Potential impact of abiotic resource use considering country-specific supply chain: consumption-based characterization and normalization utilizing a multi-regional input–output model
Der Artikel geht auf das kritische Thema der abiotischen Ressourcennutzung und ihrer Auswirkungen auf die Umwelt ein und betont die Rolle globaler Lieferketten. Es führt einen neuartigen Ansatz ein, der ein multiregionales Input-Output-Modell zur Charakterisierung und Normalisierung der potenziellen Auswirkungen der Ressourcengewinnung und des Ressourcenverbrauchs verwendet. Die Studie konzentriert sich auf 29 mineralische Ressourcen und drei fossile Brennstoffe, die 189 Länder und Regionen abdecken, um eine detaillierte Bewertung des ökologischen Fußabdrucks im Zusammenhang mit der Nutzung abiotischer Ressourcen zu liefern. Durch die Aufschlüsselung des Bergbausektors in der Eora MRIO-Datenbank identifiziert die Forschung die Minenproduktion, die von Verbraucherländern verursacht wurde, und bietet eine einzigartige Perspektive auf die Verantwortung der verschiedenen Nationen für die Umwelt. Die Ergebnisse zeigen signifikante Unterschiede bei verbrauchsbasierten Charakterisierungsfaktoren und Normalisierungswerten, was die Bedeutung der Berücksichtigung länderspezifischer Lieferketten bei der Bewertung der Umweltauswirkungen unterstreicht. Der Artikel beleuchtet auch die Grenzen bestehender Modelle auf globaler Ebene und die Notwendigkeit regional eindeutigerer Daten, um die Umweltauswirkungen der abiotischen Ressourcennutzung präzise zu bewerten. Anhand einer Fallstudie zu Lithium-Ionen-Batterien zeigt die Forschung die praktische Anwendung verbrauchsabhängiger Charakterisierungsfaktoren auf und zeigt, wie unterschiedliche Verbraucherländer unterschiedliche Umweltauswirkungen haben können. Die Studie schließt mit der Diskussion der Auswirkungen auf nachhaltiges Ressourcenmanagement und des Potenzials für zukünftige Forschung in diesem Bereich.
KI-Generiert
Diese Zusammenfassung des Fachinhalts wurde mit Hilfe von KI generiert.
Abstract
Purpose
Life cycle impact assessment (LCIA) quantifies the potential impacts of environmental loads in the characterization step and evaluates their significance in the normalization step. This study develops consumption-based characterization factors (CFs) and normalization values (NVs) for abiotic resource use using a multi-regional input–output model. Consumption-based accounting in these steps is critical for ensuring consistency between the scope of these steps in LCIA and the overall evaluation scope of the life cycle assessment (LCA) study.
Methods
We calculated the consumption-based CFs and NVs for the target year 2015, covering 189 countries and regions for 29 mineral resources and three fossil fuels. These calculations were based on estimates of induced mine production and the adoption of country-level mine production-based CFs. The mining sector in the Eora multi-regional input–output database was disaggregated using country-level mine production data and trade statistics, enabling the estimation of induced mine production for each consuming country. A user cost model was adopted to calculate country-level mine production-based CFs.
Results and discussion
Consumption-based CFs varied significantly, with the maximum values being up to 3,300 times higher than the minimum, depending on the consuming countries. This highlights the importance of considering supply chain differences when assessing the potential impacts of abiotic resource use. The USA had the largest consumption-based NVs, followed by China and Brazil, with fossil fuels as the primary contributing resources. Some countries (e.g., Japan and Germany) exhibited notably higher consumption-based NVs compared to mine production-based NVs, reflecting their scarcity of primary resources in-country and underscoring the relevance of consumption-based NVs. Compared to previous studies with lower resolution for target resources, the consumption-based NVs developed in this study, which differentiate a greater variety of resources, offer more plausible results and enable more flexible analyses targeting specific resources.
Conclusions
Consumption-based accounting of the potential impacts associated with resource use can support LCA practitioners in conducting region-specific analyses without the need to identify mining countries for abiotic resources. The findings can also be utilized to analyze the responsibilities of consuming countries, sustainable supply chain management, and country-level supply risk assessments. Future work should focus on improving the disaggregation of metal-related sectors using smelter, refinery, and manufacturing data.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
1 Introduction
Human activities and economic growth are underpinned by consumption activities, which results in diverse environmental impacts and reduced resource availability worldwide through the global supply chain. Achieving sustainable development requires effectively reducing these environmental impacts associated with consumption activities while fulfilling our needs. Life cycle assessment (LCA) is a standardized methodology used to support sustainable development by accounting for environmental loads and evaluating the environmental impacts associated with products, services, and activities along the supply chains (ISO 2006). Life cycle impact assessment (LCIA), the third phase of LCA, quantifies the potential impacts associated with the inputs and outputs of the processes (i.e., the results of the life cycle inventory (LCI) analysis phase) in the characterization step and, as an optional step, evaluates their relative environmental significance in the normalization step.
Characterization quantifies the potential impacts of environmental loads (i.e., emissions of substances and resource extraction) by multiplying the environmental loads (i.e., LCI results) by the characterization factors (CFs), which represent the potential impact per unit of emission or extraction. Since the impacts caused by emissions (or resource extraction) depend on several factors such as the quantity of emitted substances (or extracted resources), the properties of substances (or resources), the characteristics of the emitting sources (or extracted sources), and the nature of the receiving environment (Finnveden et al. 2009), characterization of region-specific impacts has been developed for various impact categories including abiotic resource use (Mutel et al. 2019; Pennington et al. 2004; Potting and Hauschild 2006; Yokoi et al. 2024a). Although spatial differentiation in characterization is significant for increasing the reliability of LCA results, it requires not only the development of regional CFs but also regionally explicit LCI data, and linking LCI and LCIA data is a significant challenge for the regionalization of LCA (Mutel et al. 2019; Pfister et al. 2020).
Anzeige
Normalization calculates the magnitude of environmental impacts (i.e., characterization results) by relating them to reference values (i.e., dividing them by normalization values (NVs)) and can facilitate the interpretation of characterization results for different impact categories (ISO 2006; Pizzol et al. 2017). The normalization result for each impact category is generally expressed as dimensionless and thus allows for the comparison of the relative magnitude of different environmental impacts caused by the target activity as relative values to the NVs, which also serves as preparation for the integration of different impacts into a single value in the next step (i.e., weighting). NVs significantly influence LCIA results and have been proposed for different geographical scales and perspectives (i.e., production and consumption-based accounting). A typical example of an NV is the global environmental impact for the target year. Using global-scale NV, characterization results are evaluated as values relative to global-scale impacts. Alternatively, NVs can be calculated on a regional or country scale, and these NVs could be suitable when evaluators focus on specific regions (Li et al. 2024). Unlike global-scale NVs, there are two approaches to calculate regional-level NVs: production- and consumption-based approaches (Laurent and Hauschild 2015). Production-based NVs are commonly applied because they are easier to calculate with less data compared to consumption-based NVs (De Laurentiis et al. 2023). However, considering the globalized nature of supply chains, most LCA studies need to cover environmental loads induced by activities beyond the target region, in addition to those generated within it. Therefore, consumption-based NVs are important for ensuring the consistency of the system boundary and scope between normalization and LCI (Crenna et al. 2019) (Fig. 1).
