Abstract
Cities in China are on the frontline of low-carbon transition which requires monitoring city-level emissions with low-latency to support timely climate actions. Most existing CO2 emission inventories lag reality by more than one year and only provide annual totals. To improve the timeliness and temporal resolution of city-level emission inventories, we present Carbon Monitor Cities-China (CMCC), a near-real-time dataset of daily CO2 emissions from fossil fuel and cement production for 48 major high-emission cities in China. This dataset provides territory-based emission estimates from 2020-01-01 to 2021-12-31 for five sectors: power generation, residential (buildings and services), industry, ground transportation, and aviation. CMCC is developed based on an innovative framework that integrates bottom-up inventory construction and daily emission estimates from sectoral activities and models. Annual emissions show reasonable agreement with other datasets, and uncertainty ranges are estimated for each city and sector. CMCC provides valuable daily emission estimates that enable low-latency mitigation monitoring for cities in China.
Introduction
Urban areas in China are responsible for over 80% of the nation's fossil-fuel $\text{CO}_2$ emissions, making them central to achieving the "Dual Carbon" goals of peaking emissions by 2030 and reaching neutrality by 2060. However, effective climate action is often hindered by a lack of timely data. Traditional emission inventories, such as CEADs or MEIC, typically suffer from a reporting lag of one to two years and often lack the temporal resolution needed to understand the impact of specific events like heatwaves, holidays, or pandemic-related lockdowns.
To address this gap, the Carbon Monitor Cities-China (CMCC) dataset offers a near-real-time, daily-resolution inventory for 48 major high-emission cities. By integrating traditional bottom-up statistical accounting with high-frequency activity data, CMCC provides a high-resolution lens through which urban decarbonization efforts can be monitored with minimal latency.
Methodology
The construction of the CMCC dataset follows a three-stage framework designed to bridge the gap between annual statistics and daily activity.
- Annual Inventory Construction: The researchers first established annual bottom-up inventories for each city using IPCC territorial-based accounting. This involved analyzing energy balance tables and industrial sectoral energy consumption.
- Daily Activity Modeling: The core of the innovation lies in the use of high-frequency "proxy" data to disaggregate annual totals into daily estimates across five sectors:
- Power: Daily generation and consumption data sourced from local grid companies.
- Industry: Monthly production data for 35 key industrial products (e.g., crude steel, cement), disaggregated to daily scales using provincial power consumption patterns.
- Ground Transport: A sigmoid regression model that converts Gaode (AMAP) traffic congestion and delay indices into vehicle flux.
- Residential/Services: Building energy models (CBEM) driven by daily local temperatures, supplemented by retail sales data for the service sector.
- Aviation: Flight-by-flight reconstruction using ADS-B signals from Flightradar24.
- Data Integration and Gap Filling: Annual totals are disaggregated based on these daily variations. For cities with missing industrial data, a City Similarity Matrix was employed, using socioeconomic indicators (GDP, population, industrial structure) to "borrow" sectoral coefficients from similar, data-rich cities.
Key Insights and Results
The CMCC dataset reveals critical patterns in urban emissions that were previously obscured by annual reporting:
- High Impact of Selected Cities: The 48 cities in the dataset account for approximately 44% of China’s total $\text{CO}_2$ emissions, despite housing only 21% of the population. This underscores the potential for targeted urban policies to drive national-scale change.
- Capturing the "COVID Rebound": The daily resolution clearly illustrates the sharp drop in emissions during the early 2020 lockdowns and the subsequent "V-shaped" recovery in 2021.
- Sectoral Fingerprints: Different cities exhibit unique emission profiles. For instance, Shanghai shows significant spikes driven by summer cooling demand, while Harbin's emissions are dominated by winter heating. Tourism hubs like Haikou show distinct aviation-heavy profiles that fluctuate with holiday seasons.
- Accuracy and Uncertainty: The dataset demonstrates strong agreement with established annual inventories ($R^2$ of 0.82–0.83 for key sectors). Total emission uncertainty ranges from 2.54% to 30%, with lower uncertainty in cities that provide direct energy reporting.
Conclusion and Future Directions
While CMCC represents a significant advancement in emission monitoring, several areas for future development remain. The current scope excludes emissions from land use and forestry (LULUCF), and the ground transportation model relies heavily on calibration data from Beijing. Future iterations aim to incorporate more city-specific traffic data and expand the coverage beyond the initial 48 cities to include more of China’s 300+ prefectural-level cities. By providing low-latency, high-resolution data, CMCC serves as a vital tool for policymakers to track and adjust climate strategies in real-time.
Summary updated