National transport model of Croatia (NTMC) consists of network, demand and assignment submodels. Whilst network model (nodes, links, zones) is common, approaches for demand and assignment differ between passenger and freight transport. Each type of transport requires specific methods. This paper describes development of the freight demand model, namely its input data, modelling methods, calibration and validation. The main inputs for modelling of freight transport are socioeconomic data and national statistics on production and trade flows (import / export). Socioeconomic data was collected on the zoning level. For production and import / export, the statistical data was available at national level and was broken down to zoning level according to the socioeconomic data of each zone. Main source for socioeconomic data (population, employment, working places by economic sector) were national statistics (Croatian Bureau of Statistics). Sources for production data depend of the type of commodity, while import / export data was derived from the national trade statistics. The freight demand model considers all transport modes relevant from a national perspective, i.e. road (HGV and LGV), rail and vessel. The modal freight trip matrices are calculated with a commodity based multi-modal model using an enhanced 4-step approach. For the adequate consideration of commodity-specific affinities, the freight traffic is segmented into 50 different commodities; each of them having specific generation and attraction rates as well as specific modal transport cost rates. The calculation steps of the enhanced 4-step approach are conducted separately for each commodity. The multi-modal freight model, which mainly generates the long-distance trips, is supplemented by a submodel for generating the freight trips from / to the airports and by a sub-model for the local (short-distance) distribution trips by HGV and LGV.

transport model; freight demand model; calibration; validation