The Late Roman defensive walls of Aquileia

Proxy models of the reconstruction 

Representation models of the Late Roman wall of Aquileia

View of the 3D reconstruction

One of the entrances located along the external wall

Since 2018, the University of Verona (prof. Patrizia Basso and prof. Diana Dobreva) is carring out its fieldwork in the southern sector of Aquileia (Fondi ex Pasqualis) where the presence of defensive walls, the so-called "M2" (dated from the 3rd century AD to the 4th century AD) and "M3" (built during the 5th century AD) was already known.

In 2019 a collapsed portion of the external surface of the "M2" wall was discovered. This evidence has allowed to propose a reconstruction of the elevation of the external wall, constructed on the bank of the Natissa river.

During the 5th century AD, another defensive wall ("M3"), smaller than the previous one, was built 10 meters towards the river. Entrances were placed along the wall in order to easily connect the mercantile squares with the river itself.

The reconstructive proposals of the "M2" and "M3" defensive walls were made considering: fieldwork data,  historical data, geological data, geophysics results, archaeological comparisons, litterary sources, and iconographic sources.

The Extended Matrix method was used for mapping, validating and representing the 3D reconstruction and the reconstructive process of the Late Roman defensive walls of Aquileia. The application of the EM method allowed to improve the analysis of both the archaeological context and the connected sources in order to suggest a virtual reconstruction based on scientific data, transparent and easy to interact with.

Extended Matrix

Ortographic view of the 3D reconstruction of the Late Roman wall 


The 3D reconstruction of the Late Roman defensive walls of Aquileia was the subject of a thesis in Quaternary, Prehistory e Archaeology at the University of Ferrara (title: “Le mura tardoantiche di Aquileia: l’area dei Fondi Ex Pasqualis. Dallo studio alla ricostruzione”) realized by Nicola Del Barba and supervised by Prof. Patrizia Basso (University of Verona) and Emanuel Demetrescu (DHILab, CNR-ISPC of Rome).