White Rose University Consortium logo
University of Leeds logo University of Sheffield logo York University logo

Greater Manchester PTE New Railway Station Demand Prediction Model

Preston, J.M and Aldridge, D.M. (1991) Greater Manchester PTE New Railway Station Demand Prediction Model. Working Paper. Institute of Transport Studies, University of Leeds , Leeds, UK.

Full text available as:
[img]
Preview
Text
ITS236_WP326_uploadable.pdf
Available under License : See the attached licence file.

Download (732Kb)
Published Version: http://www.its.leeds.ac.uk/

Abstract

This paper reports on work that was initiated in February 1989 to develop a simple model that can accurately predict the usage of new stations in the Greater Manchester area. The starting point for this study was the work carried out by Moss in 1988 in which an attempt was made to develop a new station demand model based on patronage data for 9 of the 10 new stations that had been opened in Greater Manchester since 1984. In his research the main explanatory variable was the population within 1,000m and the sub-divisions 0-300m, 300-600m, 600-800m and 800-1,000m were also examined. Arbitrary dummy variables were studied to assess the effect of service frequency, car ownership, alternative routes and park and ride. The main finding was that population, on its own, did not appear to be an adequate explanatory variable. For example, Mills Hill was found to attract 7.9 times as many passengers per 1,000 households as Hag Fold. In other words, there was very large variation in the trip rates at these nine new stations, with a mean of 36.2 and a standard deviation of 22.4 daily trips per thousand households.

An alternative approach is that developed in West Yorkshire based on multiple regression techniques (Preston, 1987). This model predicts the number of rail trips between two stations as a function of: (i) the population within 800 metres of the origin station, (ii) the proportion of that population in social classes I and II, (iii) the population between 800 metres and 2 kilometres of the origin station, (iv) the number of jobs within 800 metres of the destination station, (v) the generalised cost of rail, (vi) the generalised cost of competing modes (bus and car).

This model is a form of direct demand model in that it forecasts the number of trips (T) between origin i and destination j by mode k (ie. Tijk). In this paper, we shall develop a simpler version of this model which will simply predict the number of trips from origin i by mode k (ie. Tik). We shall call this a trip end model.

The West Yorkshire model (called the Aggregate Simultaneous Model - ASM) was calibrated for 39 existing stations based on patronage data collected in the early 1980s. In this work, we shall attempt to calibrate a similar model for 36 existing stations in Greater Manchester, based on patronage data collected in 1987/8. These stations are listed in Appendix 1. Of these stations 16 are on what we have termed the Oldham Loop, 9 are on the Bury Line, 9 are on the Altrincham Line and the remaining 2 are on the Buxton Line. It was felt that this sample was reasonably representative of Greater Manchester new stations although well used commuter stations may be unavoidably over represented due to the availability of patronage data which dictated at data set. In developing a new station model for Greater Manchester, we have borne in mind comments made about Moss's earlier work by Greater Manchester PTE, Greater Manchester Transportation Unit, Manchester City Council and British Rail, Provincial (Midland) and in particular the consideration of existing trip patterns, especially to central Manchester. Whilst we encountered difficulties in obtaining relevant data, we were able to explicitly incorporate a number of other points. Our work will be based on multiple regression and will be able to examine the effect of distance, frequency and car ownership directly. Use will be made of 0-800m and 800m - 2km populations (rather than households within 1km), adjusted to take into account overlapping catchment areas. Figure 1 shows the zoning scheme used. Having dealt with the background to this study, the rest of the report will be as follows: - in section 2 we outline the data sought and made available for our study; - in section 3 we describe the process of calibrating a simple trip end model; - in section 4 our simple trip end model is developed further and in greater depth with the aim of maximising the goodness of fit; - in section 5 we develop a more generalised framework; - in section 6 we calibrate a simple trip end model for walk access patrons; -in section 7 we comment on the criteria of model choice, discuss statistical problems related to cross-sectional data and summarise our findings;

Item Type: Monograph (Working Paper)
Copyright, Publisher and Additional Information: Copyright of the Institute of Transport Studies, University Of Leeds
Academic Units: The University of Leeds > Faculty of Environment (Leeds) > Institute for Transport Studies (Leeds)
Depositing User: Adrian May
Date Deposited: 10 May 2007
Last Modified: 08 Feb 2013 17:04
Published Version: http://www.its.leeds.ac.uk/
Status: Published
Publisher: Institute of Transport Studies, University of Leeds
Identification Number: Working Paper 326
URI: http://eprints.whiterose.ac.uk/id/eprint/2234

Actions (login required)

View Item View Item