Assessing the Economic Structure, Climate Change and Decarbonisation in Europe

Anthropogenic greenhouse gas (GHG) emissions coming mainly from fossil fuel combustion for energy use are causing air temperature increases resulting in climate change. This study employs an environmentally extended input–output model to conduct an economy-wide assessment of GHG emissions in the European Union (EU). Model results indicate that the assumed growth of economic activity by 2030 will lead to a large increase in GHG emissions by 89%, assuming no technological change and no additional policy mitigation efforts. The electricity sector and agriculture create the highest direct and indirect GHG emissions per unit of economic output across the 27 EU member states (EU-27); for every 1-million-euro-increase in the final demand for the products and services of the electricity sector and agriculture, 2198 and 1410 additional tons of GHG emit, respectively. Regional climate projections under a low-decarbonisation pathway (RCP8.5), in accordance with our economic analysis, indicate a further increase of regional warming, combined with pronounced changes in the hydrological cycle. Contrariwise, following a strong mitigation pathway (RCP2.6) will result in warming levels lower than 1.5 °C with respect to the 1986–2005 reference period. Our findings reveal the importance of both direct and indirect contribution of economic sectors in the generation of GHG emissions, taking into consideration the size of the sectors and the assumed growth rates. The design and implementation of sectoral emission reduction policies from the perspective of the whole production supply chain can effectively contribute to GHG emission reduction commitments.


Introduction
Economic and population growth have contributed to increasing demand for resources, including energy. Anthropogenic greenhouse gas (GHG) emissions, coming mainly from fossil fuel combustion for energy production (mainly coal and crude oil), are causing air temperature increases resulting in climate change (IPCC 2013). According to the Intergovernmental Panel on Climate Change (IPCC 2018), human-induced global warming, reached approximately 1.0 °C above pre-industrial levels in 2017 (with a likely range of 0.8 °C and 1.2 °C). In the absence of ambitious mitigation policies in the coming years that could lead to a sharp decline in GHG emissions by 2030, global warming is likely to surpass 1.5 °C sometime between 2030 and 2045 (IPCC 2018).
The Paris Agreement, achieved at the 21st United Nations Climate Change Conference (COP21) in Paris in 2015, aims to substantially reduce global GHG emissions and limit the temperature increase by the end of this century to well below 2 °C above the pre-industrial levels, while pursuing efforts to limit the global warming to less than 1.5 °C (UNFCCC 2021). In the European Union (EU), the Governance of the Energy Union and Climate Action Regulation sets out the legislative framework for achieving the 2030 energy and climate targets in line with the Paris Agreement (Reg. (EU) 2018(EU) /1999(EU) 2018. Member states are obliged to prepare and submit integrated National Energy and Climate Plans (NECPs). These plans aim to support the binding objective for an overall economy-wide reduction of at least 40% of 1 3 Published in partnership with CECCR at King Abdulaziz University GHG levels by 2030compared to 1990(European Council 2014. The European Commission has recently proposed an EU-wide net GHG emissions reduction target of at least 55% by 2030, compared to 1990 levels, to put the EU on a balanced pathway to reach climate neutrality by 2050 (European Commission 2020a). The EU Green Deal, the major initiative to move to a net-zero GHG emissions economy by 2050, aims to achieve decarbonisation in the EU within the coming years (European Commission 2020b).
Across the 27 EU countries , there is a general downward trend of GHG emissions during the last 3 decades. In particular, emissions in the EU-27 declined by 16% between 1995 and 2018, while the gross domestic product (GDP) of the EU-27 economy increased by 48% (Fig. 1-left  panel). This figure shows that the EU is on track to fulfil its commitments to meet the 2020 climate and energy package targets (European Commission 2009), i.e. to meet its 20% GHG emissions reduction target for 2020 compared to 1990 levels. Within the same period, the GHG emission intensity of the EU-27 economy has decreased by 43%, highlighting the efforts of the EU member states to decarbonise their economies ( Fig. 1-right panel) (Eurostat 2021b;EEA 2021). The contribution of the individual production sectors to the GHG emission generation substantially varies in size and from year to year. The largest emitting sectors in the EU, that is, the electricity production and the manufacturing sectors, decreased their GHG emissions between 2008 and 2019 by 31% and 22%, respectively (Eurostat 2021a). On the other hand, the agriculture-related emissions (about 15% of the total GHG emissions) have decreased by only 3%, while in other sectors, such as air transport and health services, GHG emissions have increased by 19% and 10%, respectively.
