Background & Summary

The persistent rise in global temperatures presents a complex challenge, directly contributing to a spectrum of climatic issues that threaten human survival. According to the Global Carbon Project’s 2023 report, global carbon emissions have reached an alarming level of approximately 40.9 billion tons1. A significant portion of these emissions, about 36.8 billion tons, originates from the combustion of fossil fuels1. Statistical data indicate that the global average temperature has risen by approximately 1.1 degrees Celsius (°C) since the Industrial Revolution2, and it is likely that global warming will exceed 1.5 °C before 2030 and may surpass 1.7 °C shortly thereafter1. The escalation of global temperatures has led to an increase in the frequency and intensity of extreme weather events, posing severe threats to global food and water security3,4,5,6,7,8,9,10, human health11,12,13,14,15,16, and the economic and social fabric17,18,19,20,21,22,23. The latest 28th Conference of the Parties (COP28) to the United Nations Framework Convention on Climate Change has conducted its first “Global Stocktake” to assess the progress of climate action since the signing of the Paris Agreement, underscoring an anticipated shortfall of 20.3 to 23.9 billion tonnes of CO2 equivalent in achieving the 2030 goals24. As the fifth-largest emitter, Japan’s annual fossil CO2 emissions in 2022 were 1075.07 million tons, accounting for 2.83% of the global total1. Japan’s efforts in emission reduction are of significant importance to the global response to climate change.

Japan’s carbon reduction efforts have yielded some results, with historical emissions since 1990 exhibiting a fluctuating decline across various sectors25. In 2020, Japan declared its intention to achieve carbon neutrality by the year 2050, a commitment that was enshrined in the Act on Promotion of Global Warming Countermeasures26. The subsequent year, Japan revised its Nationally Determined Contribution (NDC), elevating its reduction target from an initial 26% to a 46% cut by 2030 relative to 2013 emissions27. This revision is set to lower Japan’s absolute carbon emissions from 1.079 billion tons per year to 813 million tons per year. By the end of 2023, Japan had achieved approximately 20% of its emissions reduction target, yet the ambitious decarbonization goals still require more concerted efforts. In February 2023, the Japanese government passed the Green Transformation (GX) Basic Policy, aiming to strengthen decarbonization in key industrial sectors through the GX Alliance and to promote renewable energy as the main source of electricity28. Japan is also exploring positive measures, including the implementation of a carbon pricing policy based on the GX Promotion Act and halting the construction of new coal-fired power plants without reduction measures within the country29,30. Following the 2011 Fukushima nuclear disaster, Japan had once contemplated abandoning nuclear power development in favor of fossil fuels31. Recently, the government has been considering the restart of idle reactors and the construction of new ones28. While nuclear power may see a resurgence in Japan’s future, safety reviews, and political obstacles are likely to make it challenging for nuclear energy to contribute to Japan’s 2030 targets.

Carbon emissions inventories provide the mechanism for discerning comprehensive emission metrics and tracing source outlines within delineated geographic perimeters over established temporal spans32,33. The scientific rigor and precision of carbon emission inventories are critical in pinpointing emission sources, evaluating potential reduction measures, and shaping as well as appraising mitigation strategies34. Recent research has expanded the ambit of carbon footprint assessments, underscoring the significance of sector-specific analyses in domains such as households35,36,37,38,39, food systems10,40,41, and transportation sectors42,43,44. This underlines the need for quantifying carbon emissions with enhanced spatial and temporal resolution. With the improvement of fine-scale data and methodologies, attention is increasingly focused on compiling carbon emission inventories at the national45,46,47,48, regional34,49,50,51,52,53, and household54,55,56 levels, as well as within specific subdomains57,58. In service of decarbonization goals, there is a growing demand for more refined carbon emission inventories in future research59.

Japanese government agencies have not disclosed carbon emission data involving detailed regions and sectors31. To address this, Our previous work has developed an emissions inventory database that includes 47 prefectures and 26 sectors (including the power sector)31. Nonetheless, the database could be enhanced in two key areas: Firstly, it currently spans from 2007 to 2015, and incorporating the most recent energy consumption statistics from Japan warrants extending this timeframe to capture a more comprehensive temporal scope. Secondly, the granularity of the 26 sectors included in the database does not meet the nuanced requirements of future research; availability of updated data facilitates a more detailed sectoral breakdown.

