It is not uncommon for current regional, state and linguistic boundaries to be practically meaningless for the people, places, and artefacts which are the subjects of archaeological study. Yet, for archaeologists, they can be significant barriers to the study of the past. Administrative entities (including funding agencies, local, regional, and national governments) determine the scope and focus of research programs, temporal distinctions, data collection and data-sharing processes that may be incompatible across these boundaries. For archaeologists, it is an increasingly a time-consuming and arduous part of the research process to navigate these barriers, especially when aggregating archaeological information which is typically a layering process of clues gathered over a long period of time at a single site.
ArkeoGIS (and later, ArkeOpen) was developed to address these barriers for archaeologists working in the Upper Rhine region which is at the nexus of three countries: France, Germany, and Switzerland. The goal was to compile and analyse data produced by many researchers in two languages (three if we add English), as well as within national disciplinary traditions that do not necessarily share the same periodization of the past. For example, if we focus only on the first millennium BC, several major sites can be found on both sides of the Rhine, including the Bronze and Iron Age necropolis from Haguenau in France and the Hügelsheim in Germany; the princely site of Münsterberg at Breisach on the German side and the Britzgyberg at Illfurth in France for the Hallstatt period; the oppidum of Saverne, the Roman cities of Strasbourg (FR), Baden-Baden (DE), or Augst (CH), to name but a few. As most of these sites are close to modern-day borders, it is difficult and time-consuming to conduct research on both sides of a border, even if that border had no meaning for the people who used to inhabit it.
The initial challenge in developing ArkeoGIS was to provide online access to cross-border, diachronic data on archaeological sites and artifacts (from prehistory to the present day), as well as to integrate (paleo)environmental data. An example of the need for such a database was to take into account the effects of taphonomy (the erosion or covering of archaeological sites) which can distort present-day maps and thus lead to biases in interpreting the data. While it is difficult enough to share data amongst archaeologists, it becomes even more difficult when working with researchers in multidisciplinary contexts such as environmentalists, whose expertise is necessary to take into account the effects of erosion for the Vosges and Black Forest massifs.
Initial funding from MISHA (a regional research infrastructure in Alsace, part of a nationwide network of Maisons des Sciences de l’Homme) was secured in 2007 to create a proof-of-concept for a database that focused on archaeological research in the Upper Rhine region. This allowed partners in France and Germany to apply for European funding through INTERREG (a funding scheme meant to stimulate cooperation between regions in the EU) which funded the first versions of ArkeoGIS: a multilingual, open-source webSIG. The technology for this first version was simple: spreadsheets (.csv) are shared by each institution, researcher or student, and part of the bibliography is converted to the same format so that all available information and its sources can then be queried online. The entire GIS aspect is managed in the browser, with no additional installation needed by the user, and the response to queries can be exported and reinserted into other processing software. This first phase was carried out by project partners with a unified relative chronology and a “bottom up” vocabulary.
The project, as with many Digital Humanities projects, soon outgrew its initial technology stack and infrastructure. Luckily, in 2010 ADONIS, which later became IR* Huma-Num (which now leads France’s involvement in DARIAH-EU) was able to provide data hosting and backup. But the project team also realized that our project lacked standards (we had only followed INSPIRE norms at the time) and controlled vocabularies to provide the level of interoperability necessary to do collaborative research in a digital age. Around the same time, a white paper written by digital archaeologists associated with the Consortium-HN MASA enabled the development and sustainability of the ArkeoGIS project over time.
The French Consortium-HN MASA (whose work is continued today by the Consortium-HN MASAplus) is a funded project by Huma-Num meant to help disciplinary communities adapt to the digital turn in research. For the ArkeoGIS team, Involvement with MASA was incredibly helpful, as were meetings with other like-minded colleagues to exchange best practices, standards and common challenges. Feedback to the numerous presentations of the project, both in France and Germany, as well as at European conferences notably the CAA and the EAA, was also critical to its development.
Over time and through a series of projects ( https://arkeogis.org/manual/software/ ), ArkeoGIS has gradually aligned itself with vocabulary standards – thanks to the highly scalable PACTOLS/OPENTHESO tools (both of which are French contributions to DARIAH), which now enable information to be described using concept identifiers and the Getty vocabulary (AAT GETTY), including chronology and localization which has enabled a more standard referenced and interoperable platform, in a word: FAIR. To achieve these objectives, concept alignments were necessary, made possible by a number of grants that kept the project afloat. The question of chronologies has also arisen, and while reference to an absolute chronology is necessary – particularly for multidisciplinary research – the possibility today for each author to present their research data according to their own logic is a non-negligible component for sharing archaeological data. Over time, several chronologies have been created, and lately we have been trying to push users towards pre-existing solutions, such as Perio.do. Last but not least, spatialization has also evolved considerably over time. Initially, it was crucial to offer a wide variety of map projections in order to meet community needs. Fortunately, since then, the World Geodetic System 1984 has become the global standard, and coupled with GEONAMES, it effectively replaces its predecessors, allowing for greater standardization.