The relationship between economic activity and GHG emissions has been widely studied. A large part of the literature focuses on investigating the relationship between economic growth, energy use and GHG emissions, testing the validity of the Environmental Kuznets Curve hypothesis that postulates an inverted U-shaped relationship between environmental degradation and per capita income (Acaravci and Ozturk 2010; Lee et al. 2015;Manta et al. 2020). Bottom-up studies examining alternative approaches for reducing GHG emissions, including specific technological and managerial options per sector (Neuhoff 2005;Worrell et al. 2001), typically provide detailed information about emission reduction potential within certain sectors, but they do not capture systemic effects (Mundaca et al. 2019). Less effort has been devoted in exploring the relationship between economic structure, climate change and GHG emissions. Capturing the sectoral-level emission effects is critical to identify the key emitters and so contribute to the overall reduction of GHG emissions.
A clear understanding of any economic structure requires a sectoral categorization and an analysis of the inter-industry commodity flows (Ghosh and Roy 1998). Input-output analysis (IOA) has been considered a valid approach for studying such interdependencies between the production sectors in an economy (Miller and Blair 2009), and it has been extensively applied for policy impact analysis, structural and technical change analysis (Giannakis and Bruggeman 2017;Taliotis et al. 2020). The environmental extension of IOA models has consequently become a valuable technique for analysing interdependency among industries and GHG emissions (Hawkins et al. 2015). Tracing the GHG emissions embodied in the flows of intermediates along production supply chains can provide a more thorough understanding of how GHG are driven through economic activity (Peters 2008) and an integrated assessment of the contribution of the production sectors in the generation of GHG emissions (Suh 2006).
Computable General Equilibrium (CGE) models have also been used in the literature to study the interrelationships of production sectors and their impact on climate change (Eboli et al 2010). CGE models, unlike IO models, capture supplyside effects and allow for more flexibility, due to their non-linearity, regarding substitution effects and relative price changes Published in partnership with CECCR at King Abdulaziz University (Koks et al. 2016). Optimization models, such as the TIMES (The Integrated MARKAL-EFOM System) and the MES-SAGE (Model for Energy Supply Strategy Alternatives and their General Environmental Impacts), and accounting models, such as the LEAP (Long-range Energy Alternatives Planning system) and the GACMO (Greenhouse Gas Abatement Cost Model) have also been used for the development of the Nationally Determined Contributions (NDCs) worldwide (Haydock and McCullough 2017).
Given the highly uneven sectoral annual GHG emission generation and reduction potentials (Blok et al. 2020), an economywide assessment of the contribution of production sectors in the generation of GHG emissions is of major relevance. This paper aims to assess the direct and indirect contribution of economic sectors to GHG emissions, accounting for all monetary inter-sectoral transactions. Specifically, an environmentally extended input-output (EE-IOA) model is developed to explore the economy-wide effects of the growth of the production sectors on the GHG emissions generated in the EU-27. The paper is structured as follows: Sect. 2 outlines the methodology of the study, while Sect. 3 presents the results of the analysis. Section 4 presents complementary information on regional climate change projections for Europe under two concentration pathways for the future. The paper ends with a discussion of the results and conclusions drawn from the analysis.

Environmentally Extended Input-Output Analysis
The basic structure of an IOA model consists of a system of linear equations that account for the way in which the output of each sector i is distributed through sales to other sectors for intermediate use and final demand as follows (Miller and Blair 2009): where x i is the total output of sector i (i = 1, … , n) ; x ij describes the inter-industry sales of sector i to all sectors j(j = 1, … , n) ; y i is the final demand for sector i ′ s product.
