug using pointed chisels and further commented that it would have been unlikely for Paterson to have been able to create them by himself in just 5 years.

If not Paterson, then who had constructed the tunnels?  The area is one that has long been associated with mining and as far back as the 16^th C there would have been ample and adequate labour to have dug them but for what purpose? 

The walls today are without marks of soot and with virtually no inscriptions apart from one small cross it is hard to imagine what went on in them.  Dating structures like these is notoriously difficult and so all we can do is compare them to similar structures elsewhere and better understand the layout of the caves themselves. 

sketch plan of caves

In the 1970s excavations revealed blocked passageways suggesting that the tunnels extend beyond the current layout.  So finally this is where the geophysics comes in.  About a year ago Dr Simon Shackley and Prof Stuart Haszeldine at the University of Edinburgh asked if I thought any geophysical technique might be appropriate for exploring the extensions to the caves and I suggested that ground penetrating radar would be our best bet in this very noisy urban situation.  The main problems for the site are that to the west is built up with housing, to the north three shops have been constructed over the tunnels and to the east is a very busy main road. Combined with this the pavement and road areas are riddled with drains, pipes and buried cables.  Not ideal for high quality geophysics!

The NERC Geophysics Equipment Pool provided a Sensors and Software Radar and we used this with 500Mhz antenna.  A station spacing of 10cm along lines was used and we placed lines at 30cm apart down the pavement and as far into the road as we could safely do. The results somewhat surprised me, especially those on the pavements either side of the main road.  Here reflections were recorded in the radar data that are consistent with a void-like structure at 2-3m in the subsurface.   

These seem to align across the road, however the significant gap in data where we could not survey still means that we do not really know what the extension might look like.  At the other end of the tunnels the picture is not so clear as the ground is very uneven and to achieve better quality data would require a significant amount of ground preparation.  So a little bit further with the mystery but certainly no great Scottish enlightenment yet – just like the old days coming out of Bannermans.

We hope to be back and continue the work but as with most projects, funding is now needed plus cooperation with the council to get a few hours of road closure would help greatly.

Searching for evidence of WW I German occupation in Namibia

Searching for WW I German armaments in Namibia, 2017

So I have had a few unusual survey proposals over the years, Yagan’s head comes to mind for example, but a call back in October led to a trip at the end of the year to Namibia.  At first I thought the request a little bizarre and one in which I would be of limited help but on talking the problem through I saw that perhaps geophysics might be able to add to the investigation.  

So what was the request? Well it started with a story surrounding the retreat of German troops from Namibia during WW I.  The story goes that as the German Schutztruppe ("protection force") left the country they disposed of various armaments in the sinkholes around the town of Tsumeb. In particular the lakes of Ojikoto and Guinas were suspected to contain field guns, rifles, carriages, ordnance and even (as always in these cases) a safe full of gold!  

German WW I field gun in Ojikoto Museum

This was too much to resist!!

The architecture of the caves and even their depths were in dispute despite a long history of local divers testing their prowess at exploration. Descriptions by local divers Chris Steenkamp and Johan le Roux (who provided incredible support throughout the trip - thanks guys) talked of overhanging caves extending unknown distances with jagged roofs and a very soft sediment on the lake floors.  The region is one of limestone and dolomite with a regional fracture pattern that shows linear trends to the lakes.  Along the weakened trend the lakes open up as sink holes and thus have steep to vertical sides.  These were going to necessitate some unusual access means and a very portable geophysical survey platform. 

Lake Ojikoto

To survey the lake floor and also to stand a chance of seeing something of the lake sides would require a sonar with a wide beam pattern so I took with me both a 468kHz and a 234kHz SwathPlus.  Motion reference and positioning was provided by a TSS DMS205, a Vector Cresent and Topcon HiperPro.  

The steep sides of Lake Guinas Lake meant that we had to abseil the inflatable boat into the sinkhole before setting up the equipment.  However, once on the lake the sonar produced fantastic maps of the underground structure showing the bell-like form of the sinkholes and also providing a series of targets for the dive operations. 

In Lake Ojikoto these were at 30-40m depth and proved to be a carriage gun and boxes of shells however in Lake Guinas the depth of over 100m was way beyond my comfort dive range and so remains to be investigated with remotely operated vehicles (ROVs) and cameras.  

The sonar produced a really great map of the underside or roof of the sinkholes even where the angle was very small compared to the water surface but the data set did require some significant manual filtering as the geometries of the structure was anything but typical for sonar data. Still, it just goes to prove that often you really do not know what you will get with geophysics until you try!

