The expansion of global LNG trade and its implications for CH4 emissions mitigation

In recent years, the global trade in liquefied natural gas (LNG) has experienced significant growth, leading to a rise in the effect of embodied methane (CH4) emissions between economies. This study investigates the spatiotemporal evolution of these CH4 emissions embodied within the global LNG trade and examines the associated network characteristics between the years 2011 and 2021. The findings reveal a substantial increase of 43.3% CH4 emissions embodied in global LNG trade, reaching a peak of 2.75 Tg in 2021, which equates to a monetary value exceeding 5 billion USD in terms of natural gas. Over the study period, these emissions aggregated to a total of 1987.92 Mt CO2-eq and 718.06 Mt CO2-eq, based on the respective global warming potential values over 20 year and 100 year timeframes. Our investigation of this complex network reveals the emergence of multiple robust hub economies, which have exerted significant influence over the dynamics of supply-demand in embodied CH4 trade, as well as the structure of trade communities. Notably, the Asia-Pacific trading community has exhibited exceptional growth and prominence in this market. Such growth is directly related to an increase in embodied CH4 emissions and their overall standing within this community’s network. The US has steadily attained dominance within an export-oriented community, which encompasses economies in South and North America, as well as certain regions of Europe. Moreover, the redistribution of LNG-related CH4 emissions among economies is significantly impacted by the intensity of production emissions and the volume of LNG trade. This reveals the potential of these hub economies to drive substantial reductions in CH4 emissions by implementing targeted energy and climate policies, which they have launched. Reinforcing coalitions and fostering closer collaboration within these communities can provide a robust foundation for technological advancements and transformative changes in trade structures.


Introduction
The utilization of natural gas as a transitional fuel in the zero-emission global energy sector has gained momentum due to its lower CO 2 emissions compared to coal and oil.A remarkable surge in the international trade magnitude of natural gas, specifically in the liquefied form liquefied natural gas (LNG), is a trend that has become conspicuous in recent times.During the interval between the years 2011 and 2021, LNG trade witnessed a substantial increase of 124.7 Mt.This growth trajectory is projected to continue given the ongoing significance of energy security.Researchers such as Shi and Variam (2016) and Liu et al (2020) have diligently undertaken extensive inquiries, delving into the shifting contours of competition and the impact of economic evolution on the dynamics of the LNG trade.Meza and Koç (2021) discussed competitive considerations and future strategies across diverse economies and regions.Geng et al (2014), Chen et al (2016) and Peng et al (2021) have continuously monitored the dynamic characteristics in changes of holistic LNG trade networks.
Though many studies into the global LNG trade have been conducted, few papers have noted its implied environmental impact.It is crucial that emissions of methane (CH 4 ), a highly potent greenhouse gas (GHG), which constitutes a significant proportion of natural gas, be accounted for.Methane has a global warming potential of 82.5 times more than CO 2 over a 20 year time horizon and 29.8 times more over a 100 year time horizon, as stated in the IPCC AR6 report.Even small rates of CH 4 emissions can have a considerable influence on global GHG footprints and can diminish the emissions mitigation benefits of natural gas.Therefore, accurate estimation of CH 4 emissions associated with LNG trade and consumption is essential for future mitigation efforts.To capture embodied emission flows, previous studies have employed life-cycle assessment methods to enhance emission accounting and track GHG emissions along the full supply chains (Gilbert and Sovacool 2018, Cooper et al 2021, Roman-White et al 2021, Balcombe et al 2022).Furthermore, research has explored the implications for climate strategy and internationally transferred mitigation outcomes for both LNG exporters and importers (George et al 2018, Egging-Bratseth et al 2022, Kotagodahetti et al 2022).
However, existing examinations of CH 4 emissions embodied in LNG trade have predominantly centered on either a production-based or a consumptionbased perspective, often confining their scope to single supply chains and neglecting the intricate dynamics of emission flows.To address this research gap, there is a critical need for a more holistic approach that delves into the structural features of this complex trade.A promising avenue to bridge this gap lies in the application of complex network analysis, which can abstract trading relationships between economies into a topological network and reveal important trade characteristics (Serrano andBogun´a 2003, Garlaschelli andLoffredo 2005).This approach has been applied to studies of international trade and trade-related GHGs.For instance, Ma et al (2021) identified a health damage network related to global ammonia emissions.Wu et al (2022) explored the features of the global air pollution flow network of the oil refining industry, and Wang et al (2022) uncovered spatiotemporal patterns and network characteristics of GHG emissions embodied in global livestock products.These studies demonstrate the effectiveness of employing complex network analysis in investigating the characteristics of emissions embodied in trade.Therefore, employing complex network analysis to investigate the characteristics of CH 4 emissions embodied in global LNG trade would contribute to a quantitative assessment of the environmental responsibilities of global economies.
The aim of this study is to enable a more comprehensive understanding of the emission flow patterns embodied in LNG trade.We aim to shed light on previously unexplored aspects on interlaced CH 4 flows from a network perspective and inform targeted strategies for effective emissions mitigation efforts.Our study implements enhanced accounting methodology, involving the integration of both conventional and unconventional gas production aspects.We reveal the spatiotemporal and network characteristics of these embodied CH 4 emissions during the years between 2011 and 2021 and identify hub economies and regional hotspots along the supply chain.(2022).This comprehensive dataset encompasses diverse stages of production, gathering, and processing across onshore and offshore oil and gas facilities.LP is corresponding CH 4 emission intensity from LNG-specific production stages:

