This paper describes a method to explore the nature of the relationships between Portus and other selected ports in the Mediterranean basin, from calculating the sailing time between different points of interest derived from seasonal weather patterns data.

The sailing distance is calculated by tessellating the possible sailing distance that a ship could have moved in a given time unit and summarising the results, the resultant product is an anisotropic pattern of movement.

By altering the parameters of anisotropic pattern analysis, the seasonality of the wind/wave source   the combined wind/wave force (most frequent, maximum, minimum, average ) , efficiency of the sail it is possible to consider different sailing conditions and sailing abilities

By calculating the intersection between the different anisotropic bands and a point of interest an asymmetric directed graph can be generated of the inter port travelling times

By applying graph theory to the derived graph, the following areas can be examined:

• The calculation of the best sailing route between different ports: Generated by considering   the sailing time between N spatial located nodes with the ports forming the terminal and starting points of the route.
• Maximum and minimum travel distances for different classes of goods:  Calculated by only examining trade routes whose sailing time is less than the life expectancy of the considered goods.
• A defence analysis of the trading routes derived from considering which spatial points would be most traversed during normal sailing patterns, and therefore may require defending.
• A hazard analysis of the trading routes derived from examining which spatial points would be best avoided (such as areas  known for storms, locations of wrecks etc)  during normal sailing patterns.

This research forms part of the Roman Port Networks Project and also relates to the AHRC Portus Project, both directed by Simon Keay (http://www.portusproject.org)