Fig. 1
Concept of consumption-based characterization and normalization. CFs: characterization factors; NVs: normalization values; M-based: mine production-based; C-based: consumption-based
Among the various impact categories, this study focuses on abiotic resource use, a critical impact category included in most of the existing LCIA methods (Finnveden et al. 2009). Compared to other impact categories, abiotic resource use is uniquely characterized by the limited locations of production (i.e., mining countries), resulting in a substantial disparity between the distribution and number of mining and consuming countries. Because not only the amount of abiotic resource consumption but also supply chains and consequently sourcing countries differ among consuming countries, developing consumption-based characterization as well as normalization is relevant for abiotic resource use. However, regional CFs have predominantly focused on emission sources or mining countries (i.e., production-based characterization). In contrast, consumption-based CFs, which represent impacts induced by a unit of used abiotic resource rather than a unit of extracted abiotic resource (Fig. 1), remain underdeveloped. Although the application of production-based regional CFs for abiotic resources requires regionally explicit LCI data (i.e., mining flow), the supply chains of products are generally globalized and complicated, making it difficult for LCA practitioners to identify the mining countries of the abiotic resources contained in the products. In this context, consumption-based CF for abiotic resource use offers a more practical approach. These CFs represent the potential impacts of mining abiotic resources used in consuming countries, considering the average supply chains of those countries. They are applied to consumption flows in consuming countries rather than mining flows in mining countries, enabling regionalized characterization without identifying the mining countries of the used abiotic resources in a supply chain. On the other hand, consumption-based normalization (environmental footprint), including abiotic resource use, has been developed (Beylot et al. 2019, 2020; Breedveld et al. 1999; Corrado et al. 2020; Dahlbo et al. 2013; De Laurentiis et al. 2023; Sala and Castellani 2019; Sala et al. 2019; Sanyé-Mengual and Sala 2023). However, the scope of these studies remained limited, with gaps in the coverage of resources and countries, resulting in a lack of comprehensive consumption-based normalization for abiotic resource use (Berthet et al. 2024). For detailed information about the resources and countries covered by these studies, refer to Table 1.
Table 1
Coverage of studies on consumption-based normalization, including abiotic resource use
aBecause the process-based approach was applied, the coverage of resources can be assumed to be the same as those of the adopted LCI database and the LCIA method
Consumption-based NVs can be calculated using two approaches: process- and input–output (IO)-based approaches (De Laurentiis et al. 2023). The process-based approach allows for a detailed analysis of NVs but requires extensive data, making it difficult to track supply chains and identify mining countries. In contrast, the IO-based approach utilizes a multi-regional input–output (MRIO) model and has the potential to overcome these challenges because the MRIO model can track global supply chains. However, the IO-based approach is constrained by the heterogeneity of the industrial sectors in the IO tables, which limits both the coverage and resolution of target resources (De Laurentiis et al. 2023). The MRIO model has been widely used for consumption-based accounting to estimate the induced raw material extraction (material footprint) (e.g., Bruckner et al. 2012; Lenzen et al. 2022; Wiedmann et al. 2015) and metal-specific induced mining (e.g., Nakajima et al. 2019; Nansai et al. 2015; Vivanco et al. 2017; Wieland et al. 2022). Although the development of CFs and NVs requires wide coverage of target resources and distinguishing resource types, distinguished or targeted resource types are limited in existing studies.
Here, we tackle the two challenges mentioned above: the difficulty in identifying the mining countries of the abiotic resources contained in the target products for country-level characterization and the limited coverage of the target resources and countries for consumption-based NVs for abiotic resource use. To address these challenges, this study aimed to develop consumption-based CFs and NVs for abiotic resource use utilizing the MRIO model. By disaggregating the mining sector in the Eora MRIO database (Lenzen et al. 2012a, 2013a), we expanded the range of resource types analyzed and identified mine production in each mining country induced by each resource-consuming country. By adopting country-level mine production-based CFs for abiotic resource use, we calculated consumption-based CFs and NVs for the target year 2015 for 189 countries and regions for 29 mineral resources and three fossil fuels: silver (Ag), aluminum (Al), gold (Au), boron (B), barium (Ba), cobalt (Co), chromium (Cr), copper (Cu), fluorine (F), iron (Fe), lithium (Li), magnesium (Mg), manganese (Mn), molybdenum (Mo), niobium (Nb), nickel (Ni), phosphorus (P), lead (Pb), palladium (Pd), platinum (Pt), rhenium (Re), antimony (Sb), tin (Sn), tantalum (Ta), titanium, (Ti), uranium (U), vanadium (V), tungsten (W), zinc (Zn), oil, natural gas (NG), and coal. The concepts of consumption-based characterization and normalization are depicted in Fig. 1.
Anzeige
2 Methods
2.1 Disaggregation of the mining sectors of Eora database
This study utilized the Eora database in 2015 for MRIO tables to estimate mine production induced by resource-consuming countries. Eora is one of the largest MRIO databases, covering 189 countries and regions and containing more than 15,000 industrial sectors. It has been widely utilized to analyze the consumption-based accounting of various environmental impacts, such as greenhouse gas emissions (Kanemoto et al. 2016), air pollution (Kanemoto et al. 2014; Moran and Kanemoto 2016), biodiversity (Lenzen et al. 2012b), water (Lenzen et al. 2013b), and materials (Wiedmann et al. 2015). For this study, the mining sector for each country in the Eora database was disaggregated into 51 categories. There are various methods for disaggregating input–output tables depending on the disaggregation target (e.g., sectors, regions) and the availability of data (Wenz et al. 2015). This study employed a simplified method based on proxy data related to material flows. The disaggregation was based on mine production data in 2015, sourced from the British Geological Survey, Energy Institute, Nuclear Energy Agency, and the United States Geological Survey (BGS 2021; Energy Institute 2023; NEA 2016; USGS 2017a, 2024), as well as international trade data for 2015 obtained from the United Nations Comtrade database (UN Comtrade 2024). The 51 disaggregated mining sectors are consistent with those used in UN Comtrade, ensuring trade values for commodities are accurately captured (Table S1). This alignment enables precise assessment of the proportion of resources refined domestically versus those exported and refined abroad. The detailed method for disaggregating the mining sectors of Eora database is as follows.
First, the gross output of the original mining sector in mining country \(r\), denoted \({x}_{\text{mining}}^{r}\), was split into disaggregated mining sector-specific gross outputs \({x}_{i}^{r}\) (2015$) using the following equation:
where \({q}_{i}^{r}\) is the production volume of minerals corresponding to disaggregated mining sector \(i\) in mining country \(r\), obtained from sources such as USGS (see above) (kg), and \({p}_{i}^{r}\) is the unit price of commodities corresponding to disaggregated mining sector \(i\), calculated from export physical and monetary values in mining country \(r\) reported by UN Comtrade (2015$/kg). Where export data were unavailable, import data and global average prices were used.
