For a bit of professional development (and general interest) I took a Certificate of Higher Education module in Archaeology at Birkbeck College in the winter months leading up to Christmas last year. The course was great, with a wide range of topics and lively seminars run by our tutor, Dr Stuart Brooke - I'd thoroughly recommend it to anyone who is interested in getting a practical grounding in Archaeology. We had a choice on final essays, and I did mine on aerial imagery and how it can be used by archaeologists. If you'd like to see a practical demonstration of photogrammetry, one of the techniques I discuss in the essay, you can sign up for a webinar on X-Photo through this link.
And my essay can be read below:
Discuss the strengths and weaknesses of aerial photography as an archaeological reconnaissance technique.
Aerial and satellite imagery is now so routinely used in our lives through services such as google maps that it would be difficult to conceive of any archaeological project that would not make use of aerial reconnaissance in some way. Along with this huge expansion in accessibility, the last decade has enjoyed significant technological developments that help to improve the quality of the data available particularly through advances in laser scanning, drones and photogrammetry. With that said, as with any method of investigation, there remain difficulties in interpretation and inherent biases in data collection that archaeologists must be cognisant of, and where possible, mitigate. With that awareness, aerial photography and its allied digital imaging techniques can be used throughout a project, helping to inform further investigation. Aerial imagery then goes on to play a vital role in the presentation of the project, providing visualisations that help to drive public engagement as well as serving as a basis for Geographic Information Systems (GIS) that remain central to the practice of most modern archaeology. Finally, it provides an essential resource for heritage management as sites are monitored from the air for signs of damage over time. As the technology involved for aerial photography (and this will soon most likely incorporate advancements such as lidar - or Light Detection and Ranging as standard) becomes cheaper and more readily available, its role will continue to increase, making it even more important that archaeologists recognise the strengths and weaknesses of it as a reconnaissance technique.
Aerial photography has a long history of use in archaeology (Bourgeois & Meganck, 2005), being popularised in Britain by pioneers such as O.G.S. Crawford, who spent most of his career as the archaeological officer of the Ordnance Survey. Digital natives may take the ability to view any location from space almost for granted, but then as now it provides a revelatory new perspective: ‘It is the great merit of air-photographs that they reveal earthworks upon ploughed land which are invisible to the observer on the ground, or which appear to him as a confused tangle... From the air an orderly system is visible.’ (Crawford, 1923, 342). As Crawford explained, some archaeological features will only make sense when viewed from above, with wide ranging examples from the simple layout of a building to the Nazca lines in Peru. The camera can emphasize small differences in ground conditions caused by material remains that may not be apparent to someone at ground level. An experienced analyst of aerial photography will look for how surface water behaves in the area, for marks in crops, soil, frost, and shadows being cast by small differences in ground level. The image will either have been taken in the oblique or vertical with consequent advantages and drawbacks to each, though the vertical image is generally better for making maps, and the oblique for perspective (Renfrew & Bahn, 2020, 82). With experience and knowledge of the context, it is then possible to identify the archaeological evidence from more modern or natural features. Even as the technological tools available to the archaeologist multiply, this ability to discern the relevant features from images remains essential. As many landscapes will have been used in a plethora of different ways over their history, disentangling the remains from different periods can be challenging and require further investigation either from the air or the ground, but aerial photography is a crucial resource in understanding how each site fits into its wider landscape context as well as for discerning where would benefit from further study.
