Drones don't fear lava: Remote and ground archaeological survey during the 2021 volcano eruption in La Palma, Canary Islands

By Francesc C. Conesa, Guest Associate Editor, Landscape Archaeological Research Group, Catalan Institute of Classical Archaeology

The study of how eruptive events affected past societies and their environment has been a recurrent field of research that has gained relevant insights towards reducing present-day geohazard risks and vulnerability [1, 2, 3]. Archaeological volcanology also has close ties with the so-called “disaster archaeology” [4, 5], which aims at improving our engagement and risk management strategies in response to contemporary calamities such as floods, hurricanes, earthquakes [e.g 6, 7, 8, 9] or even nuclear accidents [e.g. Fukushima, see 10]. Much of these works have focused on the social benefits of archaeological work amongst at risk-communities and the use of remote-based approaches to map endangered Cultural Heritage in post-event scenarios. But what about recording and surveying archaeology at risk during an emergency such as an ongoing volcano eruption? 

In mid-September 2021, a volcano erupted on the island of La Palma (Canary Islands, Spain) after a whole week of an intense seismic swarm. The eruption lasted for 85 days and ceased around mid-December, making it the longest eruptive event ever recorded on the island. The lava flows covered an area of 1200 ha in the south-western Aridane Valley, forcing 7000 people to evacuate their homes and ultimately destroying c. 3000 built-up structures and other infrastructures such as roads and agricultural lands. Due to its impact on a densely populated area, it has probably been the most monitored eruption in history. The emergency services and the scientific community largely used and benefited from daily data captured by Earth Observation programmes, drones and ground surveillance teams. 

Figure 1. Extent of lava flows in mid-November (left), and detail of photo-interpreted ethnographic enclosures and huts from aerial imagery predating the eruption.

The rooted history of agro-pastoral land use in the Aridane valley shaped a rich and diverse cultural landscape dotted with indigenous settlements and rock art, agricultural plots and land features dating back to early colonial times, and hundreds of traditional Canarian housing. During the early days of the eruption, however, little was known about the disappearing cultural landscapes in the area since the region had only been partially surveyed and public heritage inventories and data catalogues were scarce. As the unstoppable lava flows advanced eastwards into the valley, the surrounding areas faced a continuous heavy ash deposition that triggered the visibility and preservation of many open-air archaeological and historical locations. 

In such an emergency context, we conducted a multidisciplinary and multi-scale research project aiming to rapidly and effectively map, monitor and, to our best extent, safeguard known and new features of archaeological, historical and ethnographic interest. Our work combined historical and geographical legacy data (e.g. historical aerial imagery and cartography that predates the eruptive event), coupled with newly acquired satellite imagery, targeted drone mapping and ground-truth surveys. 

Figure 2. Ground documentation of endangered traditional housing (left), shortly after it disappeared as shown by the red rectangle in the drone imagery (right). 

We also extensively used third-party, free and openly accessible geospatial data specifically created for the remote monitoring of the eruption. Among those, we highlight the Copernicus Emergency Management Service (EMS). Shortly after the first hours of the outbreak and over the next three months, the EMS created a record-breaking number of 64 mapping products under its activation key #EMSR546. EMS products contain raster and vector information with updated reports on volcanic activity and affected built-up infrastructure. Observations are primarily based on optical satellite imagery (e.g. the short-wave infrared band of Sentinel 2 highlight the heat of lava flows and their extent) as well as radar data ( e.g. Sentinel 1 and Cosmo Sky-Med SAR channels penetrate the dense ash cloud column and may inform about land surface change throughout the affected area). 

Figure 3. An unusual setting for conducting an archaeological survey. 

In addition to EMS mapping products, the Government of La Palma unified all the newly created geospatial data in Riesgo Volcánico, an ArcGIS-based web portal with access to daily drone imagery acquired by the emergency response teams. The use of these products only emphasises the practicality and efficacy of open geospatial information offered by all actors involved in the management of disasters. At the other end, the scientific and user community tracked and monitored the event as it happened, thus allowing us to identify and prioritise the most vulnerable areas in a timely basis. 