Next, trade flows between countries were estimated. For exports and imports between different countries (\(r\ne s\)), intermediate inputs of minerals corresponding to disaggregated mining sector \(i\) from mining country \(r\) to importing country \(s\), denoted \({T}_{i}^{r,s}\) (2015$), were calculated by allocating the original mining sector’s intermediate inputs based on export values:
where \({e}_{i}^{r,s}\) is the export value of commodities corresponding to disaggregated mining sector \(i\) from mining country \(r\) to importing country \(s\) (2015$). Domestic intermediate inputs \({T}_{i}^{r,r}\) (2015$) were calculated by subtracting total exports to other countries from mining sector-specific gross outputs:
Intermediate inputs were then decomposed further by sector. Letting \(j\) represent general sectors, the decomposition of intermediate inputs to identify which sectors consume commodities corresponding to disaggregated mining sector \(i\), denoted \({T}_{i,j}^{r,s}\) (2015$), was calculated as:
Similarly, the decomposition of intermediate inputs to identify which commodities are used as inputs to produce commodities corresponding to disaggregated mining sector \(i\), denoted \({T}_{j,i}^{s,r}\) (2015$), was calculated using the proportion of each commodity’s intermediate inputs into the mining sector:
Finally, value added for disaggregated mining sector \(i\) in mining country \(r\), denoted \({v}_{i}^{r}\) (2015$), was calculated by subtracting the sum of intermediate inputs from the mining sector-specific gross outputs:
$$v_i^r=x_i^r-\sum_{s,j}T_{j,i}^{s,r}$$
(6)
2.2 Mine production induced by the consuming countries
By disaggregating the mining sector in the Eora database, it becomes possible to calculate the induced output for each of the 51 disaggregated mining sectors by consuming country. Mine production of abiotic resource \(g\) in mining country \(r\) induced by consuming country \(t\) (\({IND}_{g}^{r,t}\)) (kg) was estimated by allocating the mine production of resource \(g\) in country \(r\) to each consuming country based on the induced outputs of the mining sectors corresponding to resource \(g\) by the final demands of the consuming countries:
where \(r\) and \(t\) are the mining and consuming countries, \(i\) and \(k\) represent the origin (disaggregated mining) and destination sectors, respectively. \({\text{Output}}_{g}^{r,t}\) represents the output of the sectors corresponding to the mining sector of resource \(g\) in country \(r\) induced by consuming country \(t\) (2015$), and \({\text{Output}}_{g}^{r,\text{Total}}\) represents the total output of the sectors corresponding to the mining sector of resource \(g\) in country \(r\) (2015$). \({d}_{g,i}^{r}\) is a row vector, where the element is 1 if sector \(i\) corresponds to the mining sector of resource \(g\) in country \(r\); otherwise, it is 0. \(L\) is the Leontief inverse matrix, which represents the output of each sector in each country induced by one unit of the final demand for each sector in each country. \({y}^{t}\) denotes the column vector of the final demand in country \(t\), and \({P}_{g}^{r}\) is the mine production of resource \(g\) in country \(r\) (kg). It should be noted that \({IND}_{g}^{r,t}\) also includes the induced mine production in cases where the consuming country and the mining country are the same (i.e., \(r=t\)). The correspondence between the target 32 abiotic resources and the disaggregated mining sectors in the Eora MRIO table is summarized in Table S2.
2.3 Consumption-based characterization of abiotic resource use
Due to differences in supply chains and sourcing countries among consuming countries, as well as the potential impacts of resource mining among mining countries, the potential impacts of resource use also differ across consuming countries. We developed consumption-based characterization factors (CFs) as the weighted average of mine production-based CFs, based on the induced mine production in mining countries by the consuming country:
where \({CF}_{g}^{t,\text{C}}\) and \({CF}_{g}^{r,\text{M}}\) denote the consumption-based CF for resource \(g\) used in country \(t\) and the mine production-based CF for resource \(g\) mined in country \(r\), respectively. In this study, we applied country-level CFs for abiotic resource use based on a user cost model for mine production-based CFs (Yokoi et al. 2024a). The user cost model assesses the external costs of abiotic resource mining and is one of the recommended characterization models for evaluating the externalities of abiotic resource use within the Area of Protection “natural resources” in international communities (Berger et al. 2020; Sonderegger et al. 2020). The user cost focuses on the potential impacts of abiotic resource mining on future resource availability and represents the investment required to sustain a constant income after a mine is closed (El Serafy 1989). The CFs based on user costs were calculated for 193 countries and regions in 2020 by Yokoi et al. (2024a), whereas the mine production-based CFs for 2015 were calculated and adjusted for the 189 target countries and regions in this study. Therefore, the consumption-based CF based on the user cost model (\({CF}_{g}^{t,\text{C}}\)) refers to the external cost (potential impact) associated with the mining of 1 kg of an abiotic resource induced by a consuming country (2015$/kg). It should be noted that other country-specific characterization models can also be applied as mine production-based CFs for abiotic resources in calculating consumption-based CFs. However, the user cost model is the only available model for country-specific mine production-based CFs at present, so this study adopted the user cost model.
2.4 Consumption-based normalization of abiotic resource use
Using the inducing mine production and mine production-based CFs, we calculated the consumption-based normalization values (\({NV}^{\text{C}}\)) (2015$) as follows:
As indicated in Eq. (11), the consumption-based NVs for a specific resource for a specific country can be calculated using the total inducing mine production and consumption-based CFs for that country, without the need to trace the global supply chain. The consumption-based NVs represent the total potential impacts that abiotic resource use in consuming countries induced during the target year (i.e., inducing impacts). In this study, mine production-based NVs (\({NV}^{\text{M}}\)) were also calculated for comparison as follows:
$$NV_g^{r,M}=P_g^rCF_g^{r,M}$$
(12)
3 Results
3.1 Consumption-based characterization factors of abiotic resource use
The distributions of the consumption-based CFs for each abiotic resource are presented in Fig. 2, arranged in descending order based on the weighted averages of the consumption-based CFs by inducing mine production (lists of mine production- and consumption-based CFs and induced mine production are provided in the Supplementary Excel file). Since the weighted averages of consumption-based CFs are the same as those of mine production-based CFs, the trends and characteristics of the order of the weighted average of consumption-based CFs follow those of mine production-based CFs discussed in a previous study (Yokoi et al. 2024a). The abiotic resources with the largest variations in consumption-based CFs, as measured by the coefficient of variation, were Li, Pt, and Ti. In contrast, Zn, Cu, and Au showed the lowest variations (Table S3). Li exhibited the highest variability between the maximum and minimum consumption-based CFs, followed by P and Al, with their ratios of the highest CF to the lowest CF being 3.3 × 103, 2.9 × 102, and 2.0 × 102, respectively (Tables 2 and 3). These findings suggest significant differences in the potential impacts of the use of some abiotic resources, depending on the consuming country. As indicated in Eq. (10), consumption-based CFs are determined by the share of inducing mine production and the mine production-based CF for each mining country. Analysis of the relationship between the coefficient of variation for consumption- and mine production-based CFs, and the share of inducing mine production, suggests that variations in consumption-based CFs across consuming countries primarily stem from differences in mine production-based CFs among mining countries (Table S3 and Fig. S1).
