Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessArticle

Peer-Review Record

A Machine Learning and Panel Data Analysis of N₂O Emissions in an ESG Framework

Sustainability 2025, 17(10), 4433; https://doi.org/10.3390/su17104433

by Carlo Drago^1,*

, Massimo Arnone² and Angelo Leogrande^3,*

Reviewer 1:

Li-Jing Liu

Reviewer 2: Anonymous

Reviewer 3:

Hamid Ghazi Sulimany

Reviewer 4:

Nurcan Kilinc-Ata

Reviewer 5:

Jorge Alejandro Silva Rodríguez De San Miguel

Sustainability 2025, 17(10), 4433; https://doi.org/10.3390/su17104433

Submission received: 22 March 2025 / Revised: 26 April 2025 / Accepted: 30 April 2025 / Published: 13 May 2025

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

N₂O and other non-CO₂ greenhouse gas emissions are increasingly becoming a research hotspot in the field of climate change. This paper employs econometric and machine learning models to explore the key influencing factors of global N₂O emissions from an ESG perspective. The research question holds certain importance and practical significance.

The authors are advised to structure the introduction section with subparagraphs to further enhance clarity.
The authors mention that the innovation lies in combining econometric models with machine learning algorithms. What are the specific advantages of this combination? What advantages do machine learning algorithms offer in analyzing N₂O influencing factors? What is the relationship between the ESG model and these two methods?
In the literature review section, the authors list numerous references but provide no summarizing explanation. What is the purpose of citing so many references? What is the relationship between the listed paragraphs?
The results section lacks summarization and refinement. For example, the regression analysis, cluster analysis, and machine learning regression analysis related to ESG are discussed separately, with insufficient attention to their interconnections. Why not discuss and model ESG within a unified framework?
The paper focuses on global N₂O emissions. At what scale were the variable data collected by the authors? Is the data volume sufficient for cluster and machine learning analysis?
Overall, the authors have done some work, but the paper's narrative is somewhat fragmented and requires further refinement.

Author Response

POINT TO POINT ANSWERS TO REVIEWER 1

Q1. The authors are advised to structure the introduction section with subparagraphs to further enhance clarity.

A1. Sub-paragraphs have been created as follows:

The study of nitrous oxide (N₂O) emissions in ESG models on a global level is a new frontier of research, characterized by increasing interest in innovative methodology for understanding the complexity of the interaction of the environmental, economic, and social determinants. While the literature on the emissions of greenhouse gases has been led by CO₂ and methane (CH₄), N₂O is relatively uncharted territory despite its high global warming potential and long-term contribution to climate change.

This research innovation is located at the nexus of environmental economics, sustainable finance, and predictive analytics, and takes an innovative methodological inspiration from the nexus of econometric techniques and machine learning with the aim of obtaining an improved understanding of the determinants of N₂O emissions and their interaction with ESG models.

The literature gap is evident on several fronts. First, much of the literature on analysis of the emissions of N₂O has been focused on sectoral analyses, the most prominent of which is agriculture and soil management practice, with little macroeconomic consideration of the mitigation role of ESG models in emissions reduction. While many studies have been done on the relationship between ESG performance and carbon emissions mitigation, few studies have made explicit reference to the consideration of the role of ESG factors in N₂O emissions on a global level. Second, traditional econometric techniques have been the overwhelming tool of examination of the effectiveness of environmental policy in the mitigation of the emissions of greenhouse gases, yet the application of machine learning methodologies remains limited in this respect.

The application of cutting-edge clustering and regression algorithms offers new leads in uncovering latent structure in the data and in model predictive power, informing a more fine-grained understanding of the underlying dynamics of N₂O emissions. The second innovative aspect of the research has to do with the application of large global datasets, including the World Bank ESG Database and other global datasets, to analyze N₂O emissions from a multidimensional lens. The data cover a long time span and include variables on economic, social, and environmental determinants of N₂O emissions, permitting close analysis of the interplay between environmental governance, economic growth, and abatement policy. The use of panel data models enables the estimation of between-country and across-time heterogeneity, and clustering routines enable the determination of clusters of countries with the same profile, both in emissions and ESG performance.

Methodologically, the research combines traditional econometric specifications, including fixed and random effects regressions, with machine learning algorithms, including density-based clustering, decision trees, regression via support vector machines, and boosting methods. This enables the comparison of predictive performance of competing models and tests the significance of ESG variables in explaining N₂O emissions. The further exploration of the importance of variables by dropout loss in machine learning models enables a quantitative assessment of the contribution of each factor to emissions and informs a deeper understanding of the nexus between ESG and environmental sustainability.

The research question is thus innovative in its potential to bring together different disciplines—economics, environmental science, and data science—to analyze a complex phenomenon such as N₂O emissions using an ESG lens. The interdisciplinary nature of the research is dictated by the necessity to address the issue of sustainability, as through it, we can develop more efficient predictive and abatement tools for emissions, and in doing so, we can ease the transition towards low-carbon economies. The proposed analysis also has important implications for environmental governance and public policy, as it provides empirical insights on the efficacy of ESG strategies in reducing the climate footprint of N₂O. In a setting in which investors and policymakers have increasing interest in sustainability, this research contributes meaningfully to data-driven emission strategies and the achievement of international climate goals. Overall, this research addresses a basic gap in the literature by suggesting a novel analysis of N₂O emissions through innovative data analytics and ESG models. The integration of econometrics and machine learning surmounts the limitations of traditional approaches and provides new empirical insights on the determinants of emissions, with implications for the development of more effective climate change mitigation strategies. This research is a step ahead in our knowledge of the economy-environment-sustainable finance nexus and provides analytical tools valuable to ESG practitioners, policymakers, and academics.

The article continues as follows. The second section presents the literature review, the third section show the data and methodology, the fourth section contains the econometric and machine learning analysis, the fifth section concludes.

Q2. The authors mention that the innovation lies in combining econometric models with machine learning algorithms. What are the specific advantages of this combination? What advantages do machine learning algorithms offer in analyzing N₂O influencing factors? What is the relationship between the ESG model and these two methods?

Within the methodology section we have added the following propositions:

A2. The article emphasizes how integrating econometric models with machine learning algorithms is innovative because it provides both predictive power and analytical depth when analyzing N2O emissions. Statistical inference, causal analysis, and control for unobserved heterogeneity across nations and over time are all made possible by econometric models like fixed and random effects regressions. Machine learning models, on the other hand, help by revealing intricate, nonlinear relationships, locating hidden structures in the data, and enhancing prediction accuracy. Additionally, these models provide variable importance metrics, such as dropout loss, which allow for a more accurate understanding of the factors that have the greatest impact on nitrous oxide emissions. In particular, machine learning is very good at determining important factors like net forest depletion, energy intensity, and forest area. Additionally, it makes it easier to group countries into clusters according to their emission characteristics and ESG performance, something that econometric models are less suitable for. The ESG model is a theoretical framework that combines governance, social, and environmental indicators to analyze how they relate to N2O emissions. This model is operationalized using both econometric and machine learning techniques: econometrics assesses statistical relationships and tests significance, whereas machine learning forecasts results and ranks ESG factors according to their influence. This collaboration enables the study to transcend conventional approaches and provide a more sophisticated, evidence-based comprehension of how environmental outcomes are influenced by ESG performance.

Q3. In the literature review section, the authors list numerous references but provide no summarizing explanation. What is the purpose of citing so many references? What is the relationship between the listed paragraphs?

A3. the literature analysis was reinterpreted in light of the research question with the addition of a summary table.

An increasing amount of research has started to examine how the Environmental, Social, and Governance (ESG) framework can be used to address nitrous oxide (N2O) emissions, one of the most powerful greenhouse gases. Researchers from a variety of fields and disciplines stress that integrated approaches are necessary for successful mitigation and that each ESG pillar interacts with N2O dynamics in a unique way.

Numerous studies highlight agriculture, land use, and energy systems as important sources of N2O from an environmental perspective. According to ESG tracking principles, Čapla et al. (2025) recommend that the agri-food sector adopt carbon footprinting, focusing on fertilizer use and livestock farming. Biochar and agricultural waste are evaluated by Park et al. (2024) and Padhi et al. (2024) as soil management techniques that directly lower N₂O flux, in line with ESG-compliant sustainable agriculture. China's decreasing cropland emissions and the ESG-driven agricultural policies that underpin this trend are discussed by Cui et al. (2022). The Bangladeshi textile industry is examined by Biswas et al. (2024) and Biswas et al. (2023), who demonstrate how ESG energy standards lower N₂O emissions during production. Lambiasi et al. (2024) call for ESG-compliant water and sanitation systems and emphasize the role of wastewater treatment in N2O emissions in the industrial and energy sectors. In its study of emissions reduction in the EU's oil and energy companies, Voicu (2023) illustrates sector-level approaches to lowering N₂O. Thermal power plant operations are linked to N2O output by An et al. (2022), which emphasizes the necessity of ESG-linked monitoring in industrial sustainability.