Methods and data
where LF represents CH 4 emission factor of liquefaction and shipping, roughly 0.1% of LNG carried (IEA 2023).VP is the vaporization factor, which contributes around 0.018% (Innocenti et al 2023).However, the fiducial value varies considerably between conventional and unconventional gas production.To enhance the accuracy of EF, as shown in figure 1, we modify Fiducial value by weighting the sum of conventional and unconventional gas production respectively, and: where CF and UF refer to economy-specific CH 4 emission factors of conventional and unconventional gas production in IEA (2022), respectively.CP and UP are the proportion of conventional and unconventional gas production, respectively.The statistics for Australia are from the Department of Climate Change, Energy, the Environment and Water (2023).Furthermore, by putting mitigation measures in place, CH 4 emissions can be lowered by more than 75% (IEA 2022).We have meticulously tailored the CH 4 intensity estimates for the US and Norway, taking into account additional data sources (Thinkstep 2017, Equinor 2021, EIA 2023, EPA 2023) in order to accurately depict their respective efforts to reduce CH 4 emissions.For the rest economies, where relevant statistics could not be obtained, we conservatively estimated them as conventional gas economies with static intensities.
The CH 4 emissions embodied in global LNG trade were determined by multiplying the volume of LNG trade with the corresponding economy specific CH 4 emission factors in the exporting economy: where MEF i refers to the CH 4 emissions embodied in LNG product imports, and ADM ij represents the import volume from economy i to economy j.XEF j represents the CH 4 emission involved in LNG product exports, where ADX ij is the export volume from i to j.

Complex network analysis
We

Degree and degree distribution
The in-degree k i in and out-degree k i out represent the quantity of economies that a given economy is importing or exporting CH 4 emissions, shown as: e ji (7) where the element e ij indicates whether CH 4 emissions are flowing from economy i to economy j, e ij = 1 when a trade flow exists from i to j, otherwise, e ji = 0 (Freeman 1979).n is the total number of economies.The node strength s is the sum of the link weightings.Considering the inconsistent CH 4 emissions in different links, in-strength s i in and out-strength s i out represent amount of imported and exported CH 4 emissions.To analyze the scale-free structure of MEN, we defined the node strength distribution p (s) as p (s) = ns n .The network can be characterized as a scale-free network if its degree distribution is well fitted by a power law distribution, that is p (s) ∼s −α , where α is the power-law index.

Node centrality
As connectivity and intermediality measures, betweenness centrality BC portrays the shortest pathways through the node to the total number of shortest paths in the network (Freeman 1977): where n st is the number of shortest paths between economy s and economy t, and n i st is the number of shortest paths between s and t that pass through economy i.
Eigenvector centrality evaluates node importance while considering the importance of its neighbors in the network, which is defined as: where λ and v j are the largest eigenvalue and associated eigenvectors, respectively.