One of the most immediate advantages of aerial photography is that unless the project is subterranean or submarine, there is already likely to be some aerial imagery of the area of archaeological interest. Even in the former cases, it is no longer impossible; Unmanned Aerial Vehicles (UAVs) can be utilised to explore caves (Zhang et al., 2017) and airborne lidar can reveal the seabed provided the depth is not too great. In the latter case, marine archaeology has seen a step change in 3D visualisation since 2009 (McCarthy et al., 2019), where aerial lidar has been combined with other techniques such as bathymetry and multibeam sonar to create detailed models that can be navigated not just by researchers but also the general public at places such as Suomenlinna museum in Finland (Koivikko, 2017, 22). These are examples of specifically-commissioned flights necessitated by the challenging conditions for these sites, but for most locations the site can be explored from the archaeologist’s desktop using the high-resolution satellite and air photography available from Google Earth. So much so that Renfrew and Bahn have labelled the introduction of Google Earth as a true “aerial revolution” , citing inter alia that ‘in 2008 it revealed 500 new caves in South Africa alone’ (Renfrew & Bahn, 2020, 87). With the variety of data publicly available from IKONOS, QuickBird, LANDSAT, GeoEye and the Cold War CORONA satellites before even visiting a site, we can see why Andrew Bevan described archaeology as being subject to a ‘data deluge’, facing ‘floods of new evidence about the human past that are largely digital, frequently spatial, increasingly open and often remotely sensed.’ (Bevan, 2015, 1473). These developments have been hugely beneficial in saving time expended on basic surveying but present their own challenges in sifting through the voluminous data available and interpreting it.
With so much extant aerial imagery, some might argue against the creation of more to add to the mountain of data, but such an approach would leave the researcher at the mercy of certain inherent biases in the data. Whether in the case of satellite or conventional aerial photography, the images are just snapshots of that time under those conditions, which change constantly. This is seen whenever the UK faces drought conditions and crop circles emerge as a good news story in an otherwise unwelcome situation (BBC, 2018). Other than soil aridity, there are many factors that will affect what can be seen from the air including the crops, vegetation, climatic conditions and the nature of the material remains.There will be deficiencies in what can be viewed, but these can be mitigated by cross-referencing or informed by further investigation. In 1906, Stonehenge became the first archaeological site to be intentionally photographed from the air (Historic England/English Heritage, 2015) and there have been plenty more in the century that followed, whether they were from the Ordnance Survey, military planes or amateur enthusiasts. They allow an archaeologist to not only increase the number of ‘snapshots’ under different conditions (and it’s not just drought that can be useful - snow can also enhance the visibility of earthworks) but can also be used to show change over time. In the case of a palimpsest landscape, this can help to discern more modern remains or may display earthworks that were later levelled or destroyed (Janik et al., 2011, 14).
It should perhaps go without saying that not everything can be seen from the air and so archaeologists have to work with an idea of scale and a clear sense of their research objectives when approaching their study. For example, long barrows have been the subject of major archaeological interest in the UK because their surviving earthworks are often visible from the air (and on the ground). But it would be a mistake to just rely on what is immediately visible as ‘excavation or further investigation sometimes reveals a greater number of ditch circuits than visible on aerial photographs; and interpretation can be confused by the large overall size of these monuments, meaning that not necessarily every part of the site will be visible on aerial photographs due to different ground or lighting conditions.’ (Janik et al., 2011, 26). Aerial photography is most useful when it is interpreted in context and then used to inform further targeted investigation. And that further investigation can also come from the air. First of all, photogrammetric techniques can be used to make up for the lack of georeferencing in existing aerial photography. In areas without heavy vegetation cover, Structure from Motion (SfM) can be used to create highly accurate Digital Surface Models (DSMs) (Historic England, 2015, 5) allowing 3D mapping with affordable low altitude devices (Fonte et al., 2012).
Another major innovation in aerial reconnaissance has been Airborne Laser Scanning (ALS). It has been used across the world to provide insight into material remains that would previously not have been possible, but it comes with its own challenges. First of all, it must be ascertained whether it will actually provide any new perspective, given the aerial resources already discussed. Just like traditional photography, laser scanners do not penetrate the ground so ‘if the archaeological features of interest are not represented on the ground surface, then lidar will not be able to record anything except the general topography of the area.’ (Historic England, 2015, 29). However, with the advent of Full Waveform (FW) laser scanning it can penetrate vegetation, making it hugely effective for use in areas that would otherwise remain hidden from a camera lens by the canopy. Though the technology has become less expensive in recent years, the next consideration on whether to commission a lidar survey will be budgetary. And finally, the expertise for interpretation and data management afterwards need to be in place before any flight takes place. With those caveats, it can be seen that ALS can prove hugely useful as at numerous sites in Ireland that were scanned as part of the Discovery Programme. At Brú na Boinne they undertook “the most successful and far reaching lidar project… in Ireland to date.” (Opitz & Cowley, 2013, 151), confirming 27 new monuments that were added to the Record of Monuments and noting over 100 possible sites for further investigation. The lidar data was then used to publish a research framework for the area and was added to Meath County Council’s GIS dataset. At the Hill of Tara, it was used to see the archaeological remains beneath the trees in an area that had already been photographed and subject to photogrammetric techniques, proving how ALS can supplement previous aerial photography. The lidar models were then utilised by NUI Galway and the Discovery Programme to identify zones of potential for a further campaign of geophysical survey, resulting in a major new discovery of a medieval component to the settlement. (Opitz & Cowley, 2013, 151). Here we see aerial photography being used for initial reconnaissance, followed by ALS and then targeted terrestrial investigation that yields a better understanding of the site, proving that aerial techniques are best employed in coordination with others.
