The geophysics work around Trajan’s harbour began in 1998, and has since been completed. The magnetometry results revealed: • A plan of the ‘Imperial Palace’ and its associated warehouses • A series of warehouses fronting the port • A part of the massive defensive wall circuit • The port’s major aqueduct, the location of which was previously unknown • Details of buildings to the southwest of the harbour, surrounding the basilica site and lining the canal which connects the harbour to the Fosse Traiana • More evidence for land reclaimation and property divisions at the confluence of the Tiber and Fossa Traiana.

The first season took place over five weeks during September and early October 2007, and was directed by Simon Keay and Graeme Earl (Southampton), assisted by Dott.ssa Lidia Paroli (Soprintendenza per I Beni Archeologici di Ostia), and involved participants from the universities of Southampton, Cambridge, Rome ‘La Sapienza’, Aix-en-Provence, Tarragona and Seville. It built upon two earlier seasons of topographic work (March 2007) and resistance tomography (June 2007). The excavations uncovered a large (250 mq) open area at the eastern edge of the Palazzo Imperiale, a key building at the centre of the port, revealing a large rectangular dock or canal that was probably of Claudian date, defined by a spectacular series of moles on the south side of the main Claudian basin of the port. This was filled with sand in the course of the first and second centuries AD, and its central stretch subsequently covered by a large circular building in the Severan period. The whole area was extensively replanned in the later fifth and sixth centuries AD.

Geophysics: The three year programme of magnetometer survey on the Isola Sacra began in the north-eastern corner of the island and has detected: • Structures to the south of the Fossa Traiana between Sant Ippolito and the Capo due Rami • that are possibly associated with the _Statio Marmorum_ • Road alignments • Field boundaries • Possible tomb structures overlooking the Tiber. Excavation: The main _focus_ of excavations remained the large open area at the eastern edge of the Palazzo Imperiale initiated in 2007. The sequence here is now clearer. Considerable emphasis was directed towards the southern side of the channel excavated in 2007. While the expected southern mole has proved elusive, the excavations uncovered the northern face and a range of rooms belonging to the large building delimiting the southern side of the channel: this runs for 250m in an east-west direction, and was c. 80 m wide. This southern wall face embodied a high complex structural sequence running from the 1st through to the later 5th centuries AD. More was learned about the circular building uncovered in 2007. It was in fact ovoid in shape (c. 42m x 35 m) and may have acted as a centre for the registration of incoming cargoes. Emphasis was also directed towards the excavation of the sequence of cisterns at its northern end. It now seems certain that these were built during the Trajanic and Hadrianic periods, undergoing an important series of modifications down into the late antique period, as well as providing evidence for limited occupation during the 11th-13th centuries AD. It is possible that these were the easternmost of a line of cisterns along the northern façade of the Palazzo Imperiale, that were fed by an aqueduct running along the south side of the channel uncovered in 2007, and which may have been used to provide freshwater for ships leaving Portus on their return journeys. Additional fieldwork included a programme of geoarchaeological coring in the excavation area (J-P Goiran, Universite de Lyons), as well as a sub-bottom profile survey of the Trajanic basin in collaboration with the Duca Sforza Cesarini.

The third and final season of excavations at Portus were undertaken for two weeks in the first half of July, and between the 26th August and the 23rd October. This involved the continuation of excavation on the western side of the side at the foot of the late antique walls, within the large building at the southern end of the site (Building 5), on the western side of the cistern block (Buildings 1 and 3) and on the western side of the path in the direction of the main body of the “Palazzo Imperiale.” Overall an area of c. 3000 square metres has now been excavated with an international team of c 50 people, resulting in the identification of seven main periods of structural development along the eastern side of the “Palazzo Imperiale.” One of the highlights of the season included the discovery of the western side of the amphitheatre, which was found to be inscribed within a monumental hemicycle, a seriously of luxuriously appointed rooms lying immediately to the west which formed the eastern edge of a garden complex. The rooms generated large quantities of imported marble, including a very fine head, possibly representing Ulysses, and environmental material. Another was the discovery of the Claudian mole and an associated series of beach deposits. In addition to the excavation, there was an intensive programme of Ground Penetrating Radar survey covering the whole of the “Palazzo Imperiale,” further environmental coring (with the Université de Lyon) and a first programme of helicopter-based infra-red photography of the whole of Portus and the Isola Sacra (with the University of Ghent).