To accommodate the emerging needs of future research, we have updated the dataset based on previous work. This dataset seeks to bridge the identified gaps by expanding the scope of Japan’s prefectural carbon emission inventory, both temporally and sectorally. The temporal coverage of the database has been expanded to encompass a broader timeframe, from 1990 to 2020, enabling a more extensive quantification of regional historical carbon emission fluctuations. Additionally, the sectors within the database have been further delineated, with the total number now increased to 32. This includes separating previously aggregated sectors and consolidating those that cannot be classified into a single category, as detailed in Table 1. Simultaneously, to match each emission result, we compiled a corresponding set of socioeconomic data, including population, Gross Domestic Product (GDP), land prices, green spaces, and roads. Consistent with the previous version, we continue to use data from Japanese power companies and power plants to calculate carbon emissions from the power sector, and additionally estimate carbon emissions from the use of coal, crude oil, and natural gas not utilized for power generation. Moreover, to eliminate the effects of price level fluctuations, the socioeconomic data set was converted to constant 2011 prices.

Table 1 Comparison between the updated and original sector classifications.
Full size table

Methodology

Sectoral emissions accounting

The scope of emission accounting includes Japan’s four major sectors: industrial sector, household sector, transportation sector and other non-energy sectors, totaling 32 sectors. The original data used for calculations come from energy consumption data from Japan’s prefecture-level natural resources and energy departments (URL: https://www.enecho.meti.go.jp/statistics/energy_consumption/ec002/results.html#headline4), including the consumption of fossil fuels such as coal, crude oil and natural gas in each department. However, there is a lack of three types of fuel consumption data for the power generation sector, so in addition to non-electricity fuel emissions, we calculated power generation emissions separately. Emissions from power generation are allocated through emissions from power plants and are not further divided by fuel type, but as a single emission source alongside the three major fuels. With the exception of the power generation sector, all other sectors are estimated using energy consumption data.

For the estimation of power generation emissions, we collected power generation data of major electric power companies in the target years from the Federation of Electric Power Companies of Japan (Fuel performance. URL: https://pdb.fepc.or.jp) and obtained capacity data for each power plant from Japan National Land Numerical Information (Category: Facilities. URL: https://nlftp.mlit.go.jp/ksj/gml/datalist/KsjTmplt-P03.html) for prefectural power generation emissions distribution. Regional power plants generally provide electricity for the local area and some surrounding cities, so the total emissions can be estimated by referring to the number and capacity scale of power companies in the prefecture. Combining the three fossil fuels, sectoral emissions of prefectures can be calculated by the following formula:

$$\begin{array}{c}E{P}_{pt}=CA{P}_{pt}\times \sum _{i=1}\left({L}_{\theta pt}^{i}\times {H}_{it}\times {I}_{it}\right)\end{array}$$
(1)
$$\begin{array}{c}{E}_{ijpt}={G}_{ijpt}\times {H}_{it}\times {I}_{it}\end{array}$$
(2)
$$\begin{array}{c}{E}_{pt}=\sum _{i=1}\sum _{j=1}{E}_{ijpt}+E{P}_{pt}\end{array}$$
(3)

Where EPpt is the power generation emissions in prefecture p in year t. CAPpt is the proportion of power plant installed capacity of prefecture p in the area where it is supported by power company θ. This parameter comes from Japan National Land Numerical Information (Category: Facilities). \({L}_{\theta pt}^{i}\) represents the total consumption of fuel type i in year t by the power company θ that provides power support for prefecture p. This parameter comes from Electricity Statistics Information of The Federation of Electric Power Companies of Japan (FEPC). Hit represents the calorific value generated per unit of fuel type i consumed in year t, and Iit represents the corresponding emission intensity. These two values come from the Prefecture Energy Statistics of Agency for Natural Resources and Energy of Japan, as shown in Table 2. The calorific value and emission intensity coefficient of each fuel type are given by year. Therefore, \({\sum }_{i=1}\left({L}_{\theta pt}^{i}\times {H}_{it}\times {I}_{it}\right)\) can be understood as the total emissions from power generation using three fossil fuels in each prefecture. In Eq. (2), Eijpt represents the total carbon dioxide emissions produced by sector j in prefecture p in year t consuming fuel type i. Gijpt represents the consumption of fuel type i in sector j in prefecture p in year t. The data is driven by prefectural energy statistics from the Prefecture Energy Statistics of Agency for Natural Resources and Energy of Japan. In Eq. (3), Ept represents the total emissions of prefecture p in year t.