In some cases however, aerial prospecting may be the only option available due to the nature of the site. Returning to Ireland, where the World Heritage site of Skellig Michael is only accessible in the summer and even then by precipitous paths, some of which have become dangerous through erosion, aerial photography, photogrammetry and then ALS have been crucial in developing detailed and accurate plans of the main complexes. From the Digital Surface Models generated there have been new discoveries such as previously unidentified continuations of steps, and improved definition of known archaeological features, often in places where access on the ground is not possible or extremely dangerous (Opitz & Cowley, 2013, 153). Difficulty of access to a site is not limited to topographic conditions such as Skellig Michael, but can depend on legal rights over private land, which vary from country to country or the political and security situation. In Syria and Northern Iraq, their rich archaeological heritage has faced unprecedentedly severe threats from looting, combat damage and deliberate demolition by groups such as ISIS attempting to erase the pre-Islamic past. This in turn has stimulated efforts to monitor the rate, severity, timing and location of this damage using satellite imagery. As well as the data providing a valuable resource to heritage professions in ongoing reconstruction and damage mitigation efforts, it can be used to help curb the illicit antiquities trade by exposing the scale of the looting (Casana & Laugier, 2017).
In each of these modern examples, a key factor in using the aerial imagery effectively was how the data was processed. First of all, it is more helpful if it is fully georeferenced. Luckily, the tools available to archaeologists for this have advanced substantially so that even older aerial photos can be manipulated projected as raster layers over survey data (if taken in vertical rather than oblique). It has already been described how 3D models can be created from photogrammetry and lidar but the next stage in the process is using a GIS programme to manage the resulting point clouds: moving from the data to work out how people lived in the past or how the landscape has changed over time. In our most recent example, DigiGlobe satellite imagery from the Syrian Civil War archaeological monitoring project was streamed directly into ArcGIS. This meant that researchers could turn on and off individual images as they logged observations, easily comparing photography from before the war with the most recent image to see signs of damage. But GIS programmes are not just used to document what is lost, they play an important role answering practical questions about how people navigated the space they lived in. Journey times in modern landscapes are vastly different to how they would have been experienced in previous time periods and accurate feature mapping and topographic data collected from the air can help archaeologists to overturn assumptions about the routes that people would have taken when moving about the space as it was when they encountered it. Tools such as cost surface analysis can be useful to ascertain ancient pathways, which may be confirmed to some extent using aerial photography by them coinciding with modern trackways (Bell et al., 2002, 185). With more data to enrich the GIS alongside the aerial photography, such as geophysical survey and artifact mapping, we are better able to conceive of the landscape as a whole rather than being limited to single sites. Though as described earlier, this is not without its pitfalls, as Brouwer Burg and Howey argue that ‘archaeological modelers must beware of incautious data selection. Targeted, thoughtful, and reflexive data selection must be implemented to help us make the best use of the double-edged sword of big data.’ (Brouwer Burg & Howey, 2020)
Sometimes the data is just too much for one person to be able to make a considered selection, for example in satellite imagery of the entirety of Peru. Building on her work in Egypt documenting and discovering new sites through analysis of satellite imagery (Parcak, 2007), Sarah Parcak set up Global Xplorer, an organisation that set out to crowd source the archaeological analysis of a huge volume of satellite photography. The aim was to identify sites that were at risk as well as discover new ones. Volunteers sift through the images - with the first stage being the elimination of ones that simply show cloud cover - and look out for patterns that would indicate material remains. Locations are kept anonymous in order to avoid fuelling the looting that it seeks to curb, and over the project ‘over 70,000 citizen scientists [analysed and voted on] 14,620,932 individual satellite images of Peru’. The project announced that the crowd ‘identified 19,084 features of archaeological interest, dating all the way back from Caral, the oldest civilization in Peru c. 3200 BCE, all the way to the fall of the Inca in 1572 CE’ (GlobalXplorer, 2018). One of the most headline-grabbing achievements of these “armchair archaeologists” was the discovery of fifty new lines in the Nasca-Palpa region, which were then more fully documented using drones. Currently humans remain more efficient at the task of accurately recognising archaeological features from the imagery but the pace of change in machine learning and AI means that may not be the case for long (McCoy, 2020, 108).