Fig. 2
Violin plot of consumption-based characterization factors of abiotic resource use in 2015 in descending order of weighted averages by inducing mine production. A logarithmic scale is used for the vertical axis
Countries with the highest consumption-based characterization factors and their top-inducing countries in 2015. C-based, consumption-based; M-based, mine production-base
Country
C-based CF [2015$/kg]
Top inducing country (share)
M-based CF [2015$/kg]
Weighted average of CF [2015$/kg]
Ag
Viet Nam
3.5E + 02
China (79%)
3.7E + 02
2.8E + 02
Al
Malaysia
1.4E + 00
Malaysia (71%)
1.8E + 00
6.8E − 01
Au
Ghana
2.6E + 04
Ghana (88%)
2.7E + 04
2.2E + 04
B
Argentina
1.0E − 01
Argentina (92%)
1.1E − 01
4.9E − 02
Ba
Morocco
2.5E − 01
Morocco (97%)
2.5E − 01
1.7E − 01
Co
Botswana
1.7E + 01
South Africa (69%)
2.2E + 01
7.8E + 00
Cr
Turkey
9.7E + 00
Turkey (90%)
1.0E + 01
7.3E + 00
Cu
Canada
2.9E + 00
Canada (52%)
3.5E + 00
2.0E + 00
F
Morocco
2.6E − 01
Morocco (94%)
2.7E − 01
2.1E − 01
Fe
South Africa
5.6E − 02
South Africa (35%)
8.7E − 02
4.3E − 02
Li
Myanmar
1.4E + 01
Zimbabwe (95%)
1.5E + 01
8.6E − 01
Mg
Spain
7.6E − 01
Spain (76%)
9.4E − 01
2.7E − 01
Mn
Gabon
1.8E + 00
Gabon (99%)
1.8E + 00
1.0E + 00
Mo
Iran
9.0E + 00
Iran (78%)
1.0E + 01
3.1E + 00
Nb
Canada
1.2E + 01
Canada (90%)
1.3E + 01
5.2E + 00
Ni
British Virgin Islands
8.9E + 00
Philippines (94%)
9.2E + 00
5.2E + 00
P
China
3.3E − 02
China (96%)
3.4E − 02
2.2E − 02
Pb
Viet Nam
1.6E + 00
China (96%)
1.6E + 00
1.3E + 00
Pd
Kazakhstan
1.6E + 04
Russia (98%)
1.6E + 04
8.7E + 03
Pt
Kazakhstan
2.3E + 04
Russia (93%)
2.5E + 04
4.8E + 03
Re
Chile
6.2E + 02
Chile (99%)
6.2E + 02
5.4E + 02
Sb
Hong Kong
6.0E + 00
China (87%)
6.3E + 00
5.8E + 00
Sn
Myanmar
2.0E + 01
Myanmar (99%)
2.0E + 01
1.5E + 01
Ta
Brazil
1.0E − 02
DR Congo (61%)
2.8E − 03
5.1E − 03
Ti
Viet Nam
7.7E + 00
Viet Nam (92%)
8.2E + 00
2.0E + 00
U
Kazakhstan
3.1E + 01
Kazakhstan (99%)
3.2E + 01
2.1E + 01
V
China
2.7E − 02
China (90%)
2.9E − 02
2.0E − 02
W
Viet Nam
2.2E + 01
Viet Nam (83%)
2.3E + 01
1.8E + 01
Zn
Ireland
1.6E + 00
Ireland (52%)
1.8E + 00
1.3E + 00
Oil
Aruba
5.7E − 01
Colombia (80%)
6.5E − 01
2.6E − 01
NG
Bangladesh
3.7E − 01
Bangladesh (92%)
3.8E − 01
2.0E − 01
Coal
Viet Nam
8.1E − 02
Viet Nam (54%)
1.2E − 01
3.2E − 02
Table 3
Countries with the lowest consumption-based characterization factors and their top-inducing countries in 2015. C-based: Consumption-based; M-based: Mine production-based
Country
C-based CF [2015$/kg]
Top inducing country (share)
M-based CF [2015$/kg]
Weighted average of CF [2015$/kg]
Ag
Australia
1.5E + 02
Australia (66%)
8.0E + 01
2.8E + 02
Al
Suriname
7.2E − 03
Suriname (98%)
4.3E − 05
6.8E − 01
Au
Philippines
1.6E + 04
Australia (71%)
1.4E + 04
2.2E + 04
B
Chile
2.6E − 02
Chile (84%)
1.6E − 02
4.9E − 02
Ba
Turkmenistan
6.2E − 03
Kazakhstan (89%)
7.8E − 05
1.7E − 01
Co
British Virgin Islands
5.4E + 00
Philippines (56%)
5.2E + 00
7.8E + 00
Cr
Kazakhstan
3.7E + 00
Kazakhstan (91%)
3.3E + 00
7.3E + 00
Cu
North Korea
1.6E + 00
Peru (66%)
1.4E + 00
2.0E + 00
F
Botswana
1.3E − 02
South Africa (93%)
4.2E − 05
2.1E − 01
Fe
Tajikistan
9.5E − 03
Russia (65%)
1.6E − 04
4.3E − 02
Li
Chile
4.2E − 03
Chile (97%)
2.1E − 08
8.6E − 01
Mg
Greece
2.0E − 02
Greece (87%)
1.0E − 08
2.7E − 01
Mn
Brazil
2.3E − 01
Brazil (87%)
1.1E − 01
1.0E + 00
Mo
China
1.5E + 00
China (82%)
7.5E − 01
3.1E + 00
Nb
Brazil
4.4E + 00
Brazil (99%)
4.4E + 00
5.2E + 00
Ni
Papua New Guinea
2.2E + 00
Australia (62%)
9.4E − 01
5.2E + 00
P
Morocco
1.1E − 04
Morocco (99%)
5.3E − 24
2.2E − 02
Pb
Kazakhstan
8.9E − 01
Russia (67%)
8.7E − 01
1.3E + 00
Pd
Luxembourg
2.9E + 03
South Africa (80%)
5.1E + 00
8.7E + 03
Pt
Luxembourg
9.0E + 02
South Africa (95%)
7.7E + 00
4.8E + 03
Re
Romania
1.2E + 02
Armenia (69%)
8.9E − 01
5.4E + 02
Sb
Bolivia
1.6E + 00
Bolivia (84%)
8.1E − 01
5.8E + 00
Sn
Malaysia
8.3E + 00
Australia (38%)
4.5E + 00
1.5E + 01
Ta
Rwanda
2.8E − 03
Rwanda (97%)
2.8E − 03
5.1E − 03
Ti
Papua New Guinea
2.7E − 01
Australia (86%)
8.1E − 02
2.0E + 00
U
Australia
4.8E + 00
Australia (75%)
1.3E − 02
2.1E + 01
V
Kazakhstan
2.2E − 03
Russia (97%)
1.6E − 03
2.0E − 02
W
Kenya
5.7E + 00
UK (66%)
1.6E − 03
1.8E + 01
Zn
Uzbekistan
9.4E − 01
Kazakhstan (73%)
8.1E − 01
1.3E + 00
Oil
Libya
1.0E − 02
Libya (96%)
1.6E − 06
2.6E − 01
NG
Turkmenistan
6.7E − 03
Turkmenistan (95%)
1.1E − 03
2.0E − 01
Coal
Kazakhstan
2.0E − 03
Kazakhstan (91%)
1.3E − 05
3.2E − 02
Countries with higher consumption-based CFs typically rely on mine production in countries with higher mine production-based CFs, and vice versa (Eq. (10), Tables 2 and 3). Generally, consuming countries depend on mine production in multiple countries. Even in cases where consuming countries primarily depend on mine production in countries with significantly low mine production-based CFs, the resulting consumption-based CFs show relatively high values compared to the mine production-based CF of the mining countries, unless the number of inducing countries is limited to one. This is because mining countries with higher mine production-based CFs have a greater effect on the calculated consumption-based CFs, as indicated in Eq. (10). This trend is evident in the consumption-based CFs for specific cases such as Suriname for Al, Botswana for F, Chile for Li, Morocco for P, and Libya for oil (Table 3). Therefore, consumption-based CFs showed relatively small variations compared to mine production-based CFs, despite some countries having significantly low mine production-based CFs (Yokoi et al. 2024a) (Table S3 and Fig. S1).