Once we were able to standardise the data, it was possible to envisage an open, interoperable version of the project, ArkeOpen. As its name suggests, this evolution of ArkeoGIS (the 5th version) is founded on the best principles of open access and thus offers the possibility of even wider data harvesting. Tools such as OpenArchaeo, which can dynamically query several datasets using semantic web (which itself can only work with controlled vocabularies such as OPENTHESO) made it an essential part of the ArkeOpen infrastructure. ArkeOpen is fully integrated into the MASAplus Consortium-HN ecosystem allowing the interoperability with other tools and services, such as ArcheoVIZ,which is specialized in vizualizing prehistoric artefacts,or Epicherchell for North African epigraphy, both of which are also MASAplus resources. The current version of ArkeoGIS includes a deeper focus on reciprocal sharing of data and interoperability with other archaeological platforms and tools, Indeed, connecting archaeological research projects, whether they are past, current, or future, is much easier today and allows for rapid entry on many subjects, not the least of which would be quickly understanding the state of bibliography & archaeological research!
Let’s return to Figure 0 of the Upper Rhine Valley. Looking at the results, one might conclude that they make total sense given the path of the Rhine and the propensity for settlements to be located at the banks of rivers. Now, using ArkeOpen, an archaeologist can now create a map utilising funerary data (such as burial mounds) that are often considered markers for prehistoric roads (Bernard, Wassong 2016) and Bronze Age deposits with Roman roads in two different countries. Users use an interface in their own language, and can query all datasets, whatever their language of origin. What once would have been months of tireless work is now accomplished in a matter of mouse clicks. ArkeoGIS allows researchers to compare multiple datasets, such as the correspondance -or not- of roman roads and pre-Roman (1st millenium BCE) burial tumuli, which you can see in Figures 1 & 2, or in the short demo video accessible here.
The introduction of webGIS also solves the problem of cross-border map backgrounds – a complex issue for non-specialists – by making cutting-edge tools available without requiring a time-consuming training in specialised software. If you simply want to display points on a map, ArkeoGIS is all you need. Specialists, on the other hand, can easily export the results of their query to a spreadsheet and reinsert the data into their favorite tool.
As the project has evolved, other free and open information has been added to the database. This is particularly true for DARMC, hosted by Harvard University, but also for broader environmental science projects such as carbon-14 or pollen analysis, which are often difficult for cultural archaeologists to identify.
On the strength of the successes and stability of ArkeoGIS, the project opened up to colleagues working beyond the Upper Rhine, demonstrating ArkeoGIS’s ability to function as a generalist tool for all archaeologists. This flexibility, combined with the archiving of research data in the NAKALA, the research data repository operated by IR* Huma-Num, led to the researchers behind the HYDRΩMED and WATERTRACES projects to use ArkeoGIS as part of their DMP (Data Management Plan) in 2015. At this time, funding agencies had only just begun to require projects to create DMPs, and there were not yet any standardized, functional mainstream solutions as today (the Consortium-HN MASAplus has just updated a module in DMP-OPIDOR for archaeological data). Therefore, taking in account the data storage and interoperability features offered by ArkeoGIS, the research teams elected to use ArkeoGIS as a solution, as the data shared in ArkeoGIS are necessarily FAIR and their long-term preservation and accessibility is guaranteed through NAKALA.
In 2018, ArkeoGIS was then used as a reference tool by an archaeo-environmental team for their OT-MED and RD-MED projects. These projects, conducted by a team based at Aix-Marseille University, focused on modelling the impact of climate change on ancient populations through archaeological analysis and those of the environment through geography.
While working on pluridisciplinary modelizations, we eventually realised that raw data is of limited use for non-specialists. That is to say that, without deep familiarity with the disciplinary practices of archaeology, the the bio-climatologists could not make use of the archaeological data in ArkeoGIS as they (understandably) have no idea what kind of archaeological remains have a strong impact on soils or climate. This problem goes in the other direction as well, an archaeologist has no clue which pollen diagram is the most efficient to reconstruct the climate. Raw data that is specialised for a given disciplinary community is not necessarily reusable, even if it is findable, accessible,and interoperable!
To overcome this issue, we ended up creating reference datasets for non-specialists, for example, antique (i.e. 1st millenary BCE) sites that have a left heavy impact on their environment like silver or lead mines for example, so that a bio-climatologist is not led astray by the impact of ancient civilizations on a given soil sample when trying to make broad models of past climates (Fig.2). Sharing data sometimes means rethinking the data before making it available, so that it can be truly interoperable and accessible to specialists in our field and those outside of it. The data created during these projects is still accessible and used, notably by the ANR MICA project, which at the time of writing is still sharing digital data produced more than ten years ago (starting in 2015 with the previously mentioned OT-MED & RD-Med projects), a case rare enough to warrant mention!
Finally, the system’s stability, simplicity, multilingualism and multi-chronology, enabling everyone to work in their own language with their own standards, attracted the interest of colleagues in the West Indies, a multilingual area with teams often working at a distance. A team of French researchers, led by Alain Queffelec, working in the Antilles initially used ArkeoGIS, and now display their data on ArkeOpen. While some additions were necessary, notably a new chronology and a specific ontology for stone ornaments, ArkeoGIS & ArkeOpen work well for their needs: the project is now online and open for use.