During the 2013 Field School of the Portus Project excavations were undertaken in Building 5 - “navalia”, and Buildings 8 and 3, which are part of the “Imperial Palace”. Building 5 (ca. 240 x 60 m), divided into large aisles 12 or 18 m wide separated into by corridors, was built during Trajan’s reign, when it was probably used for the construction and/or maintenance of trading ships or warships. It underwent substantial transformations at the end of the 2nd century A.D. when many of the aisles were divided, and in the late antique period when it was incorporated within the city walls and lastly occupied by burials. A new trench was opened in the large aisle investigated in 2011-2012, where the 3rd century A.D. make up, on top of which the remains of 4th-5th century A.D. _suspensurae_ were preserved, was removed. This exposed the first 2nd century A.D. levels connected with the subdivision of the aisles and then the Trajanic floor levels constituted by make-ups and sand deposits, with deep abrasions and numerous postholes. The Imperial Palace complex extends over an area of ca. 3 hectares on a peninsula at the centre of the port, situated between Trajan’s hexagonal basin and the earlier basin built at the time of Claudius. From 2012 onwards, excavations have been in progress in the north-eastern corner of the complex (Building 8) and are co-ordinated with restoration work directed by Dr. Angelo Pellegrino (SSBAR). Only the ground floor and a part of the first floor of the complex, which was originally on three levels, are preserved. Built during the reign of Trajan, it was occupied for a long period and underwent many structural alterations. Several first floor rooms have been investigated. Among these a latrine with the remains of the original Trajanic black and white mosaic floor, cut by the insertion of a new marble floor when the latrine bench was renewed and a new door opened in the 4th-5th century A.D. All of the marble was robbed out, presumably at the end of the 5th century A.D. Another room on the first floor, trapezoidal in plan and as yet of unknown function, was enriched with a showy marble floor and wall decoration. The make-ups, using long fragments of African amphorae, are preserved. Part of the portico and the open area of a peristyle were investigated, revealing two brick columns still faced with white stucco and the collapse _in_ _situ_ of a corner pillar, lying on the deposits obliterating the structures. A mosaic floor was partially exposed in the portico, of which the border formed by a polychrome plait was preserved. In the open part of the peristyle the remains of a floor of marble slabs was exposed, in phase with the cistern below it. In a corridor that was originally open, the abandonment layers dating from the 6th century A.D. onwards were investigated. Building 3 is a rectangular structure delimited by walls in _opus_ _testaceum_ and _opus_ _mixtum_, arranged around a central peristyle with vaulted portico and _opus_ _spicatum_ floors. It underwent many transformations and in the 3rd century A.D. housed a glass making workshop. The cleaning of the rubble deposits revealed the presence of numerous interior walls relating to the roofing of the corridors, of great interest for the reconstruction of the building’s walls. In a point where no interior walls were present, a trench was opened in order to investigate the accumulations of material overlying the floor.

The excavations concentrated on Building 5¸ the ‘_navalia_’ and Buildings 3, 6 and 8 that were part of the ‘Imperial Palace’. Building 5 (c. 240 x 60 m) divided into large bays 12 or 18 m wide separated in groups by corridors, was built in Trajan’s reign, when it was probably used for the construction and/or maintenance of commercial vessels or warships. The building underwent substantial alterations at the end of the 2nd century A.D., when many of the bays were subdivided, and in the late antique period when it was incorporated into the city walls and finally occupied by burials. The trench in the large bay opened in 2013 was extended as far as the eastern edge of the structure, where the 3rd century A.D. make-up was removed, on which the remains of the 3rd and 4th-5th century A.D. _suspensurae_ were preserved. This revealed the 2nd century A.D. levels relating to the subdivision of the bays and then the Trajanic floors formed by make-ups and deposits of sand, with deep abrasions and numerous postholes. The Imperial Palace is a complex that covers about 3 hectares on a peninsula at the centre of the port, situated between Trajan’s hexagonal basin and the earlier basin of Claudius. Building 3 is a rectangular structure with _opus_ _testaceum_ and _opus_ _mixtum_ walls, arranged around a central peristyle with vaulted colonnades and an _opus_ _spicatum_ floor with many rebuilds. In the 3rd century A.D., it housed a glass-making workshop. During the 2012 and 2013 campaigns the building’s east side was identified, while the latest campaign concentrated on the southern sector, in particular the layers of fill were excavated down to a strip of _opus_ _spicatum_ floor running east-west between two pillars of the south colonnade. A new trench was opened in building 6 that uncovered the remains of an east-west corridor (paved with a mosaic of large black and white tessera), of a podium and of a make-up. Building 8, built in the Trajanic era, shows a long period of use and evidence of structural alterations. This was a complex of rooms originally arranged on three levels along the north west front of the palace around a central peristyle, on the first floor corresponding to a cistern on the ground floor. During this campaign, excavations were concluded in the first floor corridor (Room 11) and the _sondage_ in the corridor (Room 3). A _sondage_ was also opened along the front facing onto the port of Claudius. In Room 11, the deposits dating to the 6th century A.D. onwards were removed. Below was the robbing level of an _opus_ _sectile_ marble floor of which only a small patch and the imprints in the make-up survived. The area of the door communicating with Room 9 was also excavated, revealing part of a polychrome mosaic already identified in another part of the same room during last season. In the southern part of room 3 on the ground floor, where the complete stratigraphic sequence was preserved, from the demolition collapses to the floor levels, the Trajanic ramp was reached. In the trench along the north front, accumulations of rubble and sand from the demolition of the north facade were investigated and medieval pottery was recovered. Below were sandy and sandy-silt levels containing abundant shells, which abutted the Trajanic wharf.