The project aims to rebuild the landscape, the ancient road system and the anthropic development in the northern part of the ancient Lombard Duchy of Friuli. The first part of the work was about the study and the analysis of several aerial photographs. We recognized and identified several traces of anomalies on the terrain. Each photo was georeferenced on a GIS and each trace was digitalized to create a complete database connected to all the castra found in Paulus Diaconus’ Historia Langobardorum and all the necropolis found during excavations from the 1800 century till our days.

As a completion of this work, several flights with an ultra light plane were made to produce oblique aerial photography. Every anomaly found during the flights was photographed and georeferenced with a portable GPS device, in correlation with the airplane GPS.

As a research on new technologies we tested a Unmanned Aerial Vehicle: the purpose was to test a low cost system for the creation of a 3D mesh regarding archaeological sites with open source program: Python Photogrammetry Toolbox, created by A. Bezzi and Meshlab, created by the University of Pisa.

The second stage of this research aimed to try to reproduce local  traffics during the Lombard period using spatial analysis. A DTM  with a 5m resolution was created as a base-map and variables to get the cost model were added: slope, altitude above sea level, hydrographical network, the line of sight. To each variable a weight was given according to the influence it had on the model: the slope had the greatest weight, then the rivers, divided according to their water flow, last the analysis of visibility.

 An anisotropic cost model was used where the cost dependent on both the direction of travel and the attributes of individual map cells. The roads reconstructed were compared with the hypothetical road network done with historical and archaeological studies. All results and data collected, were inserted into a WebGis created by Emanuel Demetrescu with open source programs.

A prediction model for answer a specific problem was also created: in our region we know where the Lombard died but we don't know where they lived. For create this model we used as variables the slope, number of necropolis and castles  into the territory, distance of the river, elevation, sun exposure, the road network. This model was created with a commercial software but thanks to Augusto Palombini, researcher in CNR, we made a similar work with an open source software.

The final part of the project was focused on the reconstruction of the landscape during the early medieval period in two samples sites, for which we have more information and archival data. To achieve this reconstructions historical maps and documents of excavations was used as well as pollen analysis which provide more detailed information in relation to the vegetative cover. All the archival data was inserted in a specific geographic software that generated the 3D reconstruction of the ancient landscape with a good approximation.