To test the developed consumption-based CFs, we conducted a case study assessing the potential impacts of the abiotic resource use associated with the production of 1 kWh of lithium-ion batteries (LIBs) (Fig. 3). The case study analyzed three types of LIBs: LiNi0.6Mn0.2Co0.2O2-Graphite LIB (NMC), LiFePO4-Li4Ti5O12 LIB (LTO), and LiFePO4-Graphite LIB (LFP). The inventories of resource use were derived from a previous study (Yokoi et al. 2024b) (see Table S4), and the consumption-based CFs for each country were applied to calculate the induced impact (i.e., characterization result) for each country using the batteries. The results indicated significant variation in the induced impacts depending on the consuming country. Moreover, even within the same consuming country, the ranking of the NMC and LTO in terms of induced impacts could differ (the LFP showed the smallest induced impacts for all consuming countries). These differences were primarily driven by variations in the consumption-based CFs for Ni, Li, and Ti. Countries exhibiting distinctive results (i.e., the largest or smallest induced impacts as shown in Fig. 3) tend to rely heavily on a limited number of countries for abiotic resource mining. Countries that rely extensively on countries with relatively high mine production-based CFs for the mining of abiotic resources that significantly contribute to the results (e.g., Ni, Li, and Ti) tend to exhibit relatively high induced impacts, and vice versa. This finding highlights the potential uncertainties introduced by differences in consuming countries with distinct supply chains in LCIA studies, which cannot be captured by existing global-scale mine production-based characterization models for abiotic resources.
Fig. 3
Induced impacts of abiotic resource use associated with the production of 1 kWh of the three lithium-ion batteries. Violin plots illustrate the variation in induced impacts across consuming countries for each battery (a). Breakdown of abiotic resource contributions for the three countries with the largest induced impacts, the induced impacts calculated using the weighted average of CFs, and the three countries with the smallest induced impacts (b–d)
3.2 Consumption-based normalization values of abiotic resource use
The consumption-based NVs for a country are calculated as the sum of the products of consumption-based CFs and inducing mine production for all abiotic resources. The sum of the consumption-based NVs for all abiotic resources and all countries in 2015 (i.e., the global external cost of abiotic resource use) was 2.3 × 1012 2015$, accounting for approximately 3.1% of the world GDP (Fig. S2) (lists of mine production- and consumption-based NVs are provided in the Supplementary Excel file).
The USA had the highest consumption-based NVs, followed by China and Brazil (these countries are responsible for approximately 47.1% of the global NVs) (Fig. 4). The top two countries (i.e., the USA and China) were consistent across both mine production- and consumption-based NVs, whereas other major countries showed notable differences. Geographical distributions also differed between mine production- and consumption-based NVs, with consumption-based NVs exceeding mine production-based NVs in 82 of 189 countries (43%) (Figs. 5 and 6) (the results for the mineral resources are shown in Figs. S3 and S4). Countries such as Japan and Germany exemplify this trend, with their consumption-based NVs being approximately 116 and 41 times higher, respectively, than their mine production-based NVs. This disparity reflects their status as major resource consumers dependent on mining activities in foreign countries due to limited domestic primary resources (Figs. S5 and S6). The share of inducing impacts in-country and in foreign countries suggests that countries with larger consumption-based NVs than mine production-based NVs tend to induce higher impacts in foreign countries than their own countries (Figs. 6, S6, and S7).
Fig. 4
Share of abiotic resources in mine production- and consumption-based normalization values for countries covering the top 10 countries for each normalization value
Geographical distributions of consumption-based normalization values (C-based NVs) (a), mine production-based (M-based) NVs (b), C-based NVs per capita (c), and C-based NVs per GDP (d) for abiotic resource use in 2015. The values represent the sum of NVs for the target abiotic resources
Scatter plots of consumption- and mine production-based normalization values (NVs), population, and GDP for all target resources. The black lines on the population and GDP scatter plots represent the global averages, i.e., the ratios of the global NV and global population (or GDP)
A notable trend was observed in the distribution of per capita consumption-based NVs: developed countries exhibited relatively high values (Figs. 5 and 6), indicating their greater responsibility for the impacts induced in mining countries. Meanwhile, consumption-based NVs per GDP showed a different trend: developed countries generally had lower NVs per GDP. The consumption-based NVs per GDP represents the ratio of inducing external costs to the size of their economy. This observation also aligns with the relationship between consumption-based NVs and final demand for all sectors (Fig. 4). For countries with relatively high NVs per GDP, addressing the impacts of abiotic resource use may require strategies such as enhancing resource efficiency or changing sourcing countries to those with lower mine production-based CFs (Maeno et al. 2025; Schandl et al. 2018; Tukker 2015). These measures are essential to decouple economic growth from the inducing impacts of abiotic resource mining. However, it is important to consider that the early stages of economic growth generally require more resources (van Vuuren et al. 1999). Focusing on mineral resources, services and functions are mainly provided by material stocks rather than material flows (Pauliuk and Müller 2014). The relationship between inducing external costs and economic growth should be carefully discussed, for example, by considering service demand, stock productivity, and inequality in material stocks (Tanikawa et al. 2021; Watari and Yokoi 2021).
Focusing on a global scale, the NVs and contributions of each abiotic resource were consistent across mine production- and consumption-based NVs. Fossil fuels (oil, NG, and coal) were the primary contributors to the total NVs, accounting for approximately 81.9% of the total NVs (Fig. S2). The mineral resource with the largest contribution was Fe, followed by Cr and Au (Fig. S2). However, the primary abiotic resources differed by country for consumption-based NVs as well as mine production-based NVs (Fig. 4). For example, in the USA—the country with the highest consumption-based NVs—oil and NG were the main abiotic resources. In contrast, in China, the second-largest contributor to consumption-based NVs, oil and coal dominated. Focusing on mineral resources, consumption-based NVs of some countries were attributed to limited resources. For example, Turkey and South Africa had a dominant share of Cr in their consumption-based NVs. These countries induced large amounts of Cr mining, resulting in high inducing impacts associated with Cr mining (Figs. S8, S9 and S10). The reason for this is considered to be the large amount of mine production of Cr in these countries. We disaggregated the mining sector in the Eora but did not disaggregate other metal-related sectors, including metal production sectors, which could lead to a larger inducing mine production in countries with large mine production. Similar trends were observed for other metals, including B in Turkey, Nb in Brazil, and U in Canada (Fig. S8). This is the main limitation of this study and a key area for future research. High inducing impacts were mainly attributed to large inducing mine production, though some countries showed high inducing impacts due to high consumption-based CFs, such as B in Argentina, Li in Taiwan, and Mg in Austria (Fig. S10). These countries have the potential to reduce their inducing impacts by changing the sourcing countries of these resources to countries with lower mine production-based CFs.
