Diagrams published by Sturman & Tapper (1) show temperatures at 2pm and 10.30pm on a calm winter’s day in Christchurch, New Zealand, in 1979. The centre of town was about 16° C in the afternoon, i.e. 2 K warmer than the coast 5 km away, but 9° C at night, i.e. 4 K warmer than the outskirts of the city. The night was clear and a 500 m deep inversion was present over the outskirts of the city. Measurements above the city at night show that lapse conditions, due to urban heating, were present to a height of about 300 m. There was a weak inversion to about 600 m and lapse conditions prevailed above that level. We conclude that on calm nights urban heating causes pollutant mixing in a deeper layer over cities than over the surrounding countryside.
The figures above are consistent with a relationship between the maximum urban heating observed in New Zealand and Australia, and city population. The maximum is about 2 K for populations of 1,000 people, 4 K for 100,000, 6 K for a city like Melbourne with about 3 million (2). The corresponding temperatures in North America are about 2 K, 8 K and 12 K, implying greater concentrations of energy consumption per unit area. The magnitude of urban heating in European cities is intermediate between Australian and US figures.
(1) Sturman, A.P. & N. Tapper, 1996: The Weather & Climate of Australia & New Zealand (Oxford), on page 332.
(2) as above but page 334, from Torok et al 1996.