The Citadel at Nimrud: An Interactive Virtual Reality Model

Paper presented in January 4, 2004, by Samuel M. Paley (University at Buffalo) and Donald H. Sanders (President, Learning Sites, Inc.) with the assistance of Richard P. Sobolewski (Polish Centre of Archaeology, Warsaw) and Thenkurussi Kesavadas (University at Buffalo) and funding from the White-Levy Program for Archaeological Publications, at the annual meetings of the Archaeological Institute of America, San Francisco, California. An early version of this paper was published in April 2004 as "The Citadel of Nimrud. A Virtual Reality Interactive Model as a Resource for World Heritage Preservation", in Enter the Past, Workshop 8, Archaologie und Computer, Staat Wien, held April 2003 (digital publication on CD-ROM with a hard copy summary).

posted here September 26, 2019


Assyrian sites mapNimrud citadel planSince 1998, we have been building a virtual reality (VR) model of the citadel at Nimrud, originally constructed by the ninth-century BC Assyrian King Ashurnasirpal II (see the map of Assyrian capital cities at the left; hover over to enlarge). We began with Ashurnasirpal's palace, the so-called Northwest Palace, and have now expanded to include the area of the eighth-century-BC so-called Central Palace of Tiglath-pileser III, as well as the temples and ziggurat in the northern part of the citadel (see the Nimrud citadel site plan at the left; hover over to enlarge). The evidence of ancient occupation is detailed enough to have embarked on the construction of a full-scale interactive VR model.

Onyx VR station at BuffaloPC version in use by studentsThe model is being built on both PC and Onyx platforms for distance learning across the Internet (see a view of the Onyx VR station at the left, showing a user with 3D glasses; and the PC version demonstrated at the AIA meetings in 2003; hover over to enlarge). Versions of the model can be viewed on the Learning Sites and University at Buffalo Virtual Site Museum's Web pages (ed. note 9/2019: the University at Buffalo Web page is no longer active).

broken lamassu at the NWPvandalized reliefs in the NWPThe deteriorating state of preservation at the site and the looting of its storerooms after the Gulf War of 1991-2, the subsequent UN sanctions which made site conservation difficult placing many of its monuments in jeopardy, the further looting of two pieces of bas-relief from the palace museum’s wall and the damage to several others last spring after Operation Iraqi Freedom, have made it extremely important that we continue to forge ahead to continue the digital documentation of Nimrud (see the two examples of such damage at the left; one to the arms of a lamassu, photo courtesy of Michael Weigl; the other looting of wall relief fragments, photo courtesy of Mark Altaweel; hover over to enlarge). We intend that this record will become one of the main resources for the preservation of Nimrud as a historical site and assist in its reconstruction as a site museum. This documentation will be available on the World Wide Web. The Website will detail the state of preservation and conservation at different periods of the site’s modern history, will assist in the recovery of stolen objects through the publication of all relevant published and unpublished material in one place, and will contribute to the future history of the site by creating a resource for the study of the history and culture of this ancient Assyrian capital.

The Central Palace part of the project is also the comprehensive final report of the Polish Center of Archaeology's excavations in the area of the Central Palace (1973 - 1976), supported with by a Levy-White grant from Harvard University.

We began the project with Ashurnasirpal’s palace since it was one of the best-preserved monuments on the site. We have over 150 years of various kinds of documentation just for this building alone. We have expanded our computer model to include the temple area north of the palace and the area just south of it, where two other buildings, one of Ashurnasirpal and one of his son, Shalmaneser III (late 9th c. BC), were found, and where the bas-reliefs from a very much destroyed (8th c. BC) palace of King Tiglath-pileser III were located. It was from this area that many of the bas-reliefs stolen from the site storerooms came. As money and time allow, we will eventually complete the reconstruction of the whole citadel. On the Web pages we have broken down the categories of information to include not only the VR models with the reasoning behind the model’s structures, but also the history of the excavation and photographic documentation of the excavation, conservation, and the present state of preservation.

 The Citadel at Nimrud Project

There are many reasons why a project like this is important for world heritage at the present moment in our history. As I stated a moment ago, because of the political situation in Iraq today, it is nearly impossible to maintain--both police and conserve--all the archaeological sites in Iraq sufficiently well to preserve very many of them; and then there are the vagaries of war, impending or in progress.