The technical coefficients a ij represent the value of the output from sector i that is required to produce one unit of output in sector j as follows: Equation (1) can then be rewritten in matrix notation as follows: (1) (3) where A is the technical coefficients matrix. Solving Eq. (3) for X , we obtain Eq. (4): where (I − A) −1 = L is the Leontief inverse or the total (direct and indirect) requirements matrix. The L matrix quantifies the direct and indirect impacts exerted by changes in final demand (ΔY) on the output of each sector (ΔX) . The output multiplier for an individual sector j is defined as the column sums of the L matrix � ∑ n i=1 l ij � . The environmental extension of the basic IOA model can be obtained by introducing an exogenous vector of emission intensity, here denoted as D = d i , that is, the amount of GHG emissions g i per unit output of each sector i x i as follows (Camanzi et al. 2017;Giannakis et al. 2019): The total (direct and indirect) GHG emissions (G) can be calculated as follows:

Data
The EU-27 symmetric IOA table for the year 2019 was derived from the Eurostat database (Eurostat 2021c). The initial scheme of 65 sectors of economic activity (Appendix Table 2) was aggregated into 25 economic sectors (Appendix Table 3). The sectoral GHG emission data were obtained by the Eurostat (2021a). Sources of uncertainty typically associated with EE-IOA include: (a) the assumption of constant returns to scale, (b) the assumption of fixed input structure, (c) the assumption of homogenous prices, (d) the assumption that the imported goods are produced with the same domestic technology, (e) uncertainties in source data (Wiedmann 2009).
In this study, we used the long-term outlook of the EU Reference Scenario 2016, that is, one of the key modelling tools of the European Commission for projecting economic activity, to estimate sectoral growth rates in the EU (Capros et al. 2016). Specifically, we explore the direct and indirect impact of the increase in the final demand for the output of the individual production sectors in the GHG emissions in the EU-27 from 2019 to 2030.
Considering the rapid warming over the region and in addition to the input-output analysis presented in the previous sections, we analysed state-of-the-art regional climate projections for the twenty-first century to assess the importance of adopting timely mitigation measures and reducing GHG emissions and concentrations. Following the temporal horizon of the IOA, we focus on the near-term changes (2021-2040), while for comparison, we also present mid-twenty-first-century (2041-2060) projections for temperature and precipitation. These projections are based on a large ensemble of high-resolution regional simulations (0.11 × 0.11°) that is part of the EURO-CORDEX initiative (Jacob et al. 2020). Our model selection is similar to Cherif et al. (2020) and is presented in Table 4 (Appendix). Our analysis is complementary to Jacob et al. (2014) and is an update in terms of future scenarios and ensemble size.
In the present study, we have assessed two future scenarios. The first Representative Concentration Pathway (RCP), RCP2.6 (Van Vuuren et al. 2011), is representative of scenarios leading to low GHG concentration levels. It is a socalled 'peak' scenario, in the sense that its radiative forcing first reaches a value around 3.1 W/m 2 (mid-century), returning to 2.6 W/m 2 by 2100. To reach such radiative forcing levels, GHG emissions and concentrations have to be reduced substantially over time. According to the latest Assessment Report of IPCC (2013), global warming under RCP2.6 stays below 2 °C above 1850-1900 levels throughout the twentyfirst century, clearly demonstrating the potential of mitigation policies. This pathway closely meets the main targets of the Paris Agreement, aiming at keeping global warming less than 2 °C above pre-industrial levels. The second pathway, RCP8.5 (Riahi et al. 2007), is representative of scenarios that lead to high GHG concentration levels. It is considered as a high-emission pathway, which, however, at least for parts of the region, is following the observed temperature trends (Zittis et al. 2019). Notably, the two pathways under investigation start to deviate significantly after the 2030s.

Results
The IOA multiplier analysis identified the most important sectors of economic activity in view of their capacity to generate economic output throughout the EU economy ( Table 1). The food industry has the highest output multiplier (O-M: 2.42), that is, for every 1-million-euro-increase in the final demand for the industry's products, the total output of the economy will increase by 2.42 million euro. Similarly, the machinery and equipment sector (O-M: 2.17) also creates strong multiplier effects in the EU-27 economy, considering its high contribution to total economic output formation, that is, 10% (Fig. 2). On the other hand, service sectors such as real estate (O-M: 1.44) and education (O-M: The EE-IOA multiplier analysis identified the contribution of the twenty-five economic sectors in the generation of GHG emissions (Table 1) In 2019, the EU-27's sectors emitted 3049 megatons of GHG (CO 2 equivalent). The assumed growth of EU economic sectors by 2030, assuming no change in technology a ij and no explicit policies and measures to reduce GHG emissions, results in a total (direct and indirect) increase of GHG emissions by 89% relative to 2019. Significant increases in the emissions of GHG are presented in the industrial sectors of machinery and equipment (332%) and chemical and plastic products (69%), which emit about 10.7% of the total GHG emissions in the EU-27. Several service sectors also exhibit large increases in their GHG emissions, such as trade (427%) and transportation warehousing services (259%); however, their share in the generation of GHG emissions is less than 3.7%. The relatively low increase of GHG emissions of agriculture (4%) and electricity (11%) sectors, i.e. the largest GHG emitters in absolute terms (Fig. 2), is mainly due to the assumed low growth rates by 2030.

