What you should expect depends on which country you are in

“Expectation is the root of all heartache.” – William Shakespeare


Or is it? Discarding any expectations will make us feel left in dark, which could produce another problem – heartnumb.


Maybe to avoid both, people invented guidelines. For transport modelling, the most extensive and probably also most respected guideline is the UK WebTAG. After spending most of my career (so far) in Australia, please allow me to have a bit bias to Australia and New Zealand guidelines. Lastly the Americans always have a say on pretty much everything.


Interestingly, the requirements by the different guidelines may reflect the personality of the people who created them.


The table below compares the criteria for highway assignment by road standard from four sources: 1 from the UK WebTag, 2 from Australia and New Zealand guideline respectively, and 1 from the US TMIP.

Validation_criteria

The UK WebTAG criteria are the most stringent, so are the British people (well, that may be an out-of-date impression stemming from the description of boarding schools in English novels – or maybe because that children’s memories about schools are not that great for most of time). Anyway, back to the guidelines, it is no doubt that the criteria in terms of the proportion of roads that should fall within a specified range are the most stringent in the WebTAG.


But, let us not to forget that British also have the longest history (just count how many Henry kings they had), and so are they in transport modelling. The most stringent requirements is a direct reflection of the guidance on best modelling practice in the UK for highway models which places a strong reliance on constant and extensive surveys to capture observed travel patterns through roadside interviews and then the use of matrix estimation to further improve the highway matrices.


Therefore, it could be fair to say that the WebTAG criteria are too stringent for more strategic models and particularly those based on the derivation of synthetic demands. And you could say so loudly if you are working on transport models outside the UK.


The Australian and New Zealand criteria are more relaxed – ‘she’ll be right, mate.’ Both of them reflect regional or strategic models and are more relaxed in terms of the proportion of roads that should fall within a specified range, and in the Australian guidance the ranges themselves even also wider – probably because Aussies have wider land, or simply more relaxed. And who can blame them for being relaxed- when you have so many fantastic scenes to enjoy, why make your life miserable by worrying about missing the target by 1%? (well, I will be quick to admit if you point out that is more than 1%.)


The TMIP criteria are probably too relaxed, actually to a degree that sounds a bit careless or arrogant. To be fair, the least stringent guideline is a reflection of the nature of the models in the US with large scale regional models being common, and the TMIP criteria recognises that for large models that there will be more variance in the modelled vs. observed comparisons and as such reasonable criteria are set out. And who knows, maybe the freedom that Americans enjoy is the fountain of their innovations when, ironically, the least is expected by them.


So, what should you expect? It may depend on which country you are in.

Categories: Blog

Leave a Reply