"outer" walls of cave

point cloud view of cave walls

The “expedition” was in part sponsored by a US Travel channel so watch out for Josh Gates and Expedition Unknown to be broadcast on March 6^th in the US on the Travel Channel. 
Check out to view. 

Unfortunately, no gold yet but we will continue our hunt for the means to support the Schools future………

Geophysics at the Gorgan Wall, Iran

Hunting the Red Snake – part II

Back in Iran, this time with my brother, Martin, to continue the hunt for the remainder of the Gorgan Wall.  Last year I was here with marine geophysical equipment to try and find the remains of both the Gorgan Wall and the Tammisheh Wall as they enter the Caspian Sea as part of a project with the University of Edinburgh and the National Museum, Iran.  From a previous blog (GorganWall I) I showed the results for the Tammisheh Wall extending well over 2km from the southern shores of the Caspian into a large lagoon behind the Miyankale Peninsula.  The sidescan sonar survey and sub-bottom profiling showed the remains of the brick wall with its accompanying ditch/canal.  Unfortunately we so no manifestation of the Gorgan Wall extension – so back to try again.

The Gorgan wall projection to the west towards the Caspian Sea

The Gorgan Wall (the Red Snake) extends from the Alborz Mountains in the east across the Gorgan Plains and was built from the 4^th C AD as a defensive structure to keep out the marauding Turks from the north.  The wall was constructed of bricks fired from local clay in kilns spaced 40-80m apart along the length of the wall. It was at least 2m wide and tall expanding to a more substantial feature at numerous towers and where defensive forts occurred.  In front, to the north, of the wall was a 10-30m wide, up to 3m deep, canal.  The remains of the wall and canal are relatively well documented to the east. H,owever to the west, evidence of it remains elusive with historical writings describing the wall splitting into different parts and extending down to long-lost towns.

Despite being of similar age to the Tammisheh wall and being relatively close to each other the geomorphological setting for both is somewhat different.  The Tammisheh wall extends into the Caspian at a place where the sea is relatively sheltered with fairly benign sedimentation gently silting up the almost closed lagoon.  The projection of the Gorgan Wall however is across an area of coast where there have been highly dynamic changes due to the interplay between the Gorgan Ricer discharge and longshore drift bringing sediment onshore from the Caspian.  The present day shoreline demonstrates these active processes where the waves can reach impressive magnitudes due to a fetch stretching the length of the Caspian.

CMD Mini-explorer and Explorer ground conductivity meters

This year we have returned to attack the challenge of finding the remaining western part of the Gorgan wall as it progresses towards the Caspian Sea. We come armed with electromagnetic instruments, the GFInstruments CMD Explorer and Min-explorer.  Both are frequency domain ground conductivity meters, each with three coil spacings, in order to look to a range of depths.  We have a Leica dGPS for positioning control and for ground truth work a more traditional shovel and trowel.

The western limit of the wall is manifest by a series of “robber” pits that were used by locals to mine the decaying bricks for use in other projects.  The wall and remains of the canal that ran along the north side of the wall run in a straight course here for at least 20km.  Our first geophysical grids were laid over the robber pits and known wall/ditch locations in order to characterise their geophysical signature. The figure below shows this mapped on the satellite images.  Note that the background satellite images clearly define the agricultural system of 200x200m fields and also parts of the old natural landscape with its river channels and back barrier bars with sloughs.

EM Conductivity over robbed-out portion of wall

We continued to map the wall to the west and after approximately 1km a large modern drainage ditch was encountered and gave us the opportunity to ground truth the geophysics.  A 2D interpretation of the geophysics together with the surface mapping suggested that wall and ditch might exist here.  Excavation proved this to be true and samples have now been taken for analysis.

Using this characteristic signature mapping was continued a further 5km to the west with a complex pattern developing of linear features mixed with sinuous features.  Our current thinking has an interpretation based on a landscape evolution of shorelines behind which the old Gorgan River fights its way to the sea and through which the wall and canal/ditch is cut.  It is likely that the dynamic natural landscape was always an issue for maintaining the wall and canal here and a much larger geophysical and ground truth investigation is going to be necessary before a clearer picture of the history is revealed.

New geophysical results showing extension to the west from last known position

So, more geophysics to come hopefully in 2017. 

During the trip we also tried making maps with and AUV operated by Georgian colleagues.  While the purpose was to make maps along the wall projection we also took time to investigate some of the more interesting geological features.

Dormant mud volcano,Gorgan, Iran

This was way too short a trip as usual but a highly enjoyable one.  Our biggest issue this time was getting the equipment through customs – lessons for the next trip.  Our biggest hazard encountered (apart from the driving)  ........... watch out for those camels!