Alexander Stille’s recent work on the presentation of the past is fair warning of the many dangers to archaeological sites, through war, overuse or neglect. Speaking of Egypt, he writes: "We are faced with a dramatic paradox: soon we may have virtual realities of the tombs of the Valley of the Kings in classrooms all around the world--a genuine democratic advance--but the tombs themselves, most of which looked almost new when they were discovered in the late nineteenth and early twentieth century, may no longer exist" (p. xiii). We hope, of course, that VR will not be one of the only ways to remember Nimrud.

example of digital documentationBut, Nimrud is also a site from which, for over the last 150 or more years, cultural heritage has been dispersed to dozens of museums and private collections across the world, including those in Iraq. Many of the examples of bas-relief are fragments, broken or cut up, that were spirited away to be loaded on ships bound for Europe or America, either because 19th-century transport technology was not up to the task or, more recently, to obscure the origins of stolen pieces. Digital documentation is one rather inexpensive way to document the site from the excavation record and, in some cases at Nimrud, the database must be reconstructed from different styles of record keeping (see an example from our Northwest Palace model at the left; showing how photographic evidence can be used to accurately build textures in the virtual world; hover over to enlarge). Also, broken monuments can be restored digitally to approximate the way they were found and with good scholarly surmise even reconstruct their original state. In the event of a solution to the present situation in Iraq, this information can help in the restoration and conservation of what survives and even become part of the exposition of the monument at the local site museum when tourists return.

AIA kid's fairFor people who would never have the chance to visit the site or never could, a virtual model would present to them as near a real experience as they would ever have. And students would have a rich resource to use in learning about the archaeology of Assyria. If you visited the Kids Fair earlier today, you would have seen school-aged children working with a portion of the VR model and participating in activities related to the VR model and Assyrian history in general (see a photo from the fair at the left; hover over to enlarge). Therefore, there is a real need to document the site digitally as it now exists and collect the material that has been taken from the site over time, to bring every bit of information together into one virtual place so that the citadel’s monuments can be studied as a whole, or at least as close to the whole as discovered by the archaeologists who excavated it and those that interpret the finds.

AIA kid's fairA working drawing of our reconstruction of one room of the palace of Ashurnasirpal II is a good example--there are 10 collections represented in this one effort (see the paper reconstruction of the relief depictions from Room I of the Northwest Palace at the left where researchers try to assemble information about the size, shape, and subject of each fragment; hover over to enlarge).

different plans of the Nimrud citadelSimilar, though not as convoluted, archaeological reconstruction is needed for nearly every building on the site. Just figuring out the physical relationships among the rooms of the Great Northern Courtyard of Ashurnasirpal’s palace, the citadel wall and the temple area on the northwestern corner of the site is full of problems as the excavation publications differ in their various iterations (see various plans of that area at the left; hover over to enlarge). Talk about the need for good data. Scholarly surmise about the various reconstructions presents us with formidable choices and difficult decisions to be made among them. In some drawings, the width of the citadel wall is indistinct and in another it is established because of a scholar's well-reasoned argument. The plan of a building may differ because of the interpretation of the excavated evidence; the place of a road or street may alter. And yet we must make some decisions or our project cannot go on and its benefits, with all the caveats, will not be realized.

our massing model of the citadelLooking at a massing model of the northwestern corner of the Nimrud citadel illustrates the problems of reconciling the data, as the widths of the city wall and the heights of the palace floor paving, temple tower, temple floor paving and palace and temple walls become relevant (note the details of our massing model at the left; hover over to enlarge).

We must be and are sensitized to the problems of reconstructing an archaeological site in virtual reality, not only because we must, from time to time, make informed guesses but also because we agree with other scholars that we have only a part of the story and that we are making our own interpretations. Stone and Molyneaux have alluded to these kinds of problems implicit in “presenting the past.” And we must also deal with and be cognizant of what has been said many times, recently by Alexander Stille, about the field of archaeology and ancient studies that “…our vision of the ‘past’ is heavily conditioned and distorted by what monuments happen to have survived the ravages of time” (p.34).

