< Previous | Contents | Next >

Guide to Transport Assessment in Scotland

APPENDIX B : DATA AVAILABILITY

1. This appendix reviews Scottish and general UK data that is available for those wishing to conduct a Transport Assessment. The scale of analysis required for a Transport Assessment will vary widely between proposed developments.

Journey Time and Distance Data

2. Journey times and distances will be used to calculate isochrones and to determine catchment areas. A large number of sources are available, ranging from simple paper based maps to sophisticated electronic transport models.

3. A distinction between data sources can be made on the basis of how journey time information is supplied. Some data sources/software packages supply information on an interrogative basis, i.e. for specific pairs of origins and destinations at a time. For instance, a rail journey planner can be used to provide journey times between two stations on the rail network. Other data sources/ systems will be able to provide journey times for a range of origin-destination (O-D) pairs at a time. Transport models for instance, where available, can be used to supply journey time matrices for car or public transport modes for any combination of O-D’s in a study area.

Modal Split

Research on Measures to Influence Modal Split

4. One area of concern is the paucity of research that can predict the effects of measures designed to alter modal split. Many research projects have been undertaken to measure the effect on modal split of measures varying from new light rail networks to traffic calming, and persuasion campaigns. Research findings often contradict one another, and the caution on the part of researchers can cast doubt that many measures would have any significant measurable effects.

5. The following comments should be made:

• Most exercises try to isolate the effects of one policy in a 'laboratory situation'. Such an approach does not generally produce results because of background influences, and the length of time needed between studies.
• Few studies measure a package of policy measures, which tends to be the situation where policy makers are trying to influence behaviour. And since one package is very unlikely to be the same as another package they are equally difficult to compare.

6. Recent publications, such as that by Rye and McGuigan (2000) "Green Commuter Plans - Do they work" are starting to provide a framework for understanding what measures may work in combination with others, and the effectiveness that can be expected. This understanding will grow rapidly as more Travel Plans are instigated and as monitoring grows.

7. It must be recognised that many measures designed to encourage access by means other than the car will only have a minor effect when introduced on their own. As part of a package of measures, however, their impact can be significant. Research to investigate the combined impact of a range of measures is developing and in the future will be of use in preparing Transport Assessments.

Trip Databases

8. Trip generation databases such as GENERATE, TRAVL and TRICS are an important source of information on car travel but their information on non-car travel is limited. Since the databases do not yet contain much information on non-car transport impacts or on developments with operational Travel Plans, they should be employed with caution in Transport Assessments — possibly as a starting point for the base level of travel and modal share for example. In general they have been created to support TIAs, with an overall focus on car travel, rather than other modes, and so do not provide all the data required for a Transport Assessment. In addition, their use may tend to replicate past travel patterns, thus encouraging developers to provide for more car travel than is necessary, rather than promoting sustainable modes as current government policy advocates.

9. Nevertheless, for the present, these databases remain necessary tools, since they provide more information on behaviour than other surveys. Furthermore they are now being changed to allow inclusion of data for modes other than the car.

Trip Generation and Trip Rates

10. TRICS (Trip Rate Information and Computing System) is a large database containing traffic count data for a number of days at a large number of sites of different categories of land-use. In addition the database also contains information on the size of the development (retail floor area, office space, number of residential units, etc.). The database contains data for a large number of sites throughout the country and is updated regularly. However most data are from the South East of England.

11. TRICS allows the user to calculate trip rates from individual or a group of selected development sites, which can be selected by the user imposing a range of criteria, such as Gross Floor Area, Retail Floor Area, number of employees etc. The database is generally constructed from vehicle counts only. Most records contain no information on trip lengths, car occupancy levels, origin-destination information, or trips by public transport, on foot or by bicycle. However, a number of pilot surveys have taken place and a few sites with multi-modal data have been incorporated.

12. Two technical terms must be understood:

• "Trip generation" refers to trips associated with a land use, such as trips to a hospital.
• "Trip rates" refer to trips per time period, and can be personal trip rates (as in trips per person per day) or trips associated with a land use per day.