When aerial archaeology was in its infancy, it probably wasn’t conceivable that archaeologists wouldn’t visit their area of study before jumping straight to a bird’s eye view but it’s more likely than not in an age where we carry detailed GIS computers in our pockets and use them to navigate our world. With so many freely available resources and limits on time and budget, it’s understandable that the modern archaeologist can be over-reliant on remote sensing rather than walking in the field and experiencing the sights and smells for themselves. There might be good reasons for not visiting areas like the cliffs of Skellig Michael or the battlefields of Syria, but where there is access, archaeologists should make use of their walking boots. The chief advantage of an airborne view is that you gain a new perspective, but this should not be at the expense of our usual one. Experience on the ground identifying earthworks feeds back to expertise in identifying them from the air and it must be remembered that we might be imposing order from above that does not exist below. Landscapes are multi-layered and we would do well to use technology to help us distinguish between them rather than adding another on top to obfuscate. ‘It is difficult to place a value on the familiarization and understanding the field surveyor gains by spending time on the ground examining the landscape and features.’ (Corns & Shaw, 2009) Ultimately aerial photography, even when supplemented with lidar can only see things that have left their mark on the surface, so excavation or geophysical survey will be required to understand what is not obvious from the air.
Aerial photography, film and laser scans can be stunning and open the world’s archaeological sites to anyone with an internet connection. With archaeologists often having to justify their work to public funding bodies, the production of engaging videography alongside general reconnaissance is not a minor benefit, in many ways, it is central to the practice of modern archaeology. As well as generating interest and headlines through the use of innovative technology, they serve a practical purpose in helping to explain the project to the communities around it. They can explore the site without physically going there (with the advantage of avoiding attendant visitor damage to the site). It can also be seen as it was in earlier time periods, constructed in virtual reality from the lidar data collected. The Office of Public Works in Ireland have also found such data to be more engaging than plans and maps when liaising with those farming around the Hill of Tara (Opitz & Cowley, 2013, 151). The field of aerial archaeology is progressing as rapidly as the technology develops, and we can see it being used effectively from initial prospecting through ongoing survey, management and public presentation of findings. In all cases though, it is best when used in a collaborative manner, whether that’s through crowdsourcing “armchair archaeologists” to overcome the difficulties of a data deluge or following up on aerial photography with other techniques that benefit from its initial targeting. As previous examples have shown, aerial photography is most useful as one component of a broader range of study, with its great strength being that sites can first be found, even within a vast search area, then placed in their landscape context rather than seen in isolation. Used properly, it can create one of the most exciting resources for presenting archaeological work and visualising the past.
BBC. (2018, July 17). Hidden landscapes the heatwave is revealing. BBC News. https://www.bbc.co.uk/news/uk-44767497
Bell, T., Wilson, A., & Wickham, A. (2002). ‘Tracking the Samnites: Landscape and Communications Routes in the Sangro Valley, Italy’. American Journal of Archaeology, 106(2), 169-186. JSTOR.
Bevan, A. (2015, December 07). The data deluge. Antiquity, 89(348), 1473-1484. https://www.researchgate.net/deref/http://dx.doi.org/10.15184/aqy.2015.102
Bourgeois, J., & Meganck, M. (Eds.). (2005). Aerial Photography and Archaeology 2003. A Century of Information. Archaeological Reports Ghent University, 4.