3.3 Comparison with a previous study
A comparison with a previous study (De Laurentiis et al. 2023) is presented in Table 4. The study by De Laurentiis et al. (2023) calculated per capita mine production-based NV for the EU (EU-P) and per capita consumption-based NVs for the EU using process- and IO-based approaches (EU-C-p and EU-C-i/o) by applying the abiotic depletion potential (ADP) for CFs in the reference year 2010. To ensure consistency with De Laurentiis et al. (2023), this study also calculated per capita consumption-based NVs for EU countries using ADP (Table 4). Therefore, the differences between De Laurentiis et al. (2023) and this study are the inducing mine production, the coverage of abiotic resources, and the reference year. Notably, the per capita consumption-based NV for fossil fuels in this study aligned closely with the EU-C-i/o results. However, the per capita consumption-based NV for mineral resources was approximately one-fifth of the EU-C-i/o. This discrepancy is mainly attributed to methodological differences: De Laurentiis et al. (2023) aggregated several abiotic resources, such as “other non-ferrous metal ores” and “other minerals,” into the category “other industrial minerals,” and applied a weighted average ADP for these resources. This aggregation likely overestimated impacts of this impact category. The findings of Castellani et al. (2019) and De Laurentiis et al. (2023) identified “other industrial minerals,” as the largest contributor to EU-C-i/o followed by Au and Sn. However, the consumption-based NVs calculated using ADP in this study showed different results: the main abiotic resources were Au, Sb, and Ag (Fig. S11). These results suggest the significance of differentiating abiotic resources to calculate consumption-based NVs. Additionally, they highlight the methodological and data aggregation challenges that can influence results in LCA studies.
Table 4
Comparison with normalization values (NVs) of previous studies for EU-28
ADP, abiotic depletion potential; EU-P, European production-based NVs; EU-C-p, European consumption-based NVs using a process-based approach; EU-C-i/o, European consumption-based NVs using an input–output-based approach. EU-28 was considered to reflect the composition of the EU in the reference year (2010) of a previous study
4 Discussion
This study calculated consumption-based CFs and NVs for 29 mineral resources and three fossil fuels across 189 countries and regions. Consumption-based CFs represent the induced impacts of abiotic resource mining by reflecting the average supply chains of consuming countries. These CFs enable LCA practitioners to conduct regionally explicit assessments without identifying the mining countries for each abiotic resource used in a target product system. Our findings reveal significant variations in consumption-based CFs for some abiotic resources, primarily attributed to variations in mine production-based CFs. To ensure sustainable resource use, countries with higher consumption-based CFs can change their sourcing countries to those with lower mine production-based CFs or take measures to reduce the potential impact (i.e., mine production-based CF) in their sourcing countries. Additionally, consumption-based CFs can also be utilized to calculate and update consumption-based NVs by multiplying with inducing mine production.
However, it is important to note that the consumption-based CFs proposed in this study assume the average supply chains of the final demands of countries, which may differ from the actual supply chain of a specific product. When LCA practitioners can trace the supply chain and identify the sourcing countries of the abiotic resources used for a target product, applying mine production-based CFs to these identified mining flows is recommended. Conversely, consumption-based CFs are most effective when supply chain data is unavailable or when the focus of the assessment is on larger scales, including companies, organizations, and countries. Moreover, the mining countries induced by final demand could differ from those involved in sourcing raw materials for product manufacturing. Developing alternative CFs that reflect the average trade structure of raw materials could better capture raw material procurement structure in product manufacturing.
The country-level consumption-based NVs provided in this study can be applied to LCA studies focusing on a specific country. The results showed that some countries (e.g., Japan and Germany) have significantly higher consumption-based NVs than mine production-based NVs. This discrepancy highlights the limitations of applying mine production-based NVs, which may result in significantly different normalization results for abiotic resource use when applied to these countries, leading to a conclusion that does not represent the actual situation for abiotic resource use. Since abiotic resources have the unique characteristics of limited mining countries, consumption-based NVs for abiotic resource use are especially important compared to other impact categories. Furthermore, this study covered and differentiated a wide range of abiotic resources in calculating induced mine production. For LCA practitioners considering a limited number of abiotic resources, using NVs that account for a broader range of abiotic resources can lead to smaller normalization results, which in turn can lead to underestimation of the potential impacts of abiotic resource use. By providing resource-specific NVs, our study enables LCA practitioners to select the target abiotic resources for calculating NVs so as to be consistent with the scope (i.e., coverage of resources) of their LCA studies and LCI.
Despite these contributions, the main limitation of this study for consumption-based NVs is the disaggregation of the metal production-related sectors as well as the mining sectors in MRIO tables. Because IO-based analysis faces the challenge that mining of different abiotic resources is integrated into a single sector. To address this, as a first step, we disaggregated the mining sector in the Eora MRIO database using country-level mine production data and trade statistics. Connecting the disaggregated mining sectors with trade statistics allowed us to accurately assess the proportion of resources refined domestically versus those exported and refined abroad. However, the level of detail for refinery sectors varies by country in the Eora MRIO database. In cases where there is only a single refinery sector, different refined metals will flow in the identical proportion into the subsequent sectors. Therefore, to fully consider the supply chains of resource-consuming countries, future research should disaggregate metal production and metal-specific product sectors by integrating data on smelting, refining, manufacturing, and trade. Additionally, addressing the issue of by-products in the disaggregation of mining sectors remains a task for future research. When the original Eora MRIO database contains mining sectors for particular ores, we assigned corresponding mine production for the ores to these sectors and then disaggregated other mining sectors, resulting in 51 mining sectors. However, in some cases, the mining activities for ores produce multiple commodities, which are not fully accounted for in this model (USGS 2017b). The failure to account for by-product effects in this model may introduce uncertainty in tracking the supply chain of commodity production, although the total amount of mine production is preserved because the mine production was allocated based on induced output (Eq. (9)). To consider the effect of by-products in the disaggregation of mining sectors, integrating detailed relationships between ores and minerals into the MRIO framework utilizing ore-to-mineral ratios would be an effective approach (Berthet et al. 2024). Furthermore, the selection of the MRIO database influences the results. While this study used the Eora MRIO database, alternative MRIO databases such as EXIOBASE (Stadler et al. 2018), GLORIA (Lenzen et al. 2022), and WIOD (Dietzenbacher et al. 2013) are also available. Comparing results across these MRIO databases would help discuss the effects of MRIO database choice on the calculated consumption-based CFs and NVs, an essential direction for future research (Berthet et al. 2024). Another avenue for future research is to update consumption-based CFs and NVs over time, enabling analysis of historical trends in potential impacts of abiotic resource use.
Given the uncertainties mentioned above, validating and adjusting the estimation results is desirable. One possible approach is to calculate the consumption-based NV per GDP for each country to assess the plausibility of the results. The top 20 countries in terms of consumption-based NV per GDP across all target resources are presented in Table S5. While consumption-based NV is expressed as an external cost and is therefore not directly comparable to GDP, it is still possible to identify countries with potentially overestimated values by comparing their ratios to the global average. The global average NV per GDP is approximately 0.03, whereas Guyana shows a value of around 2.97, indicating an overestimation. Countries such as Gabon and Oman, which also rank high in Table S5, have values below 1 but still appear to be significantly overestimated. Further investigation of the results for these countries and exploration of potential model improvements represent important directions for future research. Alternatively, a reasonable approach may be to adjust the induced mine production and consumption-based NVs so that each country’s consumption-based NV per GDP falls within an acceptable range (e.g., within one standard deviation of the global average), without changing the country-level breakdown of induced mining (i.e., consumption-based CFs). This could help ensure more reasonable and consistent results.
In this study, we adopted a user cost model as the characterization model, which represents the external costs of abiotic resource mining. As a result, the consumption-based NVs calculated in this paper can be used to analyze the responsibility of consuming countries for the external costs associated with mine production. This information can support countries in managing their supply chains to ensure sustainable resource use (Yokoi et al. 2021). Furthermore, our results for inducing mine production for each consuming country can be incorporated with data on various environmental and social impact intensities associated with mining activities. This integration enables analysis of the footprint and responsibility of consuming countries for various issues, such as greenhouse gas emissions, water consumption, land use, and child labor (Islam and Murakami 2020; Islam et al. 2022; Jenkins and Yakovleva 2006; Nakajima et al. 2017; Norgate and Haque 2010; Northey et al. 2016; Sun et al. 2025; Wiedmann and Lenzen 2018). Analysis of the nexus between these impacts and consuming countries can contribute to research on sustainable supply chain management, the responsibility of consumers, and environmental justice (Mohai et al. 2009). Furthermore, inducing mine production reflects the dependency on mining activities in each mining country. Therefore, our results can be applied to assess the supply risks of abiotic resources and can contribute to country-level criticality assessments (Arendt et al. 2020; Bach et al. 2017; Koyamparambath et al. 2024; Marinova et al. 2023).
Acknowledgements
This research was partly supported by Grant-in-Aid for Scientific Research (A) (JSPS KAKENHI 21H04944) and Scientific Research (B) (JSPS KAKENHI 23 K28302).
Declarations
Conflict of interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Potential impact of abiotic resource use considering country-specific supply chain: consumption-based characterization and normalization utilizing a multi-regional input–output model
Bach V, Finogenova N, Berger M, Winter L, Finkbeiner M (2017) Enhancing the assessment of critical resource use at the country level with the SCARCE method – case study of Germany. Resour Policy 53:283–299. https://doi.org/10.1016/j.resourpol.2017.07.003CrossRef
Berger M, Sonderegger T, Alvarenga R, Bach V, Cimprich A, Dewulf J, Frischknecht R, Guinée JB, Helbig C, Huppertz T, Jolliet O, Motoshita M, Northey S, Peña CA, Rugani B, Sahnoune A, Schrijvers D, Schulze R, Sonnemann G, Valero A, Weidema BP, Young SB (2020) Mineral resources in life cycle impact assessment: part II – recommendations on application-dependent use of existing methods and on future method development needs. Int J Life Cycle Assess 25:798–813. https://doi.org/10.1007/s11367-020-01737-5CrossRef
Berthet E, Lavalley J, Anquetil-Deck C, Ballesteros F, Stadler K, Soytas U, Hauschild M, Laurent A (2024) Assessing the social and environmental impacts of critical mineral supply chains for the energy transition in Europe. Glob Environ Change 86. https://doi.org/10.1016/j.gloenvcha.2024.102841CrossRef
Beylot A, Corrado S, Sala S (2020) Environmental impacts of European trade: interpreting results of process-based LCA and environmentally extended input–output analysis towards hotspot identification. Int J Life Cycle Assess 25:2432–2450. https://doi.org/10.1007/s11367-019-01649-zCrossRef
BGS (2021) World mineral production 2015–2019. Nottingham, UK, British Geological Survey, pp 96
Breedveld L, Lafleur M, Blonk H (1999) A framework for actualising normalisation data in LCA: Experiences in the Netherlands. Int J Life Cycle Assess 4:213–220. https://doi.org/10.1007/BF02979500CrossRef
Castellani V, Beylot A, Sala S (2019) Environmental impacts of household consumption in Europe: comparing process-based LCA and environmentally extended input-output analysis. J Clean Prod 240. https://doi.org/10.1016/j.jclepro.2019.117966CrossRef
Corrado S, Rydberg T, Oliveira F, Cerutti A, Sala S (2020) Out of sight out of mind? A life cycle-based environmental assessment of goods traded by the European Union. J Clean Prod 246. https://doi.org/10.1016/j.jclepro.2019.118954CrossRef
Crenna E, Secchi M, Benini L, Sala S (2019) Global environmental impacts: data sources and methodological choices for calculating normalization factors for LCA. Int J Life Cycle Assess 24:1851–1877. https://doi.org/10.1007/s11367-019-01604-yCrossRef
Dahlbo H, Koskela S, Pihkola H, Nors M, Federley M, Seppälä J (2013) Comparison of different normalised LCIA results and their feasibility in communication. Int J Life Cycle Assess 18:850–860. https://doi.org/10.1007/s11367-012-0498-4CrossRef
De Laurentiis V, Amadei A, Sanyé-Mengual E, Sala S (2023) Exploring alternative normalization approaches for life cycle assessment. Int J Life Cycle Assess 28:1382–1399. https://doi.org/10.1007/s11367-023-02188-4CrossRef
El Serafy S (1989) The proper calculation of income from depletable natural resources. In: Ahmad YJ, El Serafy S, Lutz E (eds) Environmental accounting for sustainable development. The international bank for reconstruction and development, The World Bank, Washington, D.C., pp 10–18
Koyamparambath A, Loubet P, Young SB, Sonnemann G (2024) Spatially and temporally differentiated characterization factors for supply risk of abiotic resources in life cycle assessment. Resour Conserv Recycl 209. https://doi.org/10.1016/j.resconrec.2024.107801CrossRef