Community detection and modularity
A community is a collection of nodes that are more closely linked among themselves than they are with other nodes.This study applies the algorithm proposed by Blondel et al (2008) to discover the optimal community partition.The modularity Q is defined as: where m = 1 2 ∑ ij w ij denotes the sum of all the weights in MEN.w ij is the weighted edges between economy i and economy j, S i = ∑ j w ij denotes the weight of edge connected to economy i. c i is the community that contains node i, and δ On the LNG demand side, the implementation of the Paris Agreement acted as a catalyst, leading to a pronounced surge among emerging importers in Asia, particularly notable since 2016.This surge has been most pronounced in economies like China, driven by the growth in gas for power generation and coal-to-gas switching.In figure 2(d), we see that Japan led the imports in 2011 with 563.2 Gg of embodied CH 4 emissions, accounting for 29.3% of the global total.In the case of Japan, this is a result of the loss of nuclear power generation capacity.South Korea, Spain, the UK, and France followed as second to fifth largest importers, accounting for 14.9%, 8.8%, 7.1% and 5.8% of the global total, respectively.Over the years, the number of participants and the volume of imports increased.While Asian markets saw an upsurge in LNG demand, European economies redirected their focus to pipeline gas.China's imports grew significantly as a result of its strong environmental policy, surpassing Japan as the largest importer in 2021 with imports totaling 519.6 Gg CH 4 emissions (see also figure 2(b)).Furthermore, the Asia Pacific region, with a share increasing from 59.6% in 2011 to 66.9% in 2021, continued to be the largest market, with the growth rate outpacing the supply-side growth.

CH
The majority of CH 4 transfer fluxes related to global LNG trade were internal transfers within the Asia Pacific region, which exceeded 25% of the total in both 2011 and 2021.The second-largest embodied emission fluxes were from the Middle East to the Asia Pacific region, while the CH 4 emission flow from Africa to Europe was also substantial.Figure 3 presents the main CH 4 transfer fluxes via the LNG trade.The largest flow of embodied CH 4 emissions was from Qatar to the United Kingdom in 2011, shifting to the trade between Australia and China in 2021.The top ten trade flows between economies accounted for 40.1% of the total embodied CH 4 emissions in 2011 and 30.0% in 2021, respectively.

Topological analysis of embodied emission networks
As the number of participants and their trade expanded over the duration of the studied period, the network of embodied CH 4 emission transfers grew increasingly intricate.The fact that there are now 56 nodes instead of 43 shows that more economies are participating in the international trade of LNG products.According to table 1, the average node strength fell from 44.6 to 40.1 before 2016 and subsequently had a strong return to 49.2 in 2021 as a result of the Paris Agreement.However, in 2020, there was a slight downward trend in the number of edges and average strength, suggesting that the development of the LNG trade was hindered by the outbreak of the COVID-19 pandemic.
Commonly, the LNG trade network has been observed to exhibit scale-free properties in node degree (Geng et al 2014).This study delves into an in-depth exploration of the distribution patterns pertaining to both node in-strength and out-strength.The distributions in 2011 and 2021, as depicted in figure S1, demonstrate adherence to a powerlaw distribution, thereby highlighting the intrinsic heterogeneity within the network.The degree and strength of outward imports reflect an expanded scale of importers, whereas the inward migration of export indicators signifies a centralized tendency toward the top exporters.Notably, both the export and import node strength-to-node degree ratios exhibit an increasing trend, with exporters exhibiting a higher growth rate, indicating their greater influence within the network.
This study examines the evolution of hub economies by analyzing node strength and node degree.
In 2011, Japan and South Korea were primary economies characterized by high import strength and node degree, and in 2021, China, India, and Spain subsequently joined their ranks.In terms of exports, Qatar, Nigeria, and Algeria held the distinction of being leading exporters with high export strength and node degree in 2011, and in 2021, the US, Russia, and Australia entered this group.Corroborated by the previous accounting results, it becomes evident that several emerging economies are assuming hub roles, even as the more traditional counterparts continue to maintain their positions.
To further discern the key economies, we employ network indicators including eigenvector centrality and betweenness centrality.The involvement of China, South Korea, Japan, India, and Thailand in the MEN has deepened, establishing closer connections with vital exporters such as Qatar, Australia, the US, and Russia.In contrast, European economies including Spain, Belgium, and the UK have experienced a decline in their control and influence over discourse in recent years, yet they continue to play a significant role in the global LNG trade.Certain economies, namely the US, Indonesia, Malaysia, and the United Arab Emirates, function as important bridge entities facilitating the flow of CH 4 -related trade.However, the influence of transit economies is limited by the constraints imposed by the flexibility of LNG transportation methods.