Climate Change in Europe with and Without Decarbonisation
The regional implications of anthropogenic warming have been widely observed in Europe, and these are most profound over the last 3 decades (Kovats et al. 2014). These manifestations include an overall warming, faster than the global mean rates (Fig. 3). Particularly in the last four decades, the warming trend is 0.41 °C/decade, or 1.5 times higher than the global average (0.27 °C/decade). This regional temperature increase is mainly driven by the global concentration levels of greenhouse gases (mainly CO 2 ). These gases have a long lifetime and are, therefore, well-mixed in the atmosphere. Besides some decarbonization efforts in Europe, global GHG concentrations have increased substantially (Olivier and Peters 2020), contributing to the observed warming trend, in addition to further positive climatic feedbacks. Past precipitation changes are   Fig. 4. Under the strong mitigation pathway (RCP2.6), the projected changes for temperature will likely not exceed 1.5 °C (Fig. 4-left panels). This is the case for both time horizons under investigation. Exceptions are the northern parts of Scandinavia, where due to positive climate feedbacks, the projected warming is somehow higher (up to 3 °C in midcentury). In contrast, a high-emission pathway of low decarbonization rates (RCP8.5) indicates a much stronger European warming, particularly for the middle of the twenty-first century ( Fig. 4-right panels). For the hot-spot areas (for example, the Alps and Scandinavia), the temperature increase could reach levels of 4-5 °C. Such warming levels and reduced snow cover imply severe impacts in a range of critical socio-economic sectors, including tourism, agriculture, hydropower and more (Gobiet et al. 2014;Jacob et al. 2018). The snow-albedo positive feedback (Winter et al. 2017) is likely dominant in these areas since, in a warmer future, smaller areas are expected  : 1986-2005), based on the EURO-CORDEX ensemble of regional climate projections. Pathway RCP2.6 is presented in the left panels and pathway RCP8.5 in the right panels to be covered by snow during the winter and spring seasons. Since we have used a relatively recent reference period, about 0.6-0.7 °C should be added to these projections to approximate the regional warming levels with respect to the pre-industrial era. This comparison highlights the need for timely decarbonization at the European and global scales. For specific hot-spot areas (for example, Scandinavia, Iberia, East Europe, Anatolia, North Africa), the milder warming (about 0.5 °C less), projected under a more sustainable pathway (i.e. RCP2.6), is already evident in the next 2 decades.
The projected precipitation changes for southern and northern Europe are mainly attributed to changes in atmospheric  : 1986-2005), based on the EURO-CORDEX ensemble of regional climate projections. Pathway RCP2.6 is presented in the left panels and pathway RCP8.5 in the right panels circulation and thermodynamics. Under rising greenhouse gas concentrations, climate models project that the Hadley Cell circulation will change, the tropics will expand, and the mid-latitude westerlies and associated storm tracks will likely shift poleward (Cherif et al. 2020). This is expected to enhance subsidence and reduce storminess at the latitudes of southern Europe and the Mediterranean region, with a resulting reduction in precipitation (Fig. 5). This poleward shift of storm tracks, in addition to changes in thermodynamics, can explain the projected precipitation increase in northern Europe. For example, the Clausius-Clapeyron relationship predicts an increase in the water holding capacity of air (the saturation water vapor pressure) of approximately 7%/°C rise in temperature (Held and Soden 2006).

Discussion and Conclusions
In this study, we empirically analysed the relationship between the production activities of the EU-27's economic system and GHG emissions at the macroeconomic level. Our analysis shows the level of GHG emissions that would occur in the EU-27 by 2030 in the absence of any further policy mitigation effort and without any technological progress. In particular, the results of the EE-IOA model indicate that the assumed growth of economic activity will lead to a large increase in GHG emissions, that is, around 89%.