One important factor that makes emissions control possible is the governance aspect of ESG. Saudi Arabia's energy transition and net-zero strategy are examined by Al-Sinan et al. (2023) as a forerunner to ESG-aligned climate governance. ESG benchmarks based on climate stressors such as the Heat Index 35 require N₂O abatement, according to Drago and Leogrande (2024). Turjak (2023) highlights the need for stronger ESG regulations by criticizing EU environmental governance flaws that permit N2O emissions to continue. Schuuring (2024) confirms that institutional quality moderates ESG outcomes by highlighting the effectiveness of stronger governance institutions in controlling N₂O. Furthermore, Orsini (2022) links improved N₂O reporting to improved ESG performance in their evaluation of the UK's 2013 disclosure requirements. Blair (2021) demonstrates how the energy sector in Canada is creating standards for N₂O disclosure. In his study of Scope 3 greenhouse gas reporting in Norway, Stinchcombe (2023) argues for full ESG reporting that takes indirect N2 emissions into account. In their critique of disjointed ESG reporting standards, Kaplan and Ramanna (2021) make the case for standardized frameworks that incorporate N₂O transparency. By incorporating N2O emissions monitoring into smart city ESG planning, Gu et al. (2024) apply these insights to the urban scale.

ESG-aligned investments are becoming more and more linked to N₂O risk from a financial and market standpoint. A net-zero investment alignment model is put forth by Sidestam and Karam (2024), who highlight the financial sector's catalytic role in reducing emissions, including N₂O. ESG investment systems can support green transitions and give low-N₂O assets priority, as demonstrated by Sacco et al. (2023). While Mamatzakis and Tzouvanas (2025) investigate how disclosures of emissions affect investor behavior, Rothman (2023) finds that ESG metrics can predict firm-level N₂O emissions. Investment rerouting away from high-N₂O emitters is discussed by Yoshino and Yuyama (2021) and Kannoa (n.d.), who associate emissions with the risk of firm default. Additionally, green finance mechanisms have the potential to reduce N₂O. Green bonds are examined by Çıtak and Meo (2024) as instruments for projects aimed at reducing emissions. Brühl (2021) outlines the regulatory levers that allow for the mitigation of N₂O through financial instruments linked to ESG. The effects of carbon risk on stock performance and corporate incentives are examined by Boubaker et al. (2024) and Bolton et al. (2022), with implications for investment strategies related to N₂O. Dennis and Iscan (2024) create a climate risk transition metric that takes into account N2O exposure as well as firm-level emissions efficiency. The function of data technologies in incorporating N₂O into ESG analysis is a recurring theme in the literature. Wang et al. (2024) forecast how biochar aging will affect N2O emissions using machine learning. A digital carbon forecasting platform that can be expanded to N2O is created by Rafiee et al. (2022). Muller (2021) investigates how firm-level environmental data inform ESG investing and indirectly affect N2O mitigation, while Jiang et al. (2022) suggest blockchain-based ESG systems for emissions traceability. Long-term N₂O monitoring is crucial for wastewater treatment, according to Gruber (2021), and Zhang et al. (2024) provide techniques for calculating commodity-based emissions, such as N₂O, from company data. Geographical and social factors also come into play. In their assessment of the correlation between emissions and environmental quality across Canadian provinces, Ali et al. (2025) provide support for mitigation policies that are ESG compliant. Businesses that include N₂O reduction in their ESG disclosures gain a competitive advantage, according to Agbo et al. (2024). According to Yulianti et al. (2023), N2O control is a crucial part of ESG investment, which is positioned as a means of accomplishing sustainable development goals. Prieto (2022) looks into ESG risks in the construction and engineering sectors, which are connected to N2O emissions through material processing. The necessity of incorporating N₂O into climate-related ESG decision-making is emphasized by Squillace (2023). With possible co-benefits for N2O offsetting, Ng and Webber (2023) apply natural climate solutions to carbon accounting. Innovation policy and green development are linked by Long and Feng (n.d.), who emphasize the importance of ESG-guided technology in reducing NOx emissions.

All things considered, these studies agree that N₂O needs to be more explicitly incorporated into ESG frameworks as a signal for institutional, financial, and regulatory strategies as well as a measure of environmental performance. The ESG model is a multifaceted tool that can effectively guide mitigation across sectors and regions and capture the complexity of N2O emissions, whether through disclosure, investment, policy reform, or technological innovation.

Table 1. Synthesis of the literature review.

ESG Pillar	Macro- Theme	Focus on N₂O Emissions	Key References
Environmental (E)	Agriculture, Land Use & Industry	N₂O emissions from fertilizer, livestock, biochar, wastewater, textile and energy sectors	Čapla et al. (2025), Padhi et al. (2024), Park et al. (2024), Cui et al. (2022), Harasheh & Harasheh (2021), Biswas et al. (2023, 2024), Lambiasi et al. (2024), Voicu (2023), Prieto (2022), An et al. (2022)
Environmental (E)	Monitoring & Environmental Technologies	N₂O prediction and tracking using digital tools, ML, blockchain, and urban ESG systems	Wang et al. (2024), Rafiee et al. (2022), Jiang et al. (2022), Muller (2021), Gruber (2021), Gu et al. (2024)
Social (S)	Social Equity, Accountability & Inclusion	Social factors in emissions management, SDGs, corporate ESG disclosures, and Scope 3 N₂O tracking	Ali et al. (2025), Yulianti et al. (2023), Squillace (2023), Stinchcombe (2023), Biswas et al. (2023), Agbo et al. (2024)
Governance (G)	Policy, Regulation & Institutional Capacity	Effectiveness of ESG-aligned policies, mandatory reporting, governance quality, climate compliance	Al-Sinan et al. (2023), Orsini (2022), Schuuring (2024), Turjak (2023), Blair (2021), Guermazi et al. (2025), Kaplan & Ramanna (2021), Jiang et al. (2022), Stinchcombe (2023)
Governance (G)	Risk Management & ESG Standards	ESG standardization, default risk for polluting firms, and transition readiness	Rothman (2023), Dennis & Iscan (2024), Kannoa (n.d.)
Finance (Cross-cutting)	ESG Investment, Green Finance & Disclosure	Use of green bonds, financial risk metrics, investor preferences, and carbon/N₂O footprint accountability	Sidestam & Karam (2024), Sacco et al. (2023), Yoshino & Yuyama (2021), Micol & Costa (2023), Çıtak & Meo (2024), Brühl (2021), Mamatzakis & Tzouvanas (2025), Boubaker et al. (2024), Bolton et al. (2022), Zhang et al. (2024)
Methodological (Cross-cutting)	Predictive Analytics & ESG Benchmarking	ESG index construction, forecasting tools, and integration of N₂O into ESG scoring systems	Drago & Leogrande (2024), Long & Feng (n.d.)

Q4. The results section lacks summarization and refinement. For example, the regression analysis, cluster analysis, and machine learning regression analysis related to ESG are discussed separately, with insufficient attention to their interconnections. Why not discuss and model ESG within a unified framework?

Q5. A new results section has been introduced before the conclusions through a tabular summary of the results obtained.

Discussions of the results

Unpacking the complexity and multifaceted nature of the ESG framework in relation to nitrous oxide (N₂O) emissions requires examining the effects of each ESG component separately: environmental, social, and governance. Different but connected systems of influence are encapsulated in each pillar. The biophysical pathways of N2O emissions are directly influenced by environmental factors, including energy intensity and forest depletion. Social indicators, such as fertility, income distribution, and access to essential services, are a reflection of economic and demographic forces that change emissions, land use, and consumption patterns. Institutional capacity to implement and enforce sustainability measures is determined by governance-related factors, including legal institutions and regulatory quality. The analysis enables a more accurate identification of the channels through which emissions are generated and managed by evaluating each component separately. Because it avoids confusing the distinct effects of variables that function on various levels—physical, social, and institutional—this decomposition is methodologically relevant. The foundation for creating an integrated framework that explicitly models the interplay between ESG dimensions is also laid by these deconstructed insights. Establishing the individual influence of each pillar, for example, is necessary to understand how strong governance can either moderate or amplify the environmental and social drivers of emissions. As a result, distinct analyses are essential for both analytical clarity and the development of a cohesive, systems-based ESG approach to emissions management and climate policy.

