Modelling local rail demand in South Wales


Blainey, S. (2009) Modelling local rail demand in South Wales. In, 41st Annual Conference of the Universities' Transport Study Group, London, United Kingdom, Jan 2009.

Download

Full text not available from this repository.

Description/Abstract

The majority of the recent work which has been undertaken on modelling the absolute demand for rail travel has focused on inter-urban trips, meaning that local and suburban travel has been somewhat neglected.

This paper describes the development of flow-level absolute demand models based on LENNON ticket sales data for 85 local rail stations in South Wales. A range of log-linear regression models were calibrated on a dataset of 2,440 flows, incorporating a variety of station-specific and flow-specific independent variables. Subsets of this dataset were used to test models which incorporated measures of intermodal competition and the presence of intervening opportunities. GIS were used extensively in data processing and integration, and allowed flexible station catchment definition methods to be tested, including flow-specific catchments where population units were allocated to stations by minimising total travel time to the destination in question. To allow these methods to be validated a survey to collect information on actual station catchments was carried out on the Rhymney line and the results compared to the theoretical catchments.

The flow level model predictions did not correspond with either observed or forecast station trip totals and in an attempt to give consistency between predictions, methods were developed to constrain the number of trips predicted for each flow based on the total trips observed or predicted from origin stations. Constrained flow level predictions were first obtained by using observed trip end totals to scale the results from unconstrained models. This was not entirely successful, so probabilistic trip distribution models were also calibrated using both linear and nonlinear regression. These gave better results and also explicitly incorporated the effects of intervening opportunities in the model form.

Item Type: Conference or Workshop Item (Paper)
Additional Information: 09C Rail (I)
Related URLs:
Subjects: T Technology
T Technology > TA Engineering (General). Civil engineering (General)
Divisions: University Structure - Pre August 2011 > School of Civil Engineering and the Environment
ePrint ID: 73877
Date Deposited: 16 Mar 2010
Last Modified: 27 Mar 2014 18:52
URI: http://eprints.soton.ac.uk/id/eprint/73877

Actions (login required)

View Item View Item