The display in ArkeOpen has evolved over time, notably the automatic citation of datasets, chronologies and map backgrounds used in queries. This feature alone convinced many colleagues, often still reluctant to open up their data, to share it by guaranteeing them quick and easy citations for their research.
Like ArkeoGIS, ArkeOpen is based on the NAKALA data repository, maintained by IR* HUMA-NUM, which guarantees the security and continuity of shared data. This is a fundamental contribution to research projects, large and small. Indeed, in a world where research relies mainly on short-term funding, it is essential to have a guarantee that servers will continue to be available (maintained, up to date and secure) beyond the duration of the project funding. For researchers, it is also a guarantee of long-term access to datasets enabling reproducible research, which in the medium term should feed AI. In the short term, it’s also an excellent way of training students, engineers, archaeology professionals and researchers of conducting truly open and interoperable research, overcoming fears of data theft or plagiarism.
ArkeOpen makes it possible to rethink sharing, on a different scale and with different audiences. The tool requires robust maintenance, as the technical apparatus must balance multiple databases, multiple chronologies, and multilingual support, but the results are impressive, as the list of users, representing over 380 institutions in 37 countries and all continents attests ( https://arkeogis.org/utilisateurs/ ).
This small initial project is now integrated into the MASAplus Consortium-HN tools ecosystem, in particular Openarchaeo, which enables several other old or recent databases to be queried. Both ArkeoGIS and ArkeOpen are French contributions to DARIAH’s Tools & Services Catalogue, and are listed on the SSH Open Marketplace – both as such (ArkeoGIS & ArkeoOpen) and also as a workflow to help guide users in using the platform.
Loup Bernard, Mohammed Benkhalid et Jean-Philippe Droux, « Explorer les données archéologiques cartographiées et géospatialisées : quels outils et quelles pratiques ? D’ArkeoGIS à ArkeOpen : de la vallée du Rhin fermée au web ouvert de données, itinéraire d’une plateforme », Anabases, 39 | 2024, 353-359.
Bernard L., Wassong R., “Du Danemark au Fossé Rhénan. Un siècle d’analyse des voies de communications protohistoriques: évolution des méthodes et mise en commun des données”, in Camberlein C., Gentner S., Wassong R. (dir.), Archéologie des réseaux: économie, territoire, cartographie et spatialisation. Actes du séminaire interdisciplinaire d’archéologie de Strasbourg des 7 et 8 janvier 2015, Actualité de la recherche, revue en ligne ArcHiMedE III, 2016, p. 169-183.
(http://archimede.unistra.fr/revue-archimede/archimede-3-2016/)
Contreras DA, Hiriart E, Bondeau A, Kirman A, Guiot J, Bernard L, et al. (2018) Regional paleoclimates and local consequences: Integrating GIS analysis of diachronic settlement patterns and process-based agroecosystem modeling of potential agricultural productivity in Provence (France). PLoS ONE 13(12): e0207622. https://doi.org/10.1371/journal.pone.0207622
Schmitt, Laurent, et al. ” Holocene fluvial paleo-dynamics in the south-western compartment of the Rhine Graben (France).” Des fleuves et des hommes à l’époque mérovingienne, edited by Édith Peytremann, ARTEHIS Éditions, 2016, https://doi-org.scd-rproxy.u-strasbg.fr/10.4000/books.artehis.24169.
Loup Bernard, ArkeoGIS – eine neue und effektive digitale Plattform zur Sammlung. Darstellung und Auswertung archäologischer Daten im Elsass und in Baden. Archäologische Nachrichten aus Baden, 94 (2018), 40-48.
Alain Queffelec, Pierrick Fouéré, Jean-Baptiste Caverne (2020) A database of lapidary artifacts in the Caribbean for the Ceramic Age OSF Preprints, 7dq3b, ver. 3 peer-reviewed and recommended by PCI Archaeology https://doi.org/10.31235/osf.io/7dq3b
Olivier Marlet, Bruno Baudoin, Loup Bernard, Laure Bézard, Romain Boissat, et al.. Le Livre Blanc du consortium Mémoires des Archéologues et des Sites Archéologiques : Guide des bonnes pratiques numériques en archéologie. [Rapport de recherche] Consortium MASA. 2022, pp.38. ⟨halshs-03561376v2⟩
Daniel A. Contreras, Alberte Bondeau, Joël Guiot, Alan Kirman, Eneko Hiriart, Loup Bernard, Romain Suarez, Marianela Fader,From paleoclimate variables to prehistoric agriculture: Using a process-based agro-ecosystem model to simulate the impacts of Holocene climate change on potential agricultural productivity in Provence, France. Quaternary International,Volume 501, Part B, 2019, Pages 303-316,ISSN 1040-6182, https://doi.org/10.1016/j.quaint.2018.02.019. (https://www.sciencedirect.com/science/article/pii/S1040618216313568)
Kempf, M. Modeling multivariate landscape affordances and functional ecosystem connectivity in landscape archeology. Archaeol Anthropol Sci 12, 159 (2020). https://doi.org/10.1007/s12520-020-01127-w