Macro	Variables	Panel Data Relationships	Best Clustering algorithm	Clusters	Best ML Algorithm	Mean Dropout Loss	Mean decrease in accuracy
E	ASFND	Positive	Density-Based	2	K-Nearest Neighbors	5.537×10^(+13)
	EIPE	Negative				7.093×10^(+13)
	FA	Positive				5.537×10^(+13)
S	AGRI	Positive	Hierarchical clustering	8	Random Forest	6.167×10^(+13)
	FRT	Positive				7.710×10^(+13)
	GI	Negative				6.167×10^(+13)
	ISL20	Positive				6.167×10^(+13)
	WATER	Negative				7.471×10^(+13)
G	GDPG	Positive	Hierarchical Clustering	10	Random Forest		3.630×10^(+26)
	FMLP	Negative					2.076×10^(+27)
	RQE	Positive					2.195×10^(+27)
	RDE	Positive					1.385×10^(+27)
	STJA	Negative					2.171×10^(+27)
	SLRI	Negative					1.153×10^(+27)

In conclusion, the precise causal pathways that propel N2O emissions can be identified by dissecting the ESG model into its constituent parts. The foundation for a systemic integration, where the relationships between the environmental, social, and governance dimensions must be jointly modeled, is laid by this disaggregated analysis. A unified viewpoint is necessary for the efficacy of ESG policies since it becomes clear that strong governance can either mitigate or magnify the effects of environmental and social factors. The next critical step in tackling complicated environmental issues like nitrous oxide emissions on a global scale is to transition from a compartmentalized approach to an integrated and predictive ESG model.

Q5. The paper focuses on global N₂O emissions. At what scale were the variable data collected by the authors? Is the data volume sufficient for cluster and machine learning analysis?

A5. In the methodology and data section we have indicated the definitions of the N20 variable as follows. The data refer to 190 countries for 10 years.

Variable	Acronym	Description
Nitrous oxide emissions (metric tons of CO2 equivalent per capita)	NOE	This metric quantifies annual nitrous oxide (N₂O) emissions from agriculture, energy, waste, and industry, excluding LULUCF. Emissions are converted to carbon dioxide equivalents (CO₂e) using Global Warming Potential (GWP) factors from the IPCC’s Fifth Assessment Report (AR5), ensuring consistency in climate impact assessments.

Q6. Overall, the authors have done some work, but the paper's narrative is somewhat fragmented and requires further refinement.

A6. We tried to bring the structured metric analysis back to the ESG narrative even if probably the abundance of technical instruments applied reduces the narrative evidence. The narrative framework remains linked to ESG theories.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Literature reviews can be classified according to research topics, but current literature reviews lack a logical framework. The key theoretical framework was not mentioned, resulting in insufficient theoretical support for analyzing driving mechanisms. It is suggested that the integration of theoretical framework and literature be enhanced and the research boundaries be clarified.

The time range of the data is not clear (only mentioning the "long-term span"), and specific years need to be supplemented (such as 2010-2020)

Some conclusions contradict theoretical expectations, such as the positive correlation between regulatory quality (RQE) and N2O emissions, but the underlying mechanisms have not been thoroughly explored.

The parameter settings of the machine learning model, such as the neighborhood radius of density clustering and the k value of KNN, have not been fully explained, which may affect the reproducibility of the results. Suggest supplementing the selection criteria for algorithm hyperparameters and adding model robustness testing.

The presentation of charts is recommended to optimize their expression, as they are currently not easy to read.

Author Response

POINT TO POINT ANSWERS TO REVIEWER 2

Q1. Literature reviews can be classified according to research topics, but current literature reviews lack a logical framework. The key theoretical framework was not mentioned, resulting in insufficient theoretical support for analyzing driving mechanisms. It is suggested that the integration of theoretical framework and literature be enhanced and the research boundaries be clarified.

A1. The literature analysis was reinterpreted in light of the research question with the addition of a summary table.

Table 1. Synthesis of the literature review.

ESG Pillar	Macro- Theme	Focus on N₂O Emissions	Key References
Environmental (E)	Agriculture, Land Use & Industry	N₂O emissions from fertilizer, livestock, biochar, wastewater, textile and energy sectors	Čapla et al. (2025), Padhi et al. (2024), Park et al. (2024), Cui et al. (2022), Harasheh & Harasheh (2021), Biswas et al. (2023, 2024), Lambiasi et al. (2024), Voicu (2023), Prieto (2022), An et al. (2022)
Environmental (E)	Monitoring & Environmental Technologies	N₂O prediction and tracking using digital tools, ML, blockchain, and urban ESG systems	Wang et al. (2024), Rafiee et al. (2022), Jiang et al. (2022), Muller (2021), Gruber (2021), Gu et al. (2024)
Social (S)	Social Equity, Accountability & Inclusion	Social factors in emissions management, SDGs, corporate ESG disclosures, and Scope 3 N₂O tracking	Ali et al. (2025), Yulianti et al. (2023), Squillace (2023), Stinchcombe (2023), Biswas et al. (2023), Agbo et al. (2024)
Governance (G)	Policy, Regulation & Institutional Capacity	Effectiveness of ESG-aligned policies, mandatory reporting, governance quality, climate compliance	Al-Sinan et al. (2023), Orsini (2022), Schuuring (2024), Turjak (2023), Blair (2021), Guermazi et al. (2025), Kaplan & Ramanna (2021), Jiang et al. (2022), Stinchcombe (2023)
Governance (G)	Risk Management & ESG Standards	ESG standardization, default risk for polluting firms, and transition readiness	Rothman (2023), Dennis & Iscan (2024), Kannoa (n.d.)
Finance (Cross-cutting)	ESG Investment, Green Finance & Disclosure	Use of green bonds, financial risk metrics, investor preferences, and carbon/N₂O footprint accountability	Sidestam & Karam (2024), Sacco et al. (2023), Yoshino & Yuyama (2021), Micol & Costa (2023), Çıtak & Meo (2024), Brühl (2021), Mamatzakis & Tzouvanas (2025), Boubaker et al. (2024), Bolton et al. (2022), Zhang et al. (2024)
Methodological (Cross-cutting)	Predictive Analytics & ESG Benchmarking	ESG index construction, forecasting tools, and integration of N₂O into ESG scoring systems	Drago & Leogrande (2024), Long & Feng (n.d.)

Q2. The time range of the data is not clear (only mentioning the "long-term span"), and specific years need to be supplemented (such as 2010-2020)

A2. We have specified the time range in each equation as follows “t=[2011;2020]”.

Q3. Some conclusions contradict theoretical expectations, such as the positive correlation between regulatory quality (RQE) and N2O emissions, but the underlying mechanisms have not been thoroughly explored.

A.3 This result seems counterintuitive and paradoxical. However, if we looked at the level of Nitrous oxide emissions in energy sector (thousand metric tons of CO2 equivalent) in 2020 for the various countries, we could notice that many countries that have high levels of regulatory quality also have high levels of emissions. this condition highlights the complexity of the application of ESG models in the context of the fight against climate change.

The positive relationship between Nitrous oxide emissions (metric tons of CO2 equivalent per capita) and Regulatory Quality: Estimate. World Bank data from 2020 show that several nations with strong regulatory quality—including New Zealand (2.93 tCO₂e/capita), Australia (1.92), Ireland (1.92), Canada (1.24), and Finland (1.06)—also report some of the highest per capita nitrous oxide (N₂O) emissions worldwide. Stronger governance and regulatory systems are usually linked with more efficient environmental protection, thus this favorable link seems contradictory. But, when one takes into account the larger background of industrial structure and economic growth, the link becomes more obvious. High regulatory quality countries usually have well defined legal institutions, administrative efficiency, and strong policy implementation tools. These qualities support large-scale agricultural development—sectors that are main sources of N₂O emissions—industrial modernization, economic expansion, and others. Particularly, nations like New Zealand and Ireland are quite dependent on fertilizer-intensive livestock farming, which greatly increases N2O emissions. So, good regulatory quality does not guarantee lower short-term emissions. Instead, it shows a state's ability to develop industrially and economically—usually leading to more emissions during the expansion stage. On the other hand, future environmental control, sustainable development changes, and the incorporation of green technologies all depend on same institutional power. Effective regulatory systems are more likely to put in place emissions monitoring, adopt mitigation policies, and harmonize national development with climate goals over time. Seen in this way, the reported favorable link could indicate a transition period in which regulatory capacity initially drives growth of emissions and later promotes long-term environmental control and sustainable development.(Bueno et al., 2022; Usman et al., 2023; Molden, 2023).

Q4. The parameter settings of the machine learning model, such as the neighborhood radius of density clustering and the k value of KNN, have not been fully explained, which may affect the reproducibility of the results. Suggest supplementing the selection criteria for algorithm hyperparameters and adding model robustness testing.

A4. In the appendix the hyperparameters of all the algorithms used have been reported. As for the validation, various statistical indicators have been used to evaluate the performances for both the regression and clustering machine learning algorithms.

