Carbon emissions contribute to global warming and greenhouse effects that are mainly caused by human production and life (Du et al., 2022). Cities, as the carriers of human activities, are responsible for more than 70% of the global carbon emission ratio (Li et al., 2018). China’s carbon dioxide (CO2) emissions have been the highest in the world since 2006 (Guo and Li, 2021). It is necessary to understand the spatial distribution of carbon emissions in Chinese cities at different scales.
At present, one method of accounting for carbon emissions is the top-down inventory algorithm, and the other is to fit the carbon emissions of different countries or cities through night-time light imagery (Zhao et al., 2012; Yu et al., 2015). Carbon emissions are mainly studied at national, provincial and city scales (Yang and Li, 2022). At the national scale, Doll et al. (2000) mapped global CO2 emissions based on the correlation between night-time light levels and CO2 emissions from DMSP/OLS night-time remote sensing imagery of countries at different development stages. At the provincial scale, Shi et al. (2016) proposed a panel data analysis by integrating night-time stable light (NSL) satellite data from DMSP-OLS to simulate China's CO2 emission dynamics. This study found a positive correlation between NSL data at a 1 km resolution and CO2 emissions based on China's provincial statistics from 1997 to 2012. Clarke-Sather et al. (2011) tested China’s interprovincial inequality in CO2 emissions based on commonly used inequality measures. At the city scale, Chuai and Feng (2019) examined the interprovincial inequality of CO2 emissions in China by collecting various sources of big data and designed a new carbon emission calculation method to examine the carbon emissions of Nanjing city with a high resolution of 300 m. Patino-Aroca et al. (2022) used a top-down methodology to assign carbon emissions at a spatial resolution of 1 km and a temporal resolution of 1 hour and addressed the development of a traffic emissions inventory for Guayaquil using the International Vehicle Emissions Model (IVE). Liu et al. (2022b) believed that the intensity of carbon emissions in cities is related to the rate of urbanization and energy consumption. However, these studies are mainly established on the macroscopic scale, which cannot truly express the disparity in carbon emission levels in different regions within cities or communities (Dong and Li, 2022). To clearly observe and compare the carbon emission levels in different regions, it is essential to study the carbon emissions caused by human activities at smaller scales, such as administrative regions or urban communities. The results can provide additional valuable suggestions for building low-carbon cities and developing low-carbon planning (Liu et al., 2015b; Liu et al., 2015a).
Currently, some researchers use more elaborate urban data that focus on carbon emissions from urban industries, buildings and residential life (Wang et al., 2015; Cai et al., 2018). Intracity land use, functional dynamics and transportation systems are three important factors influencing carbon emission levels (Shu et al., 2010; Xiao et al., 2021; Lu, 2018; Bu et al., 2022). For example, Huang et al. (2022a) realized the spatial distribution of urban carbon emissions at the micro level by calculating the spatial distribution of different sectors in Shanghai at a resolution of 30 m by using point-of-interest data and web crawling techniques. Li et al. (2021) used a cubic exponential smoothing model to predict traffic CO2 emissions using 13 influencing factors. Demuzere et al. (2014) explored the impact of green urban infrastructures and green spaces on improving climate change and urban carbon emissions. Jia et al. (2020) allocated the total residential carbon emissions to each parcel in Wuhan City using the correlation between night-time light and population density and then constructed a residential building carbon emission model consisting of parcel planning factors, socioeconomic factors, and single residential. In other words, the selection of quantifiable spatial characteristic indicators and how they are used to assess the spatial layout of carbon emissions are the key points of describing the CO2 emissions of small-scale spatial units. However, current studies lack uniform accounting data and methods that focus on these issues (Lu et al., 2017; Xu, 2022; Liu et al., 2022a).
To solve the problems mentioned above, this paper takes the Sino-Singapore Tianjin Eco-city as an example. Based on multisource data such as remote sensing data, urban planning data, points of interest, mobile signalling data and basic geographic information, we take the control detailed planning plots as the measurement unit and establish the carbon emission indicator system from three dimensions, i.e., industrial consumption, residential life and transportation, according to the current development state of urban industries. By using the analytic hierarchy process (AHP), the carbon emission indicator weights are calculated. The carbon emissions of each plot were assigned. The results can provide decisions and suggestions for urban low-carbon construction, land planning, energy conservation and emission reduction.