Community partition for embodied emission networks
Changes in the LNG market's supply-demand dynamics (see figure S2) also had a significant impact on the trade patterns of embodied CH 4 emissions.The modularity and members of communities fluctuated erratically between 2011 and 2021 (see tables S2 and S3), demonstrating the changing patterns within the network of embodied CH 4 emissions over time.The modularity network showed that the MEN is roughly regionalized.However, with the increased utilization of LNG flexibility, the modularity declined from 0.38 to 0.29, showing a slight tendency towards a less tightly-knit community structure.To capture the evolution of trade patterns, the community partition is visualized in figure 4.
In 2011, the MEN was divided into four trading communities that were relatively closely connected.Community-1 included 16 economies, mostly located in Russia, the Asia Pacific region, and the Arabian Peninsula.As Japan and South Korea were the majority dominators, this community imported about 60% of its CH 4 emissions from other communities and embodied the largest CH 4 emissions, totaling 1823.1 Gg.Community-2 mainly consisted of Mediterranean economies, dominated by Algeria and Nigeria.Community-3 exhibited dispersed geospatial characters, with Qatar serving as the main hub economy.Compared to the communities mentioned above, Community-4 was the smallest community with only 141.9 Gg of CH 4 emissions, and it was geographically concentrated in South and North America.
In 2021, the MEN still encompassed four distinct communities; however, the fragility and complexity of the network, were further exacerbated by the prevailing anti-globalization sentiment, coupled with the growing significance of emerging hubs.Community-1 remained the largest community in terms of embodied CH 4 emissions with 2154.6 Gg.The composition of this community remained relatively stable, with an obvious increase in both embodied CH 4 emissions and eigenvector centrality-particularly for Russia, Australia and China.Community-2 and Community-3 exhibited notable geographic dispersion characteristics, which contributed to a more extensive distribution of LNG trade.In Community-4, the US experienced remarkable growth and dominated the economies of South and North America, as well as certain parts of Europe.Notably, Community-4 exported its 64.1% CH 4 emissions to other communities, demonstrating a need for greater responsibility beyond its direct emissions.On the other hand, the other three communities imported up to 55% of their CH 4 emissions from external sources, suggesting the importance of focusing on reducing their own emissions.