Table xxx. Boosting Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	20 Sample % of all data
Data Split Preferences	Training and Validation Test	20% for validation data
Training Parameters	Shrinkage	0.1
	Interaction Depth	1
	Min Observation in node	10
	Training data used per tree	50
	Loss Function	Gaussian
	Scale Features	Yes
	Optimized max trees	100
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Decision Tree Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Min Observations of Split	20
	Min Observations in terminal	7
	Max Interaction Depth	30
	Scale Features	Yes
	Optimized	Max complexity penalty 1
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. K-Nearest Neighbors Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Weights	Rectangular
	Distance	Euclidian
	Scale Features	1
	Optimized	Max Nearest Neighbors 10
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Linear Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Training Parameters	Include Intercept	Yes
Training Parameters	Scale Features	Yes
Data Split	Train	1544
Data Split	Test	386

Table xxx. Neural Network Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and validation data	Sample 20% for validation data
Training Parameters	Activation Function	Logistic Sigmoid
	Algorithm	rprop+
	Stopping criteria loss function	1
	Max training repetitions	100000
	Scale Features	Yes
	Population size	20
	Generation	10
	Max number of layers	10
	Max nodes in each layer	10
	Parent selection	Roulette wheel
	Crossover method	Uniform
	Mutations	Reset
	Probability	10%
	Survival Method	Fitness based
	Elitism	10%
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Random Forest Regression Hyper parameter

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Training data used per tree	50%
	Features per split	Auto
	Scale Features	Yes
	Max Trees	100
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Regularized Linear Regression Hyper parameter

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Penalty	Lasso
	Include Intercept	Yes
	Scale Features	Yes
	Optimized	Yes
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Support Vector Machine Regression Hyper parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Weights	Linear
	Tolerance of termination criterion	0.001
	Epsilon	0.01
	Scale Features	Yes
	Max violation cost	5
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Density based clustering Hyper parameters

Training Parameters
Epsilon Neighborhood size	2
Min. core points	5
Distance	Normal
Scale Features	Yes

Table xxx. Fuzzy c-Means clustering Hyper parameters

Training Parameters
Max Iterations	25
Fuziness parameter	2
Scale Features	Yes
Optimized according to	BIG
Max clusters	10

Table xxx. Hierarchical Clustering Hyper parameters

Training Parameters
Distance	Euclidean
Linkage	Average
Scale features	Yes
Optimized According to	BIC
Max Clusters	10

Table xxx. Model based clustering Hyper parameters

Training Parameters
Model	Auto
Max Iterations	25
Scale features	Yes
Optimized According to	BIC
Max Clusters	10

Table xxx. Neighborhood-Based Clustering Hyper parameters

Training Parameters
Center type	Means
Algorithm	Hartigan-Wong
Max Iterations	25
Random sets	25
Scale features	Yes
Optimized according to	BIC
Max clusters	10

Table xxx. Random Forest Clustering Hyper parameters

Training Parameters
Trees	1000
Scale features	Yes
Optimized according to	BIC
Max Clusters	10

Q5. The presentation of charts is recommended to optimize their expression, as they are currently not easy to read.

A5. All hyperparameters of all algorithms have been placed in the appendix with detailed indication of the settings.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Title

The Title needs to be revised. It should reflect what is exactly this article focused on.

Introduction

The authors mention that, “While the literature on the emissions of greenhouse gases has been led by CO₂ and methane (CH₄), N₂O is relatively uncharted territory despite its high global warming potential and long-term contribution to climate change.” How can the authors justify this claim? Any statistical evidence and references? Without evidence support, it looks like a tall claim only.

There is no discussion on the criticalness of the phenomenon of interest. The authors directly jump into contribution of the paper as can be seen from line 44.

See from line 47 to 49, reproduced here, “While many studies have been done on the relationship between ESG performance and carbon emissions mitigation, few studies have made explicit reference to the consideration of the role of ESG factors in N₂O emissions on a global level.” Why it is important to study N₂O emissions in ESG context is missing.

The lack of discussion of rationale to conduct this study is absent. Merely comparison of two techniques does not warrant the article quality, it also had to frame the idea as to what extent it contributes to theoretical knowledge, empirical evidences and extend body of knowledge.

Similarly, the methodology section shows a table having a large number of variables/factors and there is no discussion in the introduction and literature review part as to why these are important to be examined.

There is not even a single citation in all the introduction section, which looks very weird to read.

Literature Review

The literature review needs, state-of-the-art way of writing it. In current form, it looks like summarizing and putting the articles in thesis way. There is not even a single section in order to distinguish the various concepts covered by the article.

The LR section does not clearly articulate on what is the literature gap and what needs to be focused.

What are the hypotheses?

Data and Methodology

The data period is not mentioned explicitly and what countries are covered in the analysis and what are excluded is missing. The rationale for excluding those economies is also missing.

How the study has chosen a specific set of variables to examine Environment, Social and governance. For example, how GDP can be used to examine governance? All these variables must be justified representing ESG or used as a proxy for estimating ESG.

The result and discussion is missing. Just interpreting the outcome of the analysis is not justified.

While the article mentions in the findings that developed nations have better energy efficiency and environmental governance, and also that they remain significant contributors to N₂O emissions due to intensive industry and agriculture. The question is that these developed economies are having a significant share in carbon emission until now, how they have better environmental governance? This must be justified.

The article is mentioning that N₂O emissions is significant in these economies and N₂O emissions is one of the factors of environmental pollution as CO₂ emissions, and statistics shows that CO₂ emissions has a greater share in the environmental pollution globally, there what is the rationale for conducting this study?

Conclusion

Without the above justifications, Conclusion part can be mis leading and therefore, I prefer to incorporate the above changes and then write this part.

Limitations of the study and future directions should be mentioned separately.

Comments for author File: Comments.pdf

Author Response

POINT TO POINT ANSWERS TO REVIEWER 3

Q0. The Title needs to be revised. It should reflect what is exactly this article focused on.

A0. The title has been changed as follows:

A Machine Learning and Panel Data Analysis of N₂O Emissions in an ESG Framework

Introduction

Q1. The authors mention that, “While the literature on the emissions of greenhouse gases has been led by CO₂ and methane (CH₄), N₂O is relatively uncharted territory despite its high global warming potential and long-term contribution to climate change.” How can the authors justify this claim? Any statistical evidence and references? Without evidence support, it looks like a tall claim only.

A1. We have added the following graph and sentences

Figure 1. Bibliometric Evidence of Research Imbalance: N₂O Marginalization in Sustainability Literature Compared to CO₂ and CH₄.

The graph in Figure 1 provides empirical data backing the assertion that, in relation to other significant greenhouse gases, nitrous oxide (N₂O) is still quite underrepresented in scholarly discourse oriented on sustainability. Based on a structured query run using the Scopus database, publications citing "CO₂" or "carbon dioxide" in conjunction with "sustainability" number 21,618. In contrast, 3,336 people talking about "CH₄" or "methane" and "sustainability"; records connected to "N₂O" or "nitrous oxide" and "sustainability" just total 1,175. Research output on N₂O and sustainability accounts for roughly 5.4% of the pertinent CO₂-focused literature and 35.2% of that on methane, indicating a distinct bibliometric disparity. Given that N₂O has a mean atmospheric lifetime over a century and a global warming potential (GWP) nearly 300 times that of CO₂ over a 100-year horizon, its marginalization in sustainability research is both unequal and concerning. These numbers back up the claim that N₂O remains a "uncharted territory" in the literature on sustainability and climate change. The relative dearth of academic engagement with N₂O, despite its well-documented environmental impact, underscores a crucial research gap. This underlines the significance of the present research, which aims to offer a more equitable and complete management of greenhouse gases under the environmental, social, and governance (ESG) and climate mitigation frameworks.

Q2. There is no discussion on the criticalness of the phenomenon of interest. The authors directly jump into contribution of the paper as can be seen from line 44.

A2. We have added the following sentences that show the criticalness of the phenomenon of interest

Given that N₂O has a mean atmospheric lifetime over a century and a global warming potential (GWP) nearly 300 times that of CO₂ over a 100-year horizon, its marginalization in sustainability research is both unequal and concerning. These numbers back up the claim that N₂O remains a "uncharted territory" in the literature on sustainability and climate change. The relative dearth of academic engagement with N₂O, despite its well-documented environmental impact, underscores a crucial research gap. This underlines the significance of the present research, which aims to offer a more equitable and complete management of greenhouse gases under the environmental, social, and governance (ESG) and climate mitigation frameworks

Q3. See from line 47 to 49, reproduced here, “While many studies have been done on the relationship between ESG performance and carbon emissions mitigation, few studies have made explicit reference to the consideration of the role of ESG factors in N₂O emissions on a global level.” Why it is important to study N₂O emissions in ESG context is missing.