Laurent A, Hauschild MZ (2015) Normalisation. In: Hauschild M, Huijbregts MAJ (eds) Life cycle impact assessment. Springer Dordrecht, pp 271–300
Lenzen M, Geschke A, West J, Fry J, Malik A, Giljum S, Milà i Canals L, Piñero P, Lutter S, Wiedmann T, Li M, Sevenster M, Potočnik J, Teixeira I, van Voore M, Nansai K, Schandl H (2022) Implementing the material footprint to measure progress towards Sustainable Development Goals 8 and 12. Nat Sustain 5:157–166. https://doi.org/10.1038/s41893-021-00811-6CrossRef
Lenzen M, Moran D, Kanemoto K, Foran B, Lobefaro L, Geschke A (2012) International trade drives biodiversity threats in developing nations. Nature 486:109–112. https://doi.org/10.1038/nature11145CrossRef
Li X, Qian Y, Xie M, Liu D, Qiao Q (2024) An endpoint model for life cycle impact assessment in China and preliminary normalization values: a case study of vehicles. J Clean Prod 434. https://doi.org/10.1016/j.jclepro.2023.140326CrossRef
Maeno K, Tokito S, Yokoi R, Kagawa S (2025) Global supply chain restructuring towards achieving a low-carbon procurement of mineral resources for metal production. Resour Environ Sustain 20. https://doi.org/10.1016/j.resenv.2025.100215CrossRef
Marinova S, Bach V, Yokoi R, Motoshita M, Islam K, Finkbeiner M (2023) Country-level criticality assessment of abiotic resource use in Japan - application of the SCARCE method. J Clean Prod 412. https://doi.org/10.1016/j.jclepro.2023.137355CrossRef
Mutel C, Liao X, Patouillard L, Bare J, Fantke P, Frischknecht R, Hauschild M, Jolliet O, Maia de Souza D, Laurent A, Pfister S, Verones F (2019) Overview and recommendations for regionalized life cycle impact assessment. Int J Life Cycle Assess 24:856–865. https://doi.org/10.1007/s11367-018-1539-4CrossRef
Nakajima K, Noda S, Nansai K, Matsubae K, Takayanagi W, Tomita M (2019) Global distribution of used and unused extracted materials induced by consumption of iron, copper, and nickel. Environ Sci Technol 53:1555–1563. https://doi.org/10.1021/acs.est.8b04575CrossRef
Nansai K, Nakajima K, Kagawa S, Kondo Y, Shigetomi Y, Suh S (2015) Global mining risk footprint of critical metals necessary for low-carbon technologies: the case of neodymium, cobalt, and platinum in Japan. Environ Sci Technol 49:2022–2031. https://doi.org/10.1021/es504255rCrossRef
NEA (2016) Uranium 2016: Resources, production and demand. OECD Publishing, Paris
Pfister S, Oberschelp C, Sonderegger T (2020) Regionalized LCA in practice: the need for a universal shapefile to match LCI and LCIA. Int J Life Cycle Assess 25:1867–1871. https://doi.org/10.1007/s11367-020-01816-7CrossRef
Pizzol M, Laurent A, Sala S, Weidema B, Verones F, Koffler C (2017) Normalisation and weighting in life cycle assessment: quo vadis? Int J Life Cycle Assess 22:853–866. https://doi.org/10.1007/s11367-016-1199-1CrossRef
Potting J, Hauschild M (2006) Spatial differentiation in life cycle impact assessment: a decade of method development to increase the environmental realism of LCIA. Int J Life Cycle Assess 11:11–13. https://doi.org/10.1065/lca2006.04.005CrossRef
Sala S, Benini L, Beylot A, Castellani V, Cerutti A, Corrado S, Crenna E, Diaconu E, Sanyé-Mengual E, Secchi M, Sinkko T, Pant R (2019) Consumption and consumer footprint: methodology and results. EUR 29441 EN, Publications Office of the European Union, Luxembourg, ISBN 978–92–79–97256–0, JRC113607. https://doi.org/10.2760/98570
Sanyé-Mengual E, Sala S (2023) Consumption footprint and domestic footprint: assessing the environmental impacts of EU consumption and production. EUR 31390 EN, Publications Office of the European Union, Luxembourg, ISBN 978–92–76–61754–9, JRC128571. https://doi.org/10.2760/218540
Schandl H, Fischer-Kowalski M, West J, Giljum S, Dittrich M, Eisenmenger N, Geschke A, Lieber M, Wieland H, Schaffartzik A, Krausmann F, Gierlinger S, Hosking K, Lenzen M, Tanikawa H, Miatto A, Fishman T (2018) Global material flows and resource productivity: forty years of evidence. J Ind Ecol 22:827–838. https://doi.org/10.1111/jiec.12626CrossRef
Sonderegger T, Berger M, Alvarenga R, Bach V, Cimprich A, Dewulf J, Frischknecht R, Guinée JB, Helbig C, Huppertz T, Jolliet O, Motoshita M, Northey S, Rugani B, Schrijvers D, Schulze R, Sonnemann G, Valero A, Weidema BP, Young SB (2020) Mineral resources in life cycle impact assessment – part I: a critical review of existing methods. Int J Life Cycle Assess 25:784–797. https://doi.org/10.1007/s11367-020-01736-6CrossRef
Stadler K, Wood R, Bulavskaya T, Södersten CJ, Simas M, Schmidt S, Usubiaga A, Acosta-Fernández J, Kuenen J, Bruckner M, Giljum S, Lutter S, Merciai S, Schmidt JH, Theurl MC, Plutzar C, Kastner T, Eisenmenger N, Erb KH, de Koning A, Tukker A (2018) EXIOBASE 3: developing a time series of detailed environmentally extended multi-regional input-output tables. J Ind Ecol 22:502–515. https://doi.org/10.1111/jiec.12715CrossRef
Sun X, Giljum S, Maus V, Schomberg A, Zhang S, You F (2025) Robust assessments of lithium mining impacts embodied in global supply chain require spatially explicit analyses. Environ Sci Technol 59:7081–7094. https://doi.org/10.1021/acs.est.4c12749
Tanikawa H, Fishman T, Hashimoto S, Daigo I, Oguchi M, Miatto A, Takagi S, Yamashita N, Schandl H (2021) A framework of indicators for associating material stocks and flows to service provisioning: application for Japan 1990–2015. J Clean Prod 285. https://doi.org/10.1016/j.jclepro.2020.125450CrossRef
USGS (2017b) Critical mineral resources of the United States—economic and environmental geology and prospects for future supply. Professional Paper 1802. https://doi.org/10.3133/pp1802
Wenz L, Willner SN, Radebach A, Bierkandt R, Steckel JC, Levermann A (2015) Regional and sectoral disaggregation of multi-regional input-output tables – a flexible algorithm. Econ Syst Res 27:194–212. https://doi.org/10.1080/09535314.2014.987731CrossRef
Wieland H, Lenzen M, Geschke A, Fry J, Wiedenhofer D, Eisenmenger N, Schenk J, Giljum S (2022) The PIOLab: building global physical input–output tables in a virtual laboratory. J Ind Ecol 26:683–703. https://doi.org/10.1111/jiec.13215CrossRef
Yokoi R, Kataoka R, Masese T, Bach V, Finkbeiner M, Weil M, Baumann M, Motoshita M (2024) Potentials and hotspots of post-lithium-ion batteries: environmental impacts and supply risks for sodium- and potassium-ion batteries. Resour Conserv Recycl 204. https://doi.org/10.1016/j.resconrec.2024.107526CrossRef
Yokoi R, Motoshita M, Matsuda T, Itsubo N (2024) Country-specific external costs of abiotic resource use based on user cost model in life cycle impact assessment. Environ Sci Technol 58:7849–7859. https://doi.org/10.1021/acs.est.4c00100CrossRef
Yokoi R, Nansai K, Nakajima K, Watari T, Motoshita M (2021) Responsibility of consumers for mining capacity: decomposition analysis of scarcity-weighted metal footprints in the case of Japan. iScience 24:102025. https://doi.org/10.1016/j.isci.2020.102025