Table 2 Fuel types and corresponding caloric value by year.
Full size table

Sectoral consumption on electricity and its allocation

As mentioned above, the absence of precise data on fossil fuel consumption at the prefectural level within the power generation sector necessitates an estimation of emissions based on the distribution of electricity consumption. We first gathered data on electricity usage from ten prominent power companies (see power_company_cover_region.xlsx in the Excel folder in figshare) over a specified timeframe. Subsequently, emissions were reassigned according to the respective capacities of local power plants in each prefecture, contributing to the overall emissions calculation at the prefectural level. In cases where multiple power companies support a single prefecture, we designate the company with the broadest coverage as the primary contributor. Given that the most recent power generation data is available only up to 2015, we extrapolate and utilize the same statistical values from that year for subsequent years.

Socioeconomic data

The socio-economic characteristics of cities are often used as input variables in carbon emission-driven models. Multivariate regression modeling between socioeconomic and carbon emissions data can help scholars understand what characteristics influence carbon emissions and to what extent, and how this influence varies across geographies and sectors. Therefore, to match each emission result, we additionally transformed the corresponding socioeconomic data set, including population, GDP, land prices, green space, and roads. The population data of each prefecture from 1990 to 2020 comes from the Statistics Bureau of Japan (Census. URL: https://www.e-stat.go.jp/). Prefectural GDP data comes from the Prefectural Final Accounts Annual Report of the Cabinet Office of Japan (URL: https://www.cao.go.jp/). Land price, green space and road data come from the Japan Land Numerical Information website (URL: https://nlftp.mlit.go.jp/ksj/index.html). Socio-economic indicators are recorded as follows:

  1. 1.

    Population (unit: person)

  2. 2.

    GDP (unit: million yen)

  3. 3.

    Average land price (unit: yen per square meter)

  4. 4.

    Per capita green space area (unit: square meters per person)

  5. 5.

    Road density (meters per square kilometer)

Data Records

This dataset encompasses a comprehensive compilation of 72,192 data records, spanning 47 prefectures and 32 sectors (as delineated in Table 3), capturing energy consumption based on three types of fossil fuels and one secondary energy source (electricity) over 16 years (Data for 1991–2004 and 2006 are not collected in this dataset). The dataset, denominated “Extension of Japan Prefectural Emission Accounting and Enrich Socioeconomic Data 1990 to 2020,” is publicly accessible via Figshare60. To provide a detailed breakdown, the database is structured into seven main components. Firstly, it comprises an Excel folder containing data on prefecture-level heat generation, emissions, population, GDP, land prices, park area per capita, and road density in Japan from 1990 to 2020, offering valuable insights into energy consumption, economic indicators, and environmental factors. Secondly, six distinct folders house a historical dataset on prefecture-level heat generation and emissions categorized by energy sources and industries in Japan, spanning from 1990 to 2020. These folders also encompass metrics for heat generation and emissions normalized by GDP and per capita, presented as ‘.shp’ files, specifically designated as ‘Heat’, ‘Heat per capita’, ‘Heat per GDP’, ‘Emission’, ‘Emission per capita’, and ‘Emission per GDP’ respectively. Furthermore, the dataset incorporates a comprehensive socioeconomic inventory, featuring a substantial number of records, which comprehensively covers five major indicators. The flowchart of this dataset is shown as Fig. 1.

Table 3 Database Structure and Description.
Full size table
Fig. 1
figure 1

Flowchart of this dataset.