As a result, our drone flights and ground surveys were planned according to the eruption dynamics such as the active lava flows and their total extent and, undoubtedly even more important, the air quality parameters to avoid harmful gas concentrations such as CO and SO₂. Indeed, perhaps what most differs from other remotely-based survey projects is that we conducted field visits inside the exclusion zone. It was a restricted area encompassing all the evacuated lands, where scientific personnel were granted access in close coordination with the emergency command office and under strict safety measures, such as wearing gas masks and portable gas detectors at all times. 

Figure 4. Drone imagery showing the heavy ash deposition in the southern part of Las Manchas. The yellow rectangle indicates a potential indigenous settlement.

Although we could only access the exclusion area shortly after the first month of the eruption, c. 300 features of interest were remotely detected in the Aridane Valley. Those are mostly related to indigenous and historical huts, pastoral enclosures, traditional housing and a large network of historical paved paths that survived the urban and touristic expansion in the valley in recent times. A series of landmarks inside a 200m buffer along the area covered by previous and new lava flows were ground-truthed to assess its significance and potential future preservation under ash deposits. When possible, critically endangered features by encroaching lava flows were documented using drone-based photogrammetry. 

At present, the Aridane Valley and its inhabitants are slowly recovering from the eruption’s hugely social and economic impact. On our end, we continue to analyse our spatial datasets in the context of the new post-eruptive scenario while working in upcoming publications and collaborative data-sharing strategies with the local and national heritage authorities. Furthermore, the lessons learned in La Palma might advance procedures and guidelines for the future safeguarding of heritage at risk in the Canary Islands and in other similar emergency scenarios elsewhere. 

The research in La Palma was jointly coordinated by the Landscape Archaeology Research Group at the Catalan Institute of Classical Archaeology, the University of La Laguna and the Benahoarita Archaeological Museum, with the support of the Canarian Government (DGPC 94/2021-1117123435). We thank the support and field guidance of Prof. Ramon Casillas, volcanologist at the University of La Laguna and member of the PEVOLCA Scientific Committee. 

References

[1] Elson, M., Ort, M., Archaeological Volcanology, 2018. En López-Varela, S. (eds.), The Encyclopedia of Archaeological Sciences. Wiley. 

[2] Riede, F., Doing palaeo-social volcanology: Developing a framework for systematically investigating the impacts of past volcanic eruptions on human societies using archaeological datasets. Quaternary International. 499B:266-277.

[3] Riede, F., Barnes, G., Elson, M., Oetelaar, G., Holmberg, K., Sheets, P., 2020, Prospects and pitfalls in integrating volcanology and archaeology: A review. Journal of Volcanology and Geothermal Research, 401. 

[4] Torrence, R., Graham, J., 2002, The archaeology of disasters: past and future trends. En Torrence, R., Graham, J. (eds.), Natural Disasters and Cultural Change. Routledge. 

[5] Riede, F., 2017. Past-Forwarding Ancient Calamities. Pathways for Making Archaeology Relevant in Disaster Risk Reduction Research. Humanities 2017, 6.

[6] Buren, M.V., 2001, The Archaeology of El Niño Events and Other “Natural” Disasters. Journal of Archaeological Method and Theory, 8:129–149. 

[7] Bagwell, M., 2009. After the Storm, Destruction and Reconstruction: The Potential for an Archaeology of Hurricane Katrina. Archaeologies: Journal of the World Archaeological Congress 5,1:280-292.

[8] Tapete, D., Cigna, F., 2020. Poorly known 2018 floods in Bosra UNESCO site and Sergiopolis in Syria unveiled from space using Sentinel-1/2 and COSMO-SkyMed. Scientific Reports 10, 12307.

[9] Jusseret, S., 2014. Earthquake Archaeology. A future in ruins? Journal of Contemporary Archaeology, 1.2: 277-296. 

[10] Schlanger, N., Nespoulous, L. & Demoule, J., 2016. Year 5 at Fukushima: A 'disaster-led' archaeology of the contemporary future. Antiquity, 90, 350:409-424.