If anything, digital archaeology and virtual reality reconstruction can provide the means to collect and study vast amounts of data and organize material that can be retrieved and presented in ways that provide us with new possibilities to see, interpret and present. As information is added and opinions change, the VR model can be easily corrected and the new data can be added to the database.

citadel planTP III evidenceWhen we came to the publication of the Tiglath-pileser palace, we had to deal with the problem of exactly where this building was situated on the mound and how it was related to the Upper Chambers, Adad-Nirari III’s extension to the Northwest Palace of Ashurnasirpal II, the temples of Ashurnasirpal II to the north and east and toward the center of the citadel and Shalmaneser III, which were re-excavated by the Polish Center of Archaeology in the 1970s (see the location of the structures, some of the excavated evidence of the structures, and our massing model of the remains at the left; hover over to enlarge each image).

Polish excavation teamThe Polish Centre of Mediterranean Archaeology field team (photo courtesy of Richard P. Sobolewski; hover over to enlarge).

Layard's Nimrud planRichard Sobolewski had already established the fact, in an article in the Archiv fur Orientforsching, that the walls of the Tiglath-pileser palace which we know best from the publications in the Barnett and Falkner volume were in fact the lines of the 19th-century excavators’ trenches and not walls at all. Barnett and Falkner’s reconstructed “walls” were based upon a close reading of Layard’s notes and the drawings, both his own and his assisting artists. Until new excavations are done on site, it will more than likely be impossible to create any meaningful palace VR models.

The Arab Campaign reconstructedWe can propose, following Barnett and Falkner, how the bas-relief decorations on the individual walls of individual rooms were organized and this is being done. We are re-cataloguing digitally all the bas-reliefs that were published by Barnett and Falkner from 19th-century excavations. We are adding "missing" pieces from the Polish excavations (see our preliminary mockup of the Arab Campaign at the left; hover over to enlarge).

reliefs left by Layard, uncovered by the Polish teamLayard did not send home every bas-relief fragment he drew or had drawn, nor did visitors to his excavation site after he and his assistant and successor Rassam left take everything they found in his excavation trenches. We are slowly filling out the pictorial repertory of Tiglath-pileser’s artists with the several hundred fragments of sculpture that were re-excavated by the Polish Center of Archaeology (see the cache uncovered by the Polish team, at the left; photo courtesy of Richard P. Sobolewski; hover over to enlarge).

Section of the Central Palace databaseA cataloguing system has been built which works in a very simple fashion to allow us easy retrieval of information about bas-relief and other artifacts. This will assist us in the identification of motifs for bas-relief reconstruction or place the bas-relief and the artifacts, both in their iconographic context or in their archaeological find place or original setting in the palace where possible real or reconstructed (see the screen grab of the Learning Sites Central Palace online database search page at the left; hover over to enlarge).

Section of the Central Palace databaseThe database is an interactive catalogue of the individual finds from the Polish excavation with such fields as, excavator, date of excavation, find place, and detailed descriptive characteristics, including iconography, size, and state of preservation. The search engine will allow a user to look for examples of the same type of object (e.g. bas-relief, corbel, vessel, and floor slab), architectural feature (wall, doorway, and pavement) and various iconographic representations of king, courtier, soldier, enemy, weapons, etc. The search engine feature of the database will be available also on the Website.


Preservation through digital documentation of the existing remains can not only give us a picture of the current state of the ruins, but can also document all known phases of the citadel's history. We think that the only efficient way to correlate the vast amounts of data in meaningful ways for retrieval, analysis, and teaching, is digitally, allowing customized research or collaborative research on the Web among scholars at different Web portals in immersive virtual re-creations of the site. There are traditional distance education possibilities and certainly localized classroom uses. In the virtual environment, a student or scholar could do things like take measurements or change or manipulate the world for personal research, just as if one was at the site in different phases of its occupation and deterioration, or if the scholar disagreed with our interpretations. Exposition in both conventional and site museum settings is possible. The possibilities are limitless. Perhaps this “brilliant new technology”, as Stille has called it, will enhance the study of antiquity and make us more informed than we ever have been before.

Close this presentation and return to the Index