A3. Studying nitrous oxide (N₂O) emissions within the Environmental, Social, and Governance (ESG) framework helps one to acquire a more comprehensive understanding of sustainability performance. A variety of connected processes—including fertilizer use, industrial agriculture, and land degradation—directly related to ESG factors interact to generate N₂O emissions. These emissions are influenced by environmental management practices, socioeconomic disparities, and the strength of institutional governance. Looking at N₂O in relation to ESG criteria helps to identify structural causes of emissions and provides a framework for linking environmental effects to policy effectiveness and social equity. Moreover, N₂O's absence from ESG studies undermines the robustness of emissions inventories, sustainability disclosures, and green investment criteria. As ESG criteria direct corporate strategy, financial decisions, and regulatory frameworks, the absence of N₂O consideration could result in insufficient risk assessments and missed chances for mitigation. Including N₂O into ESG models increases the precision of sustainability analysis and helps to shape focused, data-informed climate action.

Q4.The lack of discussion of rationale to conduct this study is absent. Merely comparison of two techniques does not warrant the article quality, it also had to frame the idea as to what extent it contributes to theoretical knowledge, empirical evidences and extend body of knowledge.

A4. We have added the following sentences:

The justification for this research is the urgent need to handle the underrepresentation of nitrous oxide (N₂O) in the larger conversation on greenhouse gas emissions and ESG-based sustainability assessment. Although N₂O—with its great global warming potential and significance to agriculture, energy, and waste systems—has been mostly ignored, current research has thoroughly investigated the influence of ESG variables in relation to carbon dioxide (CO₂) and methane (CH₄). This exclusion reduces the completeness of ESG evaluations and compromises the policy relevance of emissions modeling. Theoretically, this paper adds to the ESG literature by including N²O emissions as a necessary environmental dimension within a multidimensional framework connecting environmental, social, and governance indicators to global sustainability outcomes. It extends current models by thinking of ESG as a macro-level analytical tool for explaining emissions dynamics rather than just as a corporate measure. Empirically, the study uses both conventional econometric methods and sophisticated machine learning algorithms to find causal links, identify hidden patterns, and improve predictive capacity. By doing this, it provides new insights on the structural drivers of N₂O emissions, therefore presenting a more complex knowledge of country-level ESG profiles. This two-pronged strategy also improves methodological approaches in sustainability and climate research. The research increases the empirical basis and theoretical frontiers of ESG-climate interactions and offers data-driven tools for more focused and inclusive emission reduction plans to decision-makers.

Q5. Similarly, the methodology section shows a table having a large number of variables/factors and there is no discussion in the introduction and literature review part as to why these are important to be examined.

A5. The choice of variables was not discretionary but rather linked to the framework used, i.e. the World Bank database. In this regard, the following note has been added to the table relating to the data section “Note. Source: World Bank Sovereign ESG Data Portal. Link: https://esgdata.worldbank.org/?lang=en accessed 20/01/2025”

Q6. There is not even a single citation in all the introduction section, which looks very weird to read.

A7. We have added the following references to the introduction

Benghzial, K., Raki, H., Bamansour, S., Elhamdi, M., Aalaila, Y., & Peluffo-Ordóñez, D. H. (2023). GHG global emission prediction of synthetic N fertilizers using expectile regression techniques. Atmosphere, 14(2), 283.

Castle, J. L., & Hendry, D. F. (2020). Climate econometrics: An overview. Foundations and trends® in econometrics, 10(3-4), 145-322.

Costantiello, A., & Leogrande, A. (2023). The Impact of Research and Development Expenditures on ESG Model in the Global Economy.

Dhaliwal, J. K., Panday, D., Robertson, G. P., & Saha, D. (2025). Machine learning reveals dynamic controls of soil nitrous oxide emissions from diverse long‐term cropping systems (Vol. 54, No. 1, pp. 132-146).

Kanter, D. R., Ogle, S. M., & Winiwarter, W. (2020). Building on Paris: integrating nitrous oxide mitigation into future climate policy. Current Opinion in Environmental Sustainability, 47, 7-12.

Kollar, A. J. (2023). Bridging the gap between agriculture and climate: Mitigation of nitrous oxide emissions from fertilizers. Environmental Progress & Sustainable Energy, 42(2), e14069.

Piñeiro‐Guerra, J. M., Lewczuk, N. A., Della Chiesa, T., Araujo, P. I., Acreche, M., Alvarez, C., ... & Piñeiro, G. (2025). Spatial variability of nitrous oxide emissions from croplands and unmanaged natural ecosystems across a large environmental gradient.

Reay, D., & Reay, D. (2015). Nitrous Oxide as a Driver of Climate Change. Nitrogen and Climate Change: An Explosive Story, 39-47.

Samy, S., Jaini, K., & Preheim, S. (2024). A Novel Machine Learning-Driven Approach for Predicting Nitrous Oxide Flux in Precision Managed Agricultural Systems. Available at SSRN 4976901.

Szeląg, B., Zaborowska, E., & Mąkinia, J. (2023). An algorithm for selecting a machine learning method for predicting nitrous oxide emissions in municipal wastewater treatment plants. Journal of Water Process Engineering, 54, 103939.

Vasilaki, V., Conca, V., Frison, N., Eusebi, A. L., Fatone, F., & Katsou, E. (2020). A knowledge discovery framework to predict the N2O emissions in the wastewater sector. Water Research, 178, 115799.

Wang, R., Bian, Y., & Xiong, X. (2024). Impact of ESG preferences on investments and emissions in a DSGE framework. Economic Modelling, 135, 106731.

Literature Review

Q7. The literature review needs, state-of-the-art way of writing it. In current form, it looks like summarizing and putting the articles in thesis way. There is not even a single section in order to distinguish the various concepts covered by the article.

A7. The literature review has been rewritten as follows:

Table 1. Synthesis of the literature review.

ESG Pillar	Macro- Theme	Focus on N₂O Emissions	Key References
Environmental (E)	Agriculture, Land Use & Industry	N₂O emissions from fertilizer, livestock, biochar, wastewater, textile and energy sectors	Čapla et al. (2025), Padhi et al. (2024), Park et al. (2024), Cui et al. (2022), Harasheh & Harasheh (2021), Biswas et al. (2023, 2024), Lambiasi et al. (2024), Voicu (2023), Prieto (2022), An et al. (2022)
Environmental (E)	Monitoring & Environmental Technologies	N₂O prediction and tracking using digital tools, ML, blockchain, and urban ESG systems	Wang et al. (2024), Rafiee et al. (2022), Jiang et al. (2022), Muller (2021), Gruber (2021), Gu et al. (2024)
Social (S)	Social Equity, Accountability & Inclusion	Social factors in emissions management, SDGs, corporate ESG disclosures, and Scope 3 N₂O tracking	Ali et al. (2025), Yulianti et al. (2023), Squillace (2023), Stinchcombe (2023), Biswas et al. (2023), Agbo et al. (2024)
Governance (G)	Policy, Regulation & Institutional Capacity	Effectiveness of ESG-aligned policies, mandatory reporting, governance quality, climate compliance	Al-Sinan et al. (2023), Orsini (2022), Schuuring (2024), Turjak (2023), Blair (2021), Guermazi et al. (2025), Kaplan & Ramanna (2021), Jiang et al. (2022), Stinchcombe (2023)
Governance (G)	Risk Management & ESG Standards	ESG standardization, default risk for polluting firms, and transition readiness	Rothman (2023), Dennis & Iscan (2024), Kannoa (n.d.)
Finance (Cross-cutting)	ESG Investment, Green Finance & Disclosure	Use of green bonds, financial risk metrics, investor preferences, and carbon/N₂O footprint accountability	Sidestam & Karam (2024), Sacco et al. (2023), Yoshino & Yuyama (2021), Micol & Costa (2023), Çıtak & Meo (2024), Brühl (2021), Mamatzakis & Tzouvanas (2025), Boubaker et al. (2024), Bolton et al. (2022), Zhang et al. (2024)
Methodological (Cross-cutting)	Predictive Analytics & ESG Benchmarking	ESG index construction, forecasting tools, and integration of N₂O into ESG scoring systems	Drago & Leogrande (2024), Long & Feng (n.d.)

Q8. The LR section does not clearly articulate on what is the literature gap and what needs to be focused.

A8. We added the following paragraph:

Literature gap. Although the current research provides useful analysis of sector-specific N₂O emission sources—especially in agriculture, energy, and industry—and emphasizes the possibilities of ESG-oriented policy tools, it is still mostly scattered and thematically compartmentalized. Without methodically including ESG aspects into a unified explanatory framework for N₂O emissions, most studies examine environmental factors in isolation or within tightly defined sectors. Moreover, the empirical methods used are usually restricted to either qualitative evaluations or conventional econometric models, with little investigation of machine learning techniques to reveal non-linear patterns or improve predictive power. To date, no research has provided a thorough, cross-country study connecting macro-level ESG indicators to N₂O emissions; nor has current work jointly used panel data analysis and machine learning algorithms to assess the interaction between governance quality, environmental deterioration, social inequality, and emission levels. Furthermore, although increasing policy interest in ESG disclosures, there is still insufficient quantitative data on how national-level ESG performance metrics relate to and possibly control N₂O emissions. By creating an integrated methodological approach that combines conventional econometric modeling with sophisticated machine learning tools to evaluate the predictive and explanatory power of ESG indicators in shaping worldwide N₂O emission patterns, this study aims to close these important gaps.

