Over the last two decades, cities in the Middle East have witnessed significant growth in economic development, resulting in population growth and higher levels of employment which, in turn, lead to increased daily trips on limited transport infrastructure.
To cope with the fast pace of development, the concerned authorities accelerated infrastructure expansion. To this end, they required robust and reliable transport models to assist them with making informed decisions about the provision of appropriate transport infrastructure. The existing transport models were simple and mostly represented only private transport modes. The authorities in the emirates of Dubai, Abu Dhabi, Sharjah, Doha and in Riyadh, KSA, began the development of comprehensive multi-modal transportation models using the latest techniques and software. Since 2004, comprehensive multi-modal transport models have been developed and applied extensively in various transport projects including in Dubai, Abu Dhabi, Jeddah and Doha. A similar model is currently under development in Sharjah.
Typically, input data include socio-economic, origin-destination and modal preference data obtained through household, roadside and stated preference surveys that rely on accurate understanding and responses from interviewees. However, this is difficult to achieve given population diversity in these cities, including significant proportions of construction workers and lower-income employees with little experience, or understanding of, the type of questions being asked. Furthermore, as of now, these cities do not have established postal addressing systems that can be used to accurately identify the origin and/or destination of reported trips. There is also a tendency to under-report key information by some segments of population. For example, high income groups, including locals and expats, tend not to reveal full information on their income or car ownership level.
The reported survey issues contribute to inaccuracies in locating trip origin and/or destination, as well as under-estimating person trip rates and daily trips. Errors in trip distribution estimation can be minimised through using large sample sizes for household and roadside interview surveys, combined with careful calibration and validation. Furthermore, the data collected should be carefully checked to ensure that it is reliable and fit for purpose.
Inevitably, this would lead to high survey fees that may not be cost-effective, nor acceptable to survey commissioners. Clearly, the use of smart data collection techniques (mobile devices, bluetooth, sensors, smartcards, video, etc) could help increase data reliability and sample size and reduce data collection costs.
Typically, demand forecasting models include the traditional four stages of trip generation, trip distribution, modal split and trip assignment. However, details within each stage are structured with a view of representing local conditions and reproducing the travel behaviour of various population groups for different journey purposes. Furthermore, a key feature of the model is the relationship between the base conditions and future projections. Internationally, it is accepted that the validated base model parameters would be applicable to future travel conditions. This assumption is probably reasonable in developed countries where there is little change in land use patterns, population mix and travel conditions between the base and future years. As such, models in developed countries are incremental, pivoting around the base conditions.
However, in developing countries and in particular in the Middle East, the patterns of land use and daily trips are continuously changing and, in most cases, completely different between the base and future years. This is the main reason for existing models in the Middle East being absolute models with no direct relationship between base and future year models, other than model parameters. However, this may cast doubt on the reliability and validity of forecast year projections. Given the extensive efforts in validation of the base year model, it would be useful to develop an approach that establishes a relationship between the base and future year values with a view of increasing the reliability of future year model output. This would, in effect, be combining the incremental and absolute modelling approach.
Model validation is an important aspect of model development process, checking the accuracy of model estimation. Extensive data and rigorous criteria are used for validating different areas of the model. Generally, guidelines and standards from the British guidance (WebTAG or DMRB) are used for validation of these demand forecasting models. However, these rigorous validation criteria would be applicable and achievable in stable conditions, such as those in developed countries, and more difficult to achieve under continuously changing conditions such as those in the Middle East. Furthermore, it is not clear that, even if model validation based on rigorous criteria could be achieved, how valuable this would be as regards the reliability and accuracy of model forecasts in an absolute model structure where there is little relationship between base and future year models. Clearly, international guidelines and criteria are important for checking the validity of model estimation, but consideration should be given to local conditions and, where possible, the criteria should be revised to reflect local conditions.
Typically, demand forecasting models are strategic in nature and applied to forecast changes in travel pattern between travel modes, along key transport corridors and across the study area, under a given scenario. The validation of these models is also at strategic level and model output at local area level is therefore not necessarily accurate. However, the developed strategic models in the Middle East are being used for local area studies including traffic impact studies. Also, these models are occasionally being applied for assessing traffic management schemes that should be done using an appropriate operational modelling tool. The impacts of a transport scheme, whether strategic or local, could accurately be estimated using an appropriate modelling tool. There is therefore a need to develop supporting city-wide operational modelling tools to complement the existing strategic models. This points to a need to develop an understanding of the strengths and limitations of available data and different modelling tools.
Generally, demand forecasting models are updated every ten years and usually coincide with a general census. However in the Middle East ,where there is a rapid and continuous change in land use development and trip pattern, there may be a need to update the model more regularly, possibly every five years. Ideally, every five years, the model should be re-based and calibrated and validated to reflect the changes in demand and supply. As this would require extensive and expensive data collection programmes, there is an urgent need to develop an approach for continuous and targetted data collection programmes using smart technology.
Reza Mohammadi, Director, Reza Mohammadi Consultancy (RMC)
TransportXtra is part of Landor LINKS
© 2019 TransportXtra | Landor LINKS Ltd | All Rights Reserved
Subscriptions, Magazines & Online Access Enquires
[Frequently Asked Questions]
Email: firstname.lastname@example.org | Tel: +44 (0) 20 7091 7959
Shop & Accounts Enquires
Email: email@example.com | Tel: +44 (0) 20 7091 7855
Advertising Sales & Recruitment Enquires
Email: firstname.lastname@example.org | Tel: +44 (0) 20 7091 7861
Events & Conference Enquires
Email: email@example.com | Tel: +44 (0) 20 7091 7865
Press Releases & Editorial Enquires
Email: firstname.lastname@example.org | Tel: +44 (0) 20 7091 7875