Newsletter #15 - 29 December 2016
We hope that you all had an enjoyable and memorable Christmas. Our teams had the initiative to be out on Boxing Day collecting signatures for our petition at the Hill Street intersection. Thank you to those who signed it.
Over the past and coming week, the NZTA control centre are watching the traffic at the Hill Street intersection and configuring their traffic light sequences to minimise queues. That’s no holiday.
OUTLOOK FOR THURSDAY...
Your guess is as good as mine… and the NZTA’s.
We’ve been comparing the NZTA’s forecasting with real-time data. There is a real disparity between previous year’s peak flows and this year’s. How can this be?
Take for example Tuesday, which NZTA forecast to have the longest peak period of 8:30am-6:30pm. Consensus is that congestion on Wednesday and Thursday has been worse.
There are three things to take into account:
1. The circulation of forecasted traffic information;
2. The reaction to that information; and
3. The desired response to the information.
The NZTA warned motorists to expect delays and heavy traffic volumes on their website, Facebook page, and Twitter feeds. They also alerted the mainstream media. At motorway electronic noticeboards as south as Greenlane, they recommended that northbound traffic use State Highway 16 to Wellsford and avoid State Highway 1 through Warkworth.
It was the NZTA's largest congestion warning ever. But who listened?
Take for example the period before Christmas. Warned about congestion, many left early – often finding that others had the same idea (http://bit.ly/2hqmwyM).
Central to congestion problems is a lot of people having the same idea of jumping into their cars and going to the same place. When warned by a forecast that the congestion or weather will be bad, many might change their schedule to another time – and many others might have the same idea. If the alternative journey window narrows, congestion could be worse than what motorists were trying to avoid.
Holiday traffic patterns are predictable. Christmas shopping, Christmas gatherings, and New Year’s gatherings are different in subtle ways. Many celebrate Christmas and New Years in the same place. Some migrate from family gatherings to social outings on the coast. Some sneak in a week of work in between.
In between Christmas and New Years there are only a few days to squeeze in a journey. Rationally, the NZTA would want to spread out the volume of traffic across the widest period. By producing forecasts and traffic warnings, however, rational outcomes don’t often occur.
The problem is that forecasts can appear to be too accurate and people rely on them too much. Seeing that off-peak starts at a certain time, people leave later - and arrive much later due to the unanticipated congestion.
If you look at the “hotspot” chart produced by NZTA, you will notice how there is red in the centre, surrounded by orange and green. Basically, it has many similarities with a “Normal Distribution” model (http://bit.ly/2islpjP). The twin peaks of Tuesday and Friday, however, are like a “Bimodal Distribution” (http://bit.ly/2iGpYKT).
Normal Distribution Model
The problem with identifying that Tuesday will be the worst day to travel is that it is the day after Boxing Day and many would see that traveling on Wednesday and Thursday looked more favourable. Some Wednesday and Friday travellers might also change to Thursday.
So, what started as two smaller peaks has become a much higher and steeper summit. Effectively, the NZTA has made congestion worse.