Q9. What are the hypotheses?

A9. We have added the following paragraph:

Main hypothesis. We have added the following main hypothesis:

H₀: At the global level, ESG indicators do not significantly account for variation in N₂O emissions.
H₁ (Main Hypothesis): Environmental, Social, and Governance (ESG) indicators significantly account for cross-country variation in nitrous oxide (N₂O) emissions. Specifically, changes in ESG performance metrics are linked to observable variations in N₂O emissions per capita, and their predictive relevance can be confirmed using both econometric estimation and machine learning-based modeling.

Data and Methodology

Q10. The data period is not mentioned explicitly and what countries are covered in the analysis and what are excluded is missing. The rationale for excluding those economies is also missing.

A10. We have added the following appendix to show the 193 countries.

Appendix 2

The countries analyzed are listed below:

Niger, Somalia, Nigeria, Chad, Sierra Leone, Central African Republic, Guinea, Mali, Benin, Burkina Faso, Congo, Dem. Rep., Equatorial Guinea, Liberia, Cote d'Ivoire, Guinea-Bissau, Lesotho, Cameroon, Mozambique, Madagascar, Pakistan, Togo, Yemen, Rep., Haiti, Zambia, Afghanistan, Sudan, Djibouti, Burundi, Eswatini, Timor-Leste, Comoros, Zimbabwe, Kiribati, Gambia, The, Tanzania, Ethiopia, Ghana, Congo, Rep., Papua New Guinea, Lao PDR, Uganda, Malawi, Myanmar, Turkmenistan, Mauritania, Gabon, Rwanda, Namibia, Senegal, Eritrea, Kenya, South Sudan, Dominica, Botswana, Dominican Republic, South Africa, Tajikistan, India, Marshall Islands, Fiji, Guyana, Nauru, Bangladesh, Nepal, Bhutan, Philippines, Micronesia, Fed. Sts., Cambodia, St. Lucia, Bolivia, Iraq, Venezuela, RB, Vanuatu, Guatemala, Algeria, Syrian Arab Republic, Indonesia, Tuvalu, Vietnam, Egypt, Arab Rep., Solomon Islands, Azerbaijan, Paraguay, Morocco, Kyrgyz Republic, Suriname, Samoa, Honduras, Mauritius, Trinidad and Tobago, Tunisia, Grenada, Palau, Korea, Dem. People's Rep., Sao Tome and Principe, West Bank and Gaza, Mongolia, Jordan, St. Kitts and Nevis, Brazil, Moldova, Peru, Uzbekistan, Panama, Seychelles, St. Vincent and the Grenadines, Cabo Verde, Nicaragua, Bahamas, The, Mexico, Colombia, Iran, Islamic Rep., Ecuador, El Salvador, Jamaica, Barbados, Brunei Darussalam, Belize, Tonga, Libya, Armenia, Kazakhstan, Oman, Kosovo, Albania, Georgia, Turkiye, Kuwait, Thailand, Lebanon, Ukraine, Costa Rica, Malaysia, Argentina, Bahrain, China, Saudi Arabia, Sri Lanka, Chile, Romania, United Arab Emirates, Bulgaria, United States, Antigua and Barbuda, Maldives, Malta, Uruguay, Bosnia and Herzegovina, Slovak Republic, Serbia, North Macedonia, Qatar, Russian Federation, Canada, Cuba, New Zealand, Croatia, France, Poland, United Kingdom, Belgium, Netherlands, Hungary, Switzerland, Australia, Austria, Greece, Latvia, Denmark, Germany, Israel, Lithuania, Ireland, Portugal, Spain, Korea, Rep., Monaco, Andorra, Cyprus, Czech Republic, Belarus, Luxembourg, Iceland, Italy, Sweden, Japan, Montenegro, Finland, Norway, Slovenia, Singapore, Estonia, San Marino.

The analyzed period is reported in the subscripts of the equations Where i=193 and t=[2011;2020].

Countries with missing or incomplete time series have been excluded.

Q11. How the study has chosen a specific set of variables to examine Environment, Social and governance. For example, how GDP can be used to examine governance? All these variables must be justified representing ESG or used as a proxy for estimating ESG.

A11. As we indicated in the section on data and methodology, the choice of variables was not random but rather guided by the database used, namely the Sovereign ESG Data Portal https://esgdata.worldbank.org/?lang=en The aforementioned database divides the variables by attributing them to each of the three macro-categories of ESG. For this reason, it appears that GDP is part of the G component, namely because it is the database itself that establishes this attribution.

Q12. The result and discussion is missing. Just interpreting the outcome of the analysis is not justified.

A12. The section “Discussions of the results” has been added. This section also contains a summary of the results obtained with both the panel data analyses and the clustering and regressions performed with machine learning.

Discussions of the results

Macro	Variables	Panel Data Relationships	Best Clustering algorithm	Clusters	Best ML Algorithm	Machine Learning Results
E	ASFND	Positive	Density-Based	2	K-Nearest Neighbors	5.537×10^(+13)	Mean Dropout Loss
	EIPE	Negative				7.093×10^(+13)
	FA	Positive				5.537×10^(+13)
S	AGRI	Positive	Hierarchical clustering	8	Random Forest	6.167×10^(+13)
	FRT	Positive				7.710×10^(+13)
	GI	Negative				6.167×10^(+13)
	ISL20	Positive				6.167×10^(+13)
	WATER	Negative				7.471×10^(+13)
G	GDPG	Positive	Hierarchical Clustering	10	Random Forest	3.630×10^(+26)	Mean decrease in accuracy
	FMLP	Negative				2.076×10^(+27)
	RQE	Positive				2.195×10^(+27)
	RDE	Positive				1.385×10^(+27)
	STJA	Negative				2.171×10^(+27)
	SLRI	Negative				1.153×10^(+27)

Q13. While the article mentions in the findings that developed nations have better energy efficiency and environmental governance, and also that they remain significant contributors to N₂O emissions due to intensive industry and agriculture. The question is that these developed economies are having a significant share in carbon emission until now, how they have better environmental governance? This must be justified.

Q14. The article is mentioning that N₂O emissions is significant in these economies and N₂O emissions is one of the factors of environmental pollution as CO₂ emissions, and statistics shows that CO₂ emissions has a greater share in the environmental pollution globally, there what is the rationale for conducting this study?

A14. Indeed, the positive relationship between regulatory quality, taken as a measure of governance goodness, and N2O emissions may seem paradoxical. However, it has been better highlighted also in light of the data relating to N2O emissions, which demonstrate how many of the countries that have very high levels of institutional governance also have high levels of N2O emissions, an issue that highlights the need to carefully use ESG formulations to fully understand the complex relationships between the variables. To better understand this issue, the following propositions have been added to replace the previous ones.

Conclusion

Q15. Without the above justifications, Conclusion part can be mis leading and therefore, I prefer to incorporate the above changes and then write this part.

A15. The conclusions have been rewritten as follows.

By tackling a major empirical and conceptual gap—the insufficient integration of nitrous oxide (N₂O) emissions into macro-level ESG (Environmental, Social, and Governance) analysis—this work contributes to the sustainability literature. Although current studies mostly concentrate on CO2 and CH4, N2O—with its great global warming potential and long atmospheric lifetime—remains underrepresented in both academic and policy debate. This work responds to that gap by building a cross-country model that systematically evaluates how ESG indicators influence N₂O emissions, using both panel data econometrics and machine learning approaches.

The empirical findings show that ESG elements greatly account for cross-national variation in N₂O emissions. While social factors like income distribution and access to water also have notable impact, environmental indicators like forest area and energy intensity rise as main predictors. Variables related to governance, especially those on regulatory quality, show unexpected outcomes suggesting that robust institutional frameworks might first correspond with more emissions from industrial development—a result emphasizing the intricate interaction between governance and environmental consequences.
The study methodologically creates ESG-N₂O connections by using inferential and predictive models. Clustering methods show significant ESG-emissions patterns across nations, therefore supporting varied policy approaches. Using several machine learning techniques increases model robustness and emphasizes variable significance.

All things considered, the research shows that ESG indicators are not only pertinent but also analytically potent instruments for controlling and comprehending N₂O emissions. These ideas provide practical routes for ESG-integrated climate governance, increase empirical knowledge, and help to theoretical development.