Our regional climate projections under a high-emission pathway of low decarbonisation (RCP8.5), in accordance with our economic analysis, indicate a considerable increase of European warming levels while for a strong mitigation pathway (RCP2.6), warming will not surpass 1.5 °C, with respect to our reference period (or about 2-2.5 °C since preindustrial levels). Near and mid-term climate projections for precipitation indicate changes of the range of ± 10% for both pathways. These ranges agree with previous assessments for the region (Gobiet et al. 2014;Jacob et al. 2018;Zittis et al. 2019;Coppola et al. 2021). Moreover, the identified hot-spot areas (for example, northeast Europe and parts of the Mediterranean) corroborate previous studies (Giorgi 2006;Giorgi and Lionello 2008). Both analyses highlight the crucial role of sectoral climate change mitigation policies and decarbonisation technologies to ameliorate the negative effects of economic growth in the generation of GHG emissions and to meet EU GHG emission reduction targets.
The results of the EE-IOA multiplier analysis revealed that the electricity and the agricultural sectors create the highest direct and indirect GHG emissions in the EU-27 per unit (million euro) of economic output produced. Electricity, agriculture and the industrial sectors of metal and non-metal products and chemical and plastic products are also the largest GHG emitters in absolute terms (tons). Our analysis reveals the importance of both direct and indirect contribution of economic sectors in the generation of GHG emissions, taking into consideration the size of the sector in terms of economic output formation and GHG emissions generation, and the assumed growth rates.
Our findings are aligned with the results of Liu et al. (2020), who found with the use of EE-IOA models that agriculture and electric power generation sectors have highemission intensities and a strong effect on other industries in Canada. The electricity, the metal and non-metal products and the land transport sectors are the most important CO 2 emitters in Cyprus . Morán and González (2007) found that the largest CO 2 emitters in Spain are the electricity and the metal and non-metal products sectors. Alcántara and Padilla (2020) showed that the sectors inducing more GHG emissions from other sectors in the Spanish economy are food manufacturing, wholesale and retail trade and construction. The metal products, the chemical products and the coal mining and petroleum processing products are the key GHG-emitting sectors in China (Guo et al. 2018;Shen et al. 2018;Yuan et al. 2020).
Identifying key GHG-emitting sectors, including both direct and indirect emissions, is, thus, crucial for formulating effective energy and environmental policies. Our findings stress the importance of an effective enforcement regime for reducing GHG emissions to counterbalance the adverse effects of economic growth. Moreover, they highlight the need for deploying advanced technologies to reduce sectoral emission intensities and contribute to the EU's medium-and long-term ambitious decarbonization targets. A broad portfolio of clean energy technologies will be needed to decarbonize all sectors of the economy, including further advancements in renewables, energy efficiency and storage and hydrogen-producing electrolyzers (IEA 2020).
Sectoral emission intensities are expected to vary greatly in the near future. Thus, if the focus is on individual GHG-emitting sectors, ignoring the complicated relations and linkages between different GHG-emitting sectors, the effectiveness of mitigation policies will be limited as less attention will be paid to sectors with low direct GHG emissions but high indirect linkages that drive other sectors to release emissions. The formulation and implementation of industrial emission reduction policies from the perspective of the whole industrial chain can be more efficient than policies focusing on specific individual sectors (Xie et al. 2016). Future research could analyse the emission intensities of the production sectors with their associated GHG abatement costs to explore the cost-effectiveness of abatement options within certain sectors to meet GHG emission reduction targets.
Published in partnership with CECCR at King Abdulaziz University Fish and other fishing products; aquaculture products; support services to fishing A03 4 Mining and quarrying B 5 Food, beverages and tobacco products C10-12 6 Textiles, wearing apparel, leather and related products C13-15 7 Wood and of products of wood and cork, except furniture; articles of straw and plaiting materials C16 8 Paper and paper products C17 9 Printing and recording services C18 10 Coke and refined petroleum products C19 11 Chemicals and chemical products C20 12 Basic pharmaceutical products and pharmaceutical preparations C21

Declarations
Conflict of interest On behalf of all authors, the corresponding author states that there is no conflict of interest.
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:// creat iveco mmons. org/ licen ses/ by/4. 0/.