Full size image

Technical Validation

Total sectoral emission by years

Annual emissions by sector from 1990 to 2020 are given in Fig. 2(B). In addition to the power generation sector, the five sectors with the highest annual emissions in the observed years are “Manufacture of Chemical and Allied Products, Oil and Coal Products”, “Manufacture of Iron and Steel”, “Non-Energy”, “Residential” and “Transportation”. Fig. 2(A) provides a further understanding of emissions from the “Manufacturing of Chemical and Allied Products, Oil and Coal Products” sector in three years: 1990, 2015 and 2020, including the proportion of carbon emissions caused by fuel use nationwide, the emission and emission proportions of the three types of fuels in each region. The results show that from 1990 to 2015, the sector’s share of national emissions caused by burning coal decreased by 7.6%, while the share of oil and gas increased by 6.9% and 0.7% respectively. The proportion of oil has almost remained unchanged from 2015 to 2020, and a small amount (0.6%) of carbon emissions has shifted from coal to gas. The sector in central Japan has gradually shed its reliance on coal and oil over time and was dominated by natural gas by 2020 Fig. 2.

Fig. 2
figure 2

(A) Prefectural carbon dioxide emission of “Manufacture of Chemical and Allied Products, Oil and Coal Products” sector (unit: 1000tCO2). The donut pie chart represents the national share of emissions from the sector’s use of three fossil fuels. The pie chart on the map represents the share of emissions from each prefecture’s use of the three fossil fuels in that sector. The size of the pie chart represents the amount of emissions from that sector. (B) Sectoral carbon dioxide emission from 1990 to 2020 (unit: MtCO2).

Full size image

Comparison with other estimation results

This data set compiles Japan’s prefecture-level emissions inventory based on energy consumption by sector. Here we compare the estimation results with The GHG Emission Data of Japan (URL: http://www.nies.go.jp/gio/en/archive/index.html) provided by the Greenhouse Gas Inventory Office of Japan (GIO) and the National Greenhouse Gas Inventory Report of Japan (NIR) released in April 2022 (URL: https://www.nies.go.jp/gio/en/archive/nir/index.html) (Fig. 3). The differences were largest in 1990, where the GIO difference ratio (assessment gap/GIO value) and NIR difference ratio (assessment gap/NIR value) were 13.2% and 15.2% respectively. The reason for the assessment gap may be changes in departmental classification. Japan’s Standard Industrial Classification was revised twice in 1993 and 2002, resulting in some newly added departments not being included in the 1990 energy consumption accounting. The biggest difference between the estimates after 2005 and those from other national estimates appeared around 2011. For example, the GIO difference ratios are 3.5% (2011) and 4.7% (2012), and the NIR difference ratios are 4.2% (2011) and 5.4% (2012). The difference in assessment mainly occurs in the assessment of power generation, because part of the power generation burden caused by the shutdown of nuclear power plants after the Great East Japan Earthquake in 2011 was transferred to private power generation (non-utility power generation facilities). The power generation and fuel consumption of these private power generation facilities are difficult to accurately survey and quantify. Therefore, the power generation estimates in this data set do not include private power generation, resulting in a gap between around 2011 and other assessment results. After 2015, the estimates from this data set gradually approached and exceeded other national estimates. The reason is that due to the lack of power generation data from 2016 to 2020, this study uses the same baseline value as in 2015 for power generation estimates in years after 2015. This may lead to overestimation of emissions from the power generation sector.

Fig. 3
figure 3

Comparison with other estimation results.

Full size image

Usage Note

The dataset is a pivotal resource for a variety of applications in environmental and socio-economic research. This extensive dataset, encompassing 96,256 records across 47 prefectures and 32 sectors, provides a detailed view of Japan’s carbon emissions and socioeconomic parameters over time. This compilation is derived from a meticulous amalgamation of data from multiple sources, including the electricity consumption dataset and publicly available data from the Japanese Government Statistics. Researchers and experts are afforded the flexibility to independently extract and process information from any database, adhering to our outlined methodology for data cleaning and enhancement.

As for the potential usage, policymakers and environmental planners can utilize this dataset to assess the impact of energy consumption in different sectors and prefectures on carbon emissions, which helps to inform targeted and effective environmental policies and initiatives at both local and national levels. Moreover, our dataset provides insights into the consumption patterns of fossil fuels and electricity across various sectors. Energy companies and consultants can analyze this data to identify trends and make informed decisions regarding energy production and distribution. The dataset’s strength lies in its detailed sectoral and regional breakdown, which, when used cautiously, can yield highly valuable insights for a wide range of applications. This dataset is openly accessible to the public, subject to the terms of the Creative Commons License with attribution (CC-BY 4.0).