Comparison of city-level carbon footprint evaluation by applying single- and multi-regional input-output tables.

City-level carbon footprint has been recognized as a useful measure of anthropogenic impact on climate change associated with citizens' activities within the administrative boundary. Although the promotion of consumer responsibility suggests rethinking urban indirect emissions, the detailed methodology is far from satisfactory for realistic applications. Due to the lack of multi-regional input-output tables for most cities, there is a wide application of single regional input-output tables. However, there still lacks further discussion on if there will be an obvious evaluation bias by applying city-level single-regional tables rather than multi-regional ones. To visualize the table coverage on its application consequence, both single- and multi-regional input-output tables were employed to compare disparities in the carbon footprint accounting in the case of Tokyo, Japan. Our analysis shows that the gap of emissions driven by Tokyo's final demand between single- and multi-regional input-output tables was considerably large. Furthermore, the results of multi-regional table were found to be 8.11 MtC higher for coal-generated emissions, 7.83 MtC for crude oil-generated emissions and 2.90 MtC for natural gas-generated emissions than those of the single-regional table. The largest deviation in emissions accounting was observed in the power, gas and heating supply sector, the construction sector and the private service sector. The gap between these two input-output tables was notable for all three types of fossil fuels (coal, crude oil and natural gas). These indicated that coal-generated emissions have been largely ignored by single-regional input-output table. The study highlighted the difference of carbon footprint accounting between these two types of input-output tables. Our findings are intended to assist policymakers and scholars in pinpointing and reallocating sectors that are likely to yield severely biased evaluation of emissions embodied in trade when a multi-regional table is not available.