Q16. Limitations of the study and future directions should be mentioned separately.

A16. The following section titled "Limitations and Future Research" has been added before the conclusions.

Although the research presents a unique combination of econometric and machine learning techniques to investigate N₂O emissions under an ESG framework, some drawbacks call for attention. First, especially for developing nations with weaker statistical infrastructures, the reliance on publicly available macro-level datasets, such those from the World Bank, could raise concerns about data completeness, granularity, and temporal alignment. This could have an impact on the accuracy of cross-country comparisons and the strength of conclusions reached. Second, although the econometric models control for unobserved heterogeneity using fixed and random effects, causal identification remains limited, and potential endogeneity between ESG variables and emissions outcomes cannot be fully ruled out. Though strong in predictive performance, the machine learning models do not offer causal justifications.

Future studies could investigate the combination of subnational or firm-level ESG measures to more precisely capture emission behavior dynamics. Temporal deep learning models or hybrid causal inference frameworks could also be included to more clearly separate the dynamic interactions between N₂O emissions and ESG indicators. Including life-cycle assessment data and Scope 3 emissions in the analysis would help to increase the applicability of ESG frameworks for sustainability accounting even more. Additionally, more work is needed to evaluate the role of specific ESG policy instruments—such as carbon pricing, green bonds, or disclosure mandates—on N₂O abatement across various governance regimes. Such extensions would improve the policy usefulness and explanatory power of ESG models in tackling global environmental issues.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The abstract starts directly from the purpose of the study. Starting with a motivation sentence before the purpose would be more academic and would emphasize the necessity of the study.

Why was N2O taken in the study, why not CO2?

The gap in the literature is explained in detail in the introduction section, but what will be the contribution of the study to the literature?

The literature is well written.

Why were these data selected? For example, "Ratio of female to male labor force participation rate" why is this data important and selected. Also, "Scientific and technical journal articles" why did you select this data?

What do ASFND, EIEP and FA stand for? Abbreviations should be clearly stated.

Policy recommendations can also be made in the conclusion section and should be concluded with the limitations of the study and possible future studies.

Author Response

POINT TO POINT ANSWERS TO REVIEWER 4

Q1. The abstract starts directly from the purpose of the study. Starting with a motivation sentence before the purpose would be more academic and would emphasize the necessity of the study.

A1. The abstract has been rewritten has follows:

Addressing climate change requires a deeper understanding of all greenhouse gases, yet nitrous oxide (N₂O)—despite its significant global warming potential—remains underrepresented in sustainability analysis and policy discourse. The paper examines N₂O emissions from an Environmental, Social, and Governance (ESG) standpoint with a combination of econometric and machine learning specifications to uncover global trends and policy implications. Results show the overwhelming effect of ESG factors on emissions, with intricate interdependencies between economic growth, resource productivity, and environmental policy. Econometric specifications identify forest degradation, energy intensity, and income inequality as the most significant determinants of N₂O emissions, which are in need of policy attention. Machine learning enhances predictive power insofar as emission drivers and country-specific trends are identifiable. Through the integration of panel data techniques and state-of-the-art clustering algorithms, the paper generates a highly differentiated picture of emission trends, separating country groups by ESG performance. The findings of the study are that while developed nations have better energy efficiency and environmental governance, they remain significant contributors to N₂O emissions due to intensive industry and agriculture. Meanwhile, developing economies with energy intensity have structural impediments to emissions mitigation. The paper also identifies the contribution of regulatory quality in emission abatement in that the quality of governance is found to be linked with better environmental performance. ESG-based finance instruments, such as green bonds and impact investing, also promote sustainable economic transition. The findings have the further implications of additional arguments for mainstreaming sustainability in economic planning, developing ESG frameworks to underpin climate targets.

Q2. Why was N2O taken in the study, why not CO2?

A2. We have written in the introduction why it is original and interesting study N2O:

Figure 1. Bibliometric Evidence of Research Imbalance: N₂O Marginalization in Sustainability Literature Compared to CO₂ and CH₄.

Q3. The gap in the literature is explained in detail in the introduction section, but what will be the contribution of the study to the literature?

A3. A subsection on the literature gap has been added within the literature review as follows:

Q4. The literature is well written.

A4. Thanks.

Q5. Why were these data selected? For example, "Ratio of female to male labor force participation rate" why is this data important and selected. Also, "Scientific and technical journal articles" why did you select this data?

A5. As we have indicated in the data and methodology section, these variables were not chosen at random. Rather, they were chosen consistently with the indications present in the database used, namely the Sovereign ESG Data Portal. This portal establishes the variables to be attributed to each of the three macro-categories, namely E, S, G. Specifically, at the following link it is possible to identify the classification based on the macro categories https://esgdata.worldbank.org/data/framework?lang=en

Q6. What do ASFND, EIEP and FA stand for? Abbreviations should be clearly stated.

A6. We have added the Appendix 3 with the list of abbreviations as follows:

Appendix 3 List of Abbreviations

Acronym	Definition	Acronym	Definition
N₂O	Nitrous Oxide	WATER	People using safely managed drinking water services (% of population)
ESG	Environmental, Social, and Governance	GDPG	GDP growth (annual %)
CO₂	Carbon Dioxide	FMLP	Ratio of female to male labor force participation rate (%) (modeled ILO estimate)
CH₄	Methane	RQE	Regulatory Quality: Estimate
GWP	Global warming potential	RDE	Research and development expenditure (% of GDP)
EU	European Union	STJA	Scientific and technical journal articles
E	Environmental	SLRI	Strength of legal rights index (0=weak to 12=strong)
S	Social	GDP	Gross Domestic Product
G	Governance	PPP	Purchasing Power Parity
NOE	Nitrous oxide emissions (metric tons of CO2 equivalent per capita)	ML	Machine Learning
ASNFD	Adjusted savings: net forest depletion (% of GNI)	MSE	Mean Squared Error
EIPE	Energy intensity level of primary energy (MJ/$2017 PPP GDP)	RMSE	Root Mean Squared Error
FA	Forest area (% of land area)	MAE	Mean Absolute Error
AGRI	Annualized average growth rate in per capita real survey mean consumption or income, total population (%)	MAD	Mean Absolute Deviation
FRT	Fertility rate, total (births per woman)	R²	Coefficient of Determination
GI	Gini index	R&D	Research and development
ISL20	Income share held by lowest 20%	ILO	International Labour Organization

Q7. Policy recommendations can also be made in the conclusion section and should be concluded with the limitations of the study and possible future studies.

A7. Two paragraphs have been added, one relating to policy recommendations and one relating to limitations and future research.

Limitations and Future Research

Policy recommendations

The results of this study highlight the pressing need to enhance the part of nitrous oxide (N₂O) within climate change mitigation and ESG-based environmental governance systems. Particularly in high-emission industries like agriculture, wastewater treatment, and energy generation, policymakers should include N₂O emissions explicitly into ESG reporting standards. This would improve the thoroughness of sustainability disclosures and match national climate goals with the whole range of greenhouse gas emissions. ESG benchmarks should also be customized to fit sector-specific dynamics, so encouraging the use of low-emission technologies like precision farming and enhanced soil management practices. Particularly in capital-intensive industries, financial tools like green bonds and ESG-aligned investment vehicles can be used to motivate N₂O reduction. The noted correlation between regulatory quality and increased N₂O emissions underlines, therefore, the necessity of institutional protections connecting economic growth with sustainability changes from the beginning. While public-private partnerships can help to deploy these technologies at scale, investments in machine learning and digital monitoring systems should be encouraged to enable predictive modeling and real-time emissions tracking. Climate policy has to include social equity aspects as well; it should guarantee that, particularly in nations with little access to fundamental infrastructure and services, emissions reduction does not aggravate inequality. At last, policy design should acknowledge the diversity of national profiles and guide different approaches and focused international assistance using ESG-emissions clustering. A more whole and data-driven approach to ESG policy is, therefore, absolutely necessary to handle the several problems of N²O emissions in the framework of sustainable development.

Author Response File: Author Response.pdf

Reviewer 5 Report

Comments and Suggestions for Authors

The manuscript investigate the impact of Environmental, Social and Governance (ESG) factors for global nitrous oxide (N₂O) emissions using a battery of econometric models including Pooled OLS, Fixed Effects, Random Effects, and Machine Learning (ML) techniques. Although this question is relevant and potentially useful, the research problem is poorly framed. “There’s no clearly articulated, testable hypothesis. This conceptual fuzziness makes it even harder for the reader to evaluate empirically whether and how the research generalizes or matters more generally.