13. The selection of a trip rate is important to estimating the travel generation of a site. The following factors need to be taken into account in selecting a rate:

• the size of the walking catchment
• the propensity to cycle;
• the type and quality of public transport provision;
• the level of on-site parking restraint;
• whether there are parking controls in the vicinity of the scheme;
• the potential for linked trip opportunities; and
• for retail sites, competition within the locality.

14. In the past TIAs have compared the proposed development with similar ones, and have been careful not to underestimate the vehicle trip generation. Typically the "85th percentile" trip rate has been chosen. This means that "of 100 similar developments, the proposed development will have trip generation characteristics corresponding to the 85th of these when their trip generation characteristics are placed in order". This approach has often resulted in TIAs assuming higher than average traffic levels. This discourages other modes of transport, given that if significant parking is provided it will reduce the likelihood of people using non-car modes.

15. In preparing Transport Assessments it is important to consider a wide range of trip generation rates such as 15th, 50th and 85th percentile trip rates. Depending on the nature of the development it is likely that a pair of lower (15th and 50th) or higher (50th and 85th) rates will be appropriate for most developments. If satisfactory measures designed to reduce car use at proposed developments are included, e.g. maximum parking standards, it is hoped that the higher rates will seldom be needed. The difference between 15th and 85th percentile rates can be large, involving factors of between 2 and 4. Care must be taken when using the lower rates, particularly with maximum parking standards, that lower car use can actually be achieved.

Person Trip Travel Surveys

16. It is important to undertake a person-trip based assessment for all developments with significant transport implications. The main sources of person trip data in Scotland are currently: the National Travel Survey; the Census; and local movement surveys (if conducted by local authorities). However, these surveys are not primarily designed for use in Transport Assessments and should be used with care:

• The National Travel Survey contains detailed information and can be useful for estimating trip generation from housing developments. However, it records journey purposes (such as "shopping") rather than trips to development types (e.g. "corner shop", "supermarket", or "town centre"), and it cannot be used for detailed local analysis.
• The National Census deals with home-based work journeys only, and records only the mode of travel for the main part of the journey. This can obscure information which is of use in a Transport Assessment such as the mode used to leave or arrive at a development.
• Local Movement Surveys can vary in their value depending on the purposes for which they have been carried out, and the detail included.

17. A major benefit of the National Travel Survey and the Census is that there is information on all modes, in contrast with the focus on car use that is common in vehicle databases. However, the NTS and Census cover trips from housing, so a likely difficulty will lie in estimating trips to and from other land uses although this can be overcome. Developers are recommended to inquire as to the availability of area data from the local authority that may assist in the preparation of a Transport Assessment.

Estimating Modal Split

18. Modal split estimates can draw on catchment area analysis, comparing travel behaviour with activity elsewhere and supported by reasoned argument about likely travel behaviour. However, it may be possible to make good estimates with less emphasis on these techniques by comparisons with similar existing developments, and including the effect of proposed measures to influence travel in the proposal. In seeking suitable sites, one should look for comparable locations, scale, public transport provision and non-car accessibility.

19. Using these techniques should provide estimates of the numbers travelling to the site, their choice of mode, and the overall modal split.

20. In general:

• Data for analysing populations within catchments according to travel times are available for certain areas to a high degree of accuracy using GIS sources linked to population data. But outside major urban areas the information, especially with regard to public transport, may be more difficult to come by.
• For rural areas access by non-car modes may be simple to calculate without data sources other than maps, and bus timetables. In between the larger urban areas and rural areas, and for smaller developments within larger urban areas, data for complex analyses will be difficult or impossible to obtain, or may be too costly for smaller developers to gain access to.
• Data on travel behaviour from which modal split predictions might be made is good at the national level, but only at the level of trip making for different journey purposes. For journeys by non-car modes to specific developments (especially 'new types' of development such as multiplex cinemas) data is not good.

Road networks

Ordnance Survey OSCAR Data

21. For the purpose of undertaking Transport Assessments four road network data sets based on OSCAR (Ordnance Survey Centre Alignment Of Roads), are available, with a varying degree of detail. Roads and the information associated with them are modelled in to a vector network of links and nodes (Source: http://www.ons.gov.uk) . Similar networks from other suppliers are also available.