Discussion and conclusion
This study delves into the spatial-temporal patterns characterizing CH 4 emission embodied in the global LNG trade.Given the escalating global concerns about energy security and the environment, it is expected that both LNG trade and corresponding CH 4 emission effects will continue to grow.This increase in demand is driven by the geographic disparity between production and consumption regions, along with the need to replace coal and environmentally harmful liquid fuels.However, the production of LNG has not been able to match the pace of demand, highlighting the importance of accounting for the combined effects of diverse emission intensities at the national level and the expanding volume of trade through emission transfers.
For LNG exporters, CH 4 intensity serve as an effective tool for assessing the potential to reduce emissions within the gas industry by marketing it.In an effort to boost environmental competitiveness in gas industry, the US has proposed charging the oil and gas sectors for their CH 4 emissions, with this policy set to take effect in 2024 (IEA 2022).Conversely, for LNG importers, the arrival of embodied CH 4 emissions could have a significant impact on their energy balance.This is due to the fact that it reduces the time needed to repurpose existing gas infrastructure and up the switch from natural gas to decarbonized energy sources.Therefore, the establishment of a comprehensive framework for measurement, reporting, and verification, along with border adjustment mechanisms, stands as essential components of effective mitigation along global supply chains.As a pioneer in this effort, the European Union's Regulation on methane emissions reduction in the energy sector exemplifies the feasibility of such initiatives (European Parliament 2023).
As revealed through the illustration of a scale-free distribution, an imminent priority emerges in establishing a system for emission mitigation responsibility that acknowledges heterogeneous levels.Within this framework, distinct economies can effectively collaborate in a concerted response commensurate with their roles and contributions within the global context (Chen et al 2018, Li et al 2020).Efforts led by hub economies, such as the Coalition for LNG Emission Abatement towards Net Zero (CLEAN) have convened major LNG exporters and importers to combat global warming.It is crucial to emphasize that hub economies lacking the capacity for mitigation require global efforts in the form of financial assistance or tailored regulations.
Trade relationships can contribute to mitigating global CH 4 emissions if importers opt to shift directly to sources with lower emission intensities (Dalin et al 2012).But such adjustments may give rise to environmental spillovers, where importers derive direct benefits from the diminished emissions, while exporters may encounter economic losses related to their exports.Therefore, it is crucial to promote collaborative efforts that enhance the interaction between upstream and downstream economies, particularly concerning climate and environmental matters (Liu et al 2022).The identification of a cluster structure through community partitioning plays a pivotal role in comprehending and governing global CH 4 emissions.This is because within a community, there are fewer trade barriers, and member states are more inclined to engage in collaborative efforts.(Ma et al 2021).Similar to the patterns observed in embodied energy-related emissions (Xu et al 2023), the network of CH 4 emissions embodied in LNG trade also exhibits community structures.By fostering collaborative efforts within these communities, economies can effectively harness their collective knowledge, resources, and influence to substantially mitigate CH 4 emissions.Given that the repercussions of emission mitigation policies in a particular economy can swiftly impact others within the same community (Liang et al 2015), collaborative efforts at the community level expand the effectiveness of emissions reduction policies and highlight the importance of united action within these communities to address emissions.A notable intergovernmental coordination is the Joint Declaration issued at COP27 by prominent economies including the US, European Union, Japan, Canada, Norway, Singapore, and the United Kingdom, most of whom are members of Community-4.
However, it is important to note that the use of static EF in this study may have introduced estimation bias.More field data and appropriate emission factors are required for in-depth precise emission estimation.Importantly, the emission impacts of large-scale unconventional gas exploration, such as shale gas and coalbed CH 4, tend to be more substantial due to their higher emission intensities.A thorough understanding requires coordinated efforts to work with inventories, as well as to assure the collection of complete data and deep consideration of emission factors.Hence, it is of great value to achieve precise measurement of CH 4 emissions and plot highresolution maps of CH 4 emissions in finer granularity.Notably, the scope of this study only considers the upstream emissions of imported LNG.A multitude of sources, encompassing oil and gas production through to consumption along the value chains, including fuel combustion, merit thorough exploration.To provide a holistic representation of the full life cycle of GHG emissions, other interrelated processes integral to LNG production and consumption should also be considered.Furthermore, emissions associated with LNG transportation, albeit a minor contributor, should not be overlooked when compiling the emission inventory.
build a directed-weighted networkemissions embodied in the global LNG trade, considering the directional nature of trade and the nonequivalence of trade links.The CH 4 emissions network (MEN) is represented by a set G = (V, E, W), where n = |V| is the economy participating in global trade.E = { e ij } is the flow in trading relationships.W = { w ij } is the CH 4 emission volumes embodied in LNG trade.

Figure 4 .
Figure 4. evolution of communities' partition in (a) 2011 and (b) 2021.The size of the circles represents trade related CH4 emissions.The width of connections indicates the amount of embodied CH4 emissions flow.

4 emissions embodied in global LNG trade
9% of global CH 4 emissions (figure 2(c)).In the following few years, economies including Indonesia, Trinidad and Tobago, Norway, and Nigeria saw a dramatic decline in exports, and participants like Yemen phased out.At the same time due to the start-up of multiple LNG projects, some emerging economies like Australia and Russia rose to 277.9 Gg and 197.0 Gg, respectively, and quickly caught up.
Australia, which in 2020, led it to surpassing Qatar as the top exporter.Additionally, the US capitalized on the shale gas revolution, enabling it to transition into a prominent exporting nation.Figure2(a) depicts temporal patterns of embodied CH 4 emissions of major exporters.In 2011, the top 5 exporters-Qatar, Nigeria, Indonesia, Algeria and Malaysia-accounted for 65.