Brouwer Burg, M., & Howey, M. (2020). Unbinding Diversity Measures in Archaeology Using GIS. Journal of Computer Applications in Archaeology, 3(1), 170-181. http://doi.org/10.5334/jcaa.55
Casana, J., & Laugier, E. J. (2017, November). Satellite imagery-based monitoring of archaeological site damage in the Syrian civil war. PLoS ONE, 12. https://doi.org/10.1371/journal.pone.0188589
Collyns, D. (2020, May 24). Scratching the surface: drones cast new light on mystery of Nazca Lines. The Guardian. https://www.theguardian.com/science/2020/may/24/nazca-lines-drones-new-discoveries-peru#:~:text=New research with drones has,as much as 1,500 years.&text=“The Nazca Lines are the,older geoglyphs,” said Isla
Corns, A., & Shaw, R. (2009, December). High resolution 3-dimensional documentation of archaeological monuments & landscapes using airborne LiDAR. Journal of Cultural Heritage, 10, 72-77.
Crawford, O. G.S. (1923, May). Air Survey and Archæology. The Geographical Journal, 61(5), 342-360. https://www.jstor.org/stable/1781831?seq=1
Fonte, J., Fábrega-Ãlvarez,, P., Parcero-Oubiña, C., & Güimil-Fariüa, A. (2012). 3D mapping with affordable low altitude devices. A case-study on the documentation of archaeological features in the surroundings of an Iron Age hillfort in Northern Portugal. AARG - Aerial Archaeology Research Group Conference.
GlobalXplorer. (2018, April 10). GlobalXplorer° Completes Its First Expedition: What the Crowd Found in Peru. Medium.com. https://medium.com/@globalxplorer/globalxplorer-completes-its-first-expedition-what-the-crowd-found-in-peru-7897ed78ce05
Historic England. (2015). Using Lidar in Archaeological Survey: The Light Fantastic. Historic England. https://historicengland.org.uk/images-books/publications/using-airborne-lidar-in-archaeological-survey/heag179-using-airborne-lidar-in-archaeological-survey/
Historic England/English Heritage. (2015). Management Plan for Avebury World Heritage Site and surrounding monuments. http://www.stonehengeandaveburywhs.org/management-of-whs/stonehenge-and-avebury-whs-management-plan-2015/
Janik, J., Dickson, A., & Priest, R. (2011). An Archaeological Aerial Survey in the Cotswold Hills: A Report for the National Mapping Programme English Heritage. Report number: 123/2011, Project No. 4755(English Heritage). https://research.historicengland.org.uk/Report.aspx?i=15677
Koivikko, M. (2017). Recycling Ships: Maritime archaeology of the UNESCO World Heritage Site, Suomenlinna. Publications of Finnish Maritime Archaeological Society, Vol. 1.
McCarthy, J., Benjamin, J., Winton, T., & van Duivenvoorde, W. (2019). 3D Recording and Interpretation for Maritime Archaeology (Vol. vol 31). Coastal Research Library. https://doi.org/10.1007/978-3-030-03635-5_1
McCoy, M. D. (2020). Maps for Time Travellers (1st ed.). University of California Press.
Opitz, R. S., & Cowley, D. C. (Eds.). (2013). Interpreting Archaeological Topography (1st ed.). Oxbow Books, Oxford, UK.
Parcak, S. (2007). Satellite Remote Sensing Methods for Monitoring Archaeological Tells in the Middle East. Journal of Field Archaeology, 32(1), 65-81. https://www.jstor.org/stable/40026043
Renfrew, C., & Bahn, P. (2020). Archaeology: Theories, Methods and Practice (8th ed.). Thames & Hudson.
Zhang, G., Shang, B., Chen, Y., & Moyes, H. (2017). SmartCaveDrone: 3D cave mapping using UAVs as robotic co-archaeologists. 2017 International Conference on Unmanned Aircraft Systems (ICUAS), Miami(USA), pp. 1052-1057. http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=7991499&isnumber=7991298