• OSCAR Asset-manager is the most detailed of these and provides a comprehensive representation of all public and selected private roads. The whole country is covered and available at levels of detail relating to 1:1250, 1:2500, and 1:10000 scales.
• OSCAR Route-Manager has been designed for regional routing and networking requirements. The entire public road network is included, with the exception of short cul-de-sacs. Complex road junctions are shown as single nodes and multi-carriageways as single links.
• OSCAR Traffic-Manager includes all public and selected private roads and includes Drive Restriction Information (DRI). Drive Restriction Information shows the restrictions, which affect the route of a vehicle, such as one-way streets or banned turns. DRI is available for all classified roads, and within the urban envelope of all Metropolitan areas and selected District Council areas. In addition, the centres of 300 towns are also available.
• OSCAR Network-Manager has more simplified road geometry than the other road products, containing only Motorways, A and B roads.

Deriving Walk and Cycle Times

22. Road networks, like OSCAR can be used to derive walk and cycle access times. Distances between origin and destination pairs via the network are calculated and converted into time using an assumed average travel speed. It will generally be necessary to use the most detailed networks available. But the following should be remembered:

• All roads in the network (including motorways and inappropriate roads) will be included in the calculation, but paths and other shortcuts will not.
• Euclidean (straight line) distances may be calculated with assumed average speeds, but that would not take into account the characteristics of the actual walking network, such as hills or pedestrian crossings.

23. Such data has to be viewed as inadequate for realistic walking and cycling catchment area calculation, unless specially adapted.

24. A three-year Scottish Cycling Development project was carried out in Scotland in 1997. One of the outputs is a Scottish Cycling database on the Internet, which contains information on cycle networks which can be used to derive cycle access times. It can be found at http://www.scottishcycling.co.uk.

Deriving Car Drive Times

25. A number of software packages are available (within GIS systems) which perform shortest path calculations through road networks and convert distances into times. In most cases the road network is divided into different road types and a database containing typical speeds by road type is used to derive link times. Routing algorithms are used to calculate a shortest path. Off-peak travel times and free flow traffic conditions are usually used in calculations. For the purposes of a Transport Assessment it might be necessary to represent a network for the AM or PM peak period. Although road speeds can be customised to reflect local conditions, the process to validate journey times in a network might be time-consuming.

26. Examples of software packages to calculate road journey times include Drivetime, RouteView, Geoconcept, and Autoroute. Drivetime contains 38 classes of roads with associated speeds. Autoroute is most suitable to derive journey times at a national level. It includes major roads only and uses 10 categories of roads in its journey time calculation. Autoroute is used on an interrogative basis and only allows the user to look at one origin-destination pair at a time.

Public Transport Networks

27. There are various paper-based data sources available providing information on public transport networks. In Scotland, the Traffic Commissioners and the relevant local authority, or Strathclyde Passenger Transport (SPT) in the areas covered by SPT, should have information on routes operating, and may have mechanisms in place for obtaining information from public transport operators. Often printed timetables are provided. Details on the rail network can be found in the Great Britain Passenger Timetable.

28. Information on the rail network, including station locations is available in digital format from a number of suppliers. Information on bus networks is not widely available in digital format, although it would be possible to represent a bus network in GIS by selecting roads served by buses as a subset of one of the road networks.

29. Calculating journey times through a public transport network is more complicated. It is not just a matter of selecting routes served by buses and using average bus speeds as various other elements need to be taken into account, including:

• The time to walk to a bus stop/ station;
• Wait time at the bus stop;
• The travel time spent on the bus;
• Interchange time (if relevant); and
• The time to walk from the bus stop to reach a destination.

Deriving Public Transport Journey Times

30. Currently, data and software to provide door-to-door public transport journey times is not readily available without a large degree of effort, unless special local exercises have been undertaken. The data generally exists to allow such exercises, but the effort involved would be beyond what would be considered appropriate for a Transport Assessment.

31. Information about railway station locations and railway lines are widely available in digital format for use with GIS. Electronic data on bus stop locations is less widely available, although databases tend to exist in the larger Metropolitan areas. Alternatively, bus operators might be able to supply information on bus stop locations. Data on service patterns and frequencies will generally be paper based and locally available, though increasingly they are available on the internet. Walking times to and from rail stations and bus stops can be calculated in a similar way as described under the heading ‘Deriving walk and cycle times’.

Bus and Coach Statistics

32. These statistics provide information about the trends in bus and coach services in Scotland, and include data on distances travelled by vehicles, numbers of bus passenger journeys, fare indices, passenger receipts, public transport support, operating costs, vehicle stock and staffing. The most recent Statistical Bulletin Trn/2001/2 Transport series Bus and Coach Statistics: 1999-2000, was published in March 2001. Bus and Coach Statistics are produced annually. The data may contribute to a Transport Assessment by providing background information.

Railplanner / Railtrack Journey Planner

33. Railplanner and the Railtrack journey planner can be used to provide journey times between any two stations served by the National Rail network. Service patterns and frequencies as well as interchanges can also be derived.

Bus Journey Planners

34. A number of bus journey planners are available through the Internet allowing the user to identify bus journey time information for a number of networks around the country (free of charge), although there appears to be much variation in the quality of information supplied. Examples include express services in Scotland, and the Scottish Citylink website at http://www.citylink.co.uk/

Telephone Enquiry Lines and "PTI" Data

35. The Transport White Papers called for a Public Transport Information (PTI) system to be systematically extended across the UK. Work is well advanced and is being led in Scotland by a consortium of transport operators and local authorities and the Scottish Executive. "Traveline" call centres have been set up in Scotland and in regions throughout the UK which can be accessed by a single telephone number. Callers are automatically connected to the nearest call centre. There is an easy transfer of calls between call centres to access information not covered by the centre a caller is connected to. The information available includes that currently provided by the National Rail Enquiry Service (NRES).

36. As part of this public transport information system, data on all public transport services in the UK are being coded for computer interrogation. PTI data on public transport is not yet made available in data form, but when it is, it should be of great value for public transport system coding. It should be possible to link this data to a GIS to use in accessibility analysis.

37. PTI systems should comprise a database of records disaggregated into bus stop locations, service network, service details, and route details, which can be interrogated in order to answer individual enquiries from passengers. GIS can be used to display bus stop locations and route networks, and accessibility analysis can be carried out on the underlying public transport service level data. The process of reformatting data can be simplified by developing an interface between the two systems.

38. The UK Public Transport Information (UK PTI) website contains links to a wide range of local and national public transport operators, with information on routes, timetables and booking. It can be found at http://www.pti.org.uk/.

Transport Modelling Software

39. Transport modelling packages may be used for performing routing calculations through public transport networks. For each O-D combination journey time, wait time, interchange time and walk time will be reported and can be included in journey time matrices. However setting up a transport model might be a complex and laborious task and, depending on the size of the study area, could require a significant resource commitment.

40. A number of software packages are available to provide information about public transport journey times. Most of these will be able to provide information on an interrogative basis providing journey details for one O-D combinations at a time. These might be helpful in the case of smaller developments. Some examples are listed below.

Land-use data

41. Information on locations of population, and facilities such as schools, hospital, etc. will be needed to calculate catchment areas. Information on land-use and facilities will also be useful when estimating trips that will be diverted from existing facilities to new developments.

National Land Information Service

42. The National Land Information Service (NLIS) is an initiative to provide national information on land and property. NLIS is the generic term for a set of services allowing the access via computer to geographically related information from a large number of sources (both public and private). The system will be based around a gazetteer, known as the National Land and Property Gazetteer (NLPG). The NLPG will be a national index of all land and property units for Great Britain.

The National Census

43. The census provides information on the number and characteristics of people and households for the whole of the country. Between censuses population estimates are updated using data from registrations of births and deaths and estimates of migration based on information on the number of people changing their general practitioner, changes in the electorate and other material from sample surveys.

44. Data is available at ‘enumeration district’ (ED) level for the 1991 census and will be available at postcode level for the 2001 census. This will enable easier linkage with other data.

Census Special Workplace Statistics (CSWS)

45. The 1991 and 2001 Censuses asked about people’s place and address of work, and mode used to travel to work. Means of Travel to Work shows the mode of transport normally used for the longest part, by distance, of the journey to work. For the 2001 Census for Scotland place of study for students and schoolchildren is also included. Data for a random stratified 10% sample of the 1991 Census was coded and this forms the CSWS. This data provides a useful source for analysing modal split according to distance, type of area and other characteristics, but would require further in-depth analysis by experts to provide useful information for developers carrying out a transport assessment.

46. CSWS provides information about persons in employment by their area of residence. The ‘resident population’ base is restricted to those who are employees or who are self-employed. It covers the full address of the workplace and how the longest part of the journey by distance of the person’s trip to work is made. Workplace address postcodes are used for the location. The distance to work has been calculated as a crow fly-related distance between the co-ordinates of the home and workplace. This data can be useful for assessing the realism of predictions made about mode split and catchments.

Information Held by Various Public Bodies

47. The table below provides an overview of various types of information with regards to land-use and facilities, the organisation holding the data and which areas are covered.

Table B1: Information held by public bodies

Local Facility Databases

48. Data on service centres and facilities (community, health, education, shopping etc) can be collected at various levels. Even the most basic of data can be useful for accessibility purposes; it can be added to as needed. Useful sources include local knowledge, local/ unitary development plans, publicly available directories such as Yellow Pages, and commercial databases of facilities.

49. Many local/unitary development plans already include maps/ lists of local facilities such as service centres. Such lists can be added to by using available data collected locally, or by use of directories such as Yellow Pages (which is available on the internet).

Commercially Held Geo-coded Data on Facilities

50. Geo-coded retail and service datasets are available from a variety of companies. They may use existing databases of such information and ‘add value’ to them by undertaking further research, cleaning and checking data, classifying and summarising records, and geo-coding. MapInfo sells databases for various land uses including post offices, banking and finance, hotels etc, public houses, petrol stations, food retailers, clothing and footwear, newsagents, household goods stores, services and estate agents. Each database costs around £1,000.

Trip Rates and Travel Behaviour

51. One of the methods to estimate the likely amount of traffic a new development might generate is to compare a development with similar existing developments and assuming that trip generation will be similar. For this purpose a number of trip databases has been compiled. It is envisaged that these databases will continue to have a similar role in the Transport Assessment process, though they will have to adapt to provide the information that will be useful for travel by all modes, and to provide stronger links between catchment areas and modal split.

52. The National Travel Survey (NTS) is also of use, as well as other sources of data on travel behaviour.

53. The Scottish Household survey is an ongoing survey funded by the Scottish Executive, and collects a wide range of data on household characteristics and behaviour, including transport. It produces regular bulletins, the most recent of which, No.6, includes information on numbers of vehicles households have access to, using an urban/rural classification and by Local Authority grouping. Statistical Bulletin TRN/2001/1 on Household Transport contained information from the Scottish Household survey on topics including the accessibility and frequency of bus services, travel to work/ school/ study and many others. It is available on the Internet at http://www.scotland.gov.uk/stats/bulletins/00062-00.asp

54. A recently initiated three year project, Scottish Neighbourhood Statistics will be creating a database of local statistics, aimed at informing community planning and regeneration processes. Amongst the areas it will cover will be access to services, within which it hopes to reference the location of key services such as doctor’s surgeries and Post Offices, and model travel times and accessibility. More details can be found at http://www.scotland.gov.uk/stats/neighbours/neighbour.asp.

Visits to Tourist Attractions

55. The database of Visits to Tourist Attractions contains information on number of visitors to United Kingdom tourist attractions with more than 10,000 visitors in the year. It includes a brief summary text, followed by many tables, showing the more successful sites by type of attraction, and lists of sites for Scotland, England, Wales and Northern Ireland. More detailed presentation of the survey information is contained in Sightseeing in the United Kingdom. (Source: Office for National Statistics — http://www.ons.org.uk/)

The National Travel Survey

56. The National Travel Survey is a survey of household and individual travel behaviour and characteristics carried out in Scotland, England and Wales. Since 1989 it has been conducted as a continuous survey with a sample of about 3,500 households per year, but is due to double in scale from 2003. The data covers:

• Variables for data pertaining to the household (such as location, vehicle availability, head of household characteristics, family structure etc);
• Individual variables (such as age, gender, driving licence);
• Variables pertaining to vehicles (such as engine size, annual mileage); and
• Variables pertaining to 'journeys' and 'stages' of journeys (for a period of one week), where a journey is defined by a trip with a 'purpose', and a stage is the use of a mode of transport or vehicle.

57. In addition, journeys over 25 miles are recorded in a three-week diary to increase information on longer journeys.

58. While it will be of great use in providing a general indication of the kinds of travel behaviour that might be expected, it is not accurate enough to provide:

• Data on travel to specialised developments, such as call-centres, since destinations of journeys are only defined in terms of their 'journey purpose'.
• Data pertaining to local areas. The data can be analysed with confidence to a regional level, but the sampling structure does not allow for disaggregation at any lower level (even as County level).

TEMPRO

59. The TEMPRO database might be useful in the Transport Assessment process to determine future year traffic forecasts and to get an understanding of traffic growth. The TEMPRO software provides access to the NTEM (National Trip-end Model) trip-end forecast database as well as the underlying car ownership and planning data projections. Data for any year from 1991 through to 2031 can be retrieved and the growth rate between a selected base and future year provided. TEMPRO calculates values by interpolation and displays the results in terms of the growth between the two selected years at Great Britain, regional, county and local authority levels.

Other Transport Models

60. In Central Scotland, the Central Scotland Transport Model (CSTM3) maintained on behalf of the Scottish Executive, provides an additional and powerful tool for forecasting traffic at the strategic level. This is soon to be extended to cover the North East. Other models held by authorities and the SITM model, maintained by Strathclyde Passenger Transport (SPT), complement this. Access to such models is not however easy and could have significant costs attached.

Scottish Transport Statistics

61. This compendium of statistics, which covers road transport vehicles, bus and coach travel, road freight, toll bridges, the road network, road traffic, injury road accidents, rail services, air transport, water transport, finance and personal and cross-modal travel for Scotland contains descriptive text, tables and charts. It is published annually, the most recent being No.20, the 2001 edition and the twentieth in the series. The Statistics also cover trends over a 10-year period in the areas, as well as some longer-term trends and compares with some key statistics for Great Britain.

Monitoring Studies of Travel Plans

62. Travel Plans and other measures are being increasingly monitored but the level of understanding about which measures have significant impacts on travel behaviour and modal choice for different developments is still at an early stage of development. Reports such as that by Rye and McGuigan "Green Commuter Plans - Do they work?" (2000) shows that we are now at the stage where informed statements can be made, but there may be a danger in transferring such research findings to proposed developments where the details of users, and measures to be utilised, are not fully known.

Conclusions

63. The data available is suitable for many Transport Assessments in Scotland. However it can be of variable quality, with much depending on the type of area in which a development is proposed, and the resources available to the developer undertaking the assessment. Data exists to analyse and predict modal split, but it may require significant resources and expertise to analyse, and the accuracy of forecasts may be unreliable.

64. If Transport Assessments are to be practicable they must be straightforward to carry out and be understandable by many people (including the developers, planning inspectors, and others). Existing data sources will adapt over time to include additional data, and new data sources may be created so that in future developers will be able to draw on a wider range of data than currently exists.

65. In the meantime, whatever the form of prediction suggested, there will be approximation, and room for possible debate between proponents and opponents of developments. Those who have access to expert opinion will be in a stronger position unless effort is made to set up shared databases that all can access freely.

66. The main conclusions of this review are that:

• Data for analysing populations within catchments according to travel times are available for certain areas to a high degree of accuracy using GIS sources linked to population data. Outside major urban areas the information, especially with regard to public transport, may be more difficult to come by.
• For rural areas access by non-car modes may be simple to calculate without data sources other than maps, and bus timetables. In between the larger urban areas and rural areas, and for smaller developments in the larger urban areas, data for complex analyses may be difficult to obtain, or may be too costly for smaller developers to gain access to.
• Data on travel behaviour from which modal split predictions might be made is good at the national level, but only at the level of trip making for different journey purposes. For journeys by non-car modes to specific developments (especially 'new types' of development such as multiplex cinemas) data is poor.

< Previous | Contents | Next >