摘要：AbstractThe more intensive use of containers for channelling the flows of international trade has led to the ports become nodes of the main supply chains. As a result, there is a strong competition among port alternatives to attract routes to their facilities in order to gain the greater share of maritime traffic. The need to identify the variables responsible for the success of a port over its competitors has led to the birth of a line of research to develop a tool for port planning. This tool, called GOAPORT, is being developed in the framework of discrete choices modes which allows us to identify these variables and forecasting market shares in different situations from current one.To calibrate it we need a data with the maximum possible detail. But surveys are very expensive and public data available is sparse and too aggregated. The problem is compounded when we try to research the Spanish port system. Due to confidentiality issues we are not allowed to know directly which ports use the Spanish companies to export and import their goods abroad.This paper focus on the development of a theoretical model which work a disaggregated level but whose data are partly aggregated making impossible to know the alternative chosen at a disaggregated level. This model, let us use the Spanish public data at the highest level of detail possible. Thus it, it can explain and predict the port choice by knowing characteristics of the company, the place where they are located, the land which they trade, trade goods, and of course, port characteristics.Furthermore, the methodology set up could be reuse and applied it in other fields as long as they have the same common structure of public data.