The contribution to literature is twofold: one, integration of econometric and ML techniques in the analysis of N₂O emissions, which remains relatively under explored in macroeconomic and ESG framework compared to its carbon counterpart. Moreover, leveraging a compression dataset (World Bank) is commendable. But the contribution theoretically and empirically is undermined by a wooden approach to the ESG framework, which tends to be spoken in more than analyzed systematically. The discourse around ESG is broad and quantitatively weak, particularly in terms of the means through which the various analytical frameworks are implemented.

Although the hybrid approach that mixes econometrics and ML is hopeful, it is not novel. There are already several studies that apply ML to environmental and emissions data. What could have made this work novel — the application of ESG metrics to N₂O emissions — is not sufficiently developed. The manuscript does not provide a clear context to previous work nor does it strongly engage critically with relevant literature in environmental economics, the assessment of ESG impacts, or computational modeling. Hence, its contribution to the knowledge of the field is still very limited.

Validation of Predictive Models No-sturdy cross-validation techniques are brought to assess model performance. One should be careful not to reach unsupported claims concerning the superiority of some algorithms (e.g. k-nearest neighbours etc.) without proper validation.

Overfitting Problem: Though the data has a high dimensionality (190 countries across 10 years), the models—especially machine learning algorithms—are capable of generalising, but they are also prone to overfitting. Not even talked about, let alone handled.

Unjustified Model Choices: Models (random forest, k-NN, support vector machines, etc.) seem to be picked at random with little justification for why they should work on this dataset. The lack of any comparative framework or rationale means it feels like a series of descriptive sections without a great deal by way of methodological rigour.

Absence of Structural Controls: The analysis does not control for important structural heterogeneity, like region, income level, or economic development category. Such omission probably biases parameter estimates and compromises generalizability.

Restrained Interaction Effects: The authors do not investigate interactions between the ESG variables, thus losing potential insight on whether associations are synergistic or antagonistic.

Weak Cluster Validation: The clustering analysis does not provide a sensitivity or robustness check. The distance metrics and parameter choices are not sufficiently justified. A glaring cluster size imbalance also deserve attention.

The results reaffirm the conclusions in the manuscript but overstate the robustness and explanatory power of the models. Although the econometric and machine learning analyses do find associations between N₂O emissions and several ESG indicators in pooled data (such as forest depletion, energy intensity, income inequality), this is undermined by a lack of hypothesis testing or model validation, and control for confounders.

While this conclusion implies substantial support for ESG-based policies and ESG-finance mechanisms, they do not test the impact of specific ESG instruments empirically (e.g., green bonds, regulatory enforcement). That said, some of the conclusions are more speculative than data-driven.

Additionally, not every core questions brought up in the introduction is fully discussed. It does not delve into how ESG elements correlate with each other or their different potential impact across a country. Likewise, although clustering is discussed as a way to identify groups of countries, the policy implications of these clusters are underdeveloped.

The number of references in this manuscript is vast, quite few of them relevant and up to date. There has been little critical engagement with foundational literature including but not limited to the environmental economics literature and transferability metrics currently in use, and which relate to potential and actual machine learning applications for environmental modeling.

These citations are repetitive (e.g., multiple mentions of the same authors without clear separation) and some references mentioned are only implicitly related to N₂O, but only form tangential basis to the papers involved empirical core. Furthermore there is a strong reliance on very recent and sometimes obscure sources and little integration of well-established studies or review papers in the field.

In addition, important methodological references have been omitted regarding machine learning models validation, econometric restrictions in panel data, or clustering methods. More specifically, most works on the reference list could use curating, sharpening its focus and critical scouring in their selection.

Tables: There are just a ton of tables. In many cases — most glaringly in the results of the regression analyses (see e.g., Table 2 and Table 8) no important statistical indicators (e.g., adjusted R², VIF — multicollinearity — or post-estimation tests) are provided or are only weakly explained.

Figures: Clustering figures references (Figures 1 and 2) as-enumerated without interpretative depth. For example, cluster visualization can be aided with clearer labeling, titles on the axes and a better explanation of what the visual pattern actually indicates either for a policy action or empirical interpretation.

Data Quality: Although the data source (World Bank ESG database and others) is reliable, the manuscript lacks a detailed description of data cleaning methods, missing data treatment, and transformations performed. Also, no standard diagnostics of panel data (stationarity test, multicolinearity check, control of heteroscedasticity, etc.) are analyzed.

Cluster Evaluation Metrics: Clustering evaluation metrics such as silhouette scores or Dunn indices are appropriate, however, I would have liked to see a bit more discussion on what exactly the poor performance of some algorithms means for the robustness of the analysis.

Author Response

POINT TO POINT ANSWERS TO REVIEWER 5

Q1. The manuscript investigate the impact of Environmental, Social and Governance (ESG) factors for global nitrous oxide (N₂O) emissions using a battery of econometric models including Pooled OLS, Fixed Effects, Random Effects, and Machine Learning (ML) techniques. Although this question is relevant and potentially useful, the research problem is poorly framed. “There’s no clearly articulated, testable hypothesis. This conceptual fuzziness makes it even harder for the reader to evaluate empirically whether and how the research generalizes or matters more generally.

A1. We have added the main hypothesis as follows:

Main hypothesis. We have added the following main hypothesis:

H₀: At the global level, ESG indicators do not significantly account for variation in N₂O emissions.
H₁ (Main Hypothesis): Environmental, Social, and Governance (ESG) indicators significantly account for cross-country variation in nitrous oxide (N₂O) emissions. Specifically, changes in ESG performance metrics are linked to observable variations in N₂O emissions per capita, and their predictive relevance can be confirmed using both econometric estimation and machine learning-based modeling.

Q2. The contribution to literature is twofold: one, integration of econometric and ML techniques in the analysis of N₂O emissions, which remains relatively under explored in macroeconomic and ESG framework compared to its carbon counterpart. Moreover, leveraging a compression dataset (World Bank) is commendable. But the contribution theoretically and empirically is undermined by a wooden approach to the ESG framework, which tends to be spoken in more than analyzed systematically. The discourse around ESG is broad and quantitatively weak, particularly in terms of the means through which the various analytical frameworks are implemented.

A2. The ESG analytical model taken into consideration starts from the analysis of the World Bank database as indicated in the Data and Methodologies section. The idea of this article is to analyze in detail the impact of N₂O emissions within the individual ESG components i.e. E-Environmental, S-Social and G-Governance. This analytical need is highlighted by the fact that for each of these components there is a panel data, clustering and machine learning analysis. It is therefore an analysis that seeks to penetrate the ESG dynamics to grasp the most relevant aspects.

Q3. Although the hybrid approach that mixes econometrics and ML is hopeful, it is not novel. There are already several studies that apply ML to environmental and emissions data. What could have made this work novel — the application of ESG metrics to N₂O emissions — is not sufficiently developed. The manuscript does not provide a clear context to previous work nor does it strongly engage critically with relevant literature in environmental economics, the assessment of ESG impacts, or computational modeling. Hence, its contribution to the knowledge of the field is still very limited.

A3. The contribution to the literature that we wanted to make is not so much related to the technical dimension or to the use of integrated regression, clustering and machine learning tools although evidently such analysis presents a remarkable level of complexity. However, the contribution that we wanted to make is related to the analysis of N₂O emission in the ESG context as indicated in the following paragraph:

Literature gap. Although the current research provides useful analysis of sector-specific N₂O emission sources—especially in agriculture, energy, and industry—and emphasizes the possibilities of ESG-oriented policy tools, it is still mostly scattered and thematically compartmentalized. Without methodically including ESG aspects into a unified explanatory framework for N₂O emissions, most studies examine environmental factors in isolation or within tightly defined sectors. Moreover, the empirical methods used are usually restricted to either qualitative evaluations or conventional econometric models, with little investigation of machine learning techniques to reveal non-linear patterns or improve predictive power. To date, no research has provided a thorough, cross-country study connecting macro-level ESG indicators to N₂O emissions; nor has current work jointly used panel data analysis and machine learning algorithms to assess the interaction between governance quality, environmental deterioration, social inequality, and emission levels. Furthermore, although increasing policy interest in ESG disclosures, there is still insufficient quantitative data on how national-level ESG performance metrics relate to and possibly control N₂O emissions. By creating an integrated methodological approach that combines conventional econometric modelling with sophisticated machine learning tools to evaluate the predictive and explanatory power of ESG indicators in shaping worldwide N₂O emission patterns, this study aims to close these important gaps.

Q4. Validation of Predictive Models No-sturdy cross-validation techniques are brought to assess model performance. One should be careful not to reach unsupported claims concerning the superiority of some algorithms (e.g. k-nearest neighbours etc.) without proper validation.

A4. Standard k-fold cross-validation was not practical given the panel structure of the dataset–193 countries over 10 years—without compromising temporal or country-level independence. A planned three-way data split (training, validation, and test sets) was thus used to guarantee that observations were not randomly spread across time and space. Model hyperparameters were optimized using the validation set; