But Jukka Heinonen and his colleague Seppo Junnila from Aalto University, Finland, have a different way of examining this issue. They believe that emissions should not be allocated to where they are produced, but to where they are consumed.

"For example, if a television is made in a big factory in the countryside, but bought by someone living in a city, the carbon emission generated from the production of that television should be allocated to the consumer, not the factory," Heinonen told environmentalresearchweb. "When you look at carbon consumption in this way it becomes almost irrelevant where someone lives and how dense the city is in which they live."

Heinonen and Junnila studied the two largest metropolitan areas in Finland: Helsinki and its two surrounding cities Espoo and Vantaa; and the important inland city of Tampere, together with the seven neighbouring semi-urban cities. The seven cities around Tampere were allocated into two groups: rural cities (RTC) and urban cities (UCT).

The pair found that carbon consumption was directly linked to income and was not necessarily related to the density of the city. "Espoo is a less dense city than Helsinki, but carbon consumption per capita is higher in Espoo than in Helsinki because Espoo is a more affluent city," said Heinonen.

To come to these conclusions, the researchers used a hybrid life cycle analysis (LCA) approach. This combines the principles of an input–output LCA – where emissions are calculated based on monetary transactions – and a process LCA, where emissions are assessed based on the energy and mass flows in the main production and supply chain processes.

Heinonen and Junnila looked at 10 consumption areas: heat and electricity; building and property; maintenance and operation; private transport; public transportation; consumer goods; leisure goods; leisure services; travelling abroad; and health, nursing and training services.

"We found that the biggest impacts on a consumer's carbon footprint are heat and electricity; the construction and maintenance of buildings; and private transport," said Heinonen. "Tampere is considerably more dense than the urban and rural cities surrounding it, but we found a negligible difference in carbon consumption between these three metropolitan areas."

The researchers believe that their study is a useful model for analysing the emissions of different urban structures that could be used in urban development when low-carbon solutions are sought. They have published their research in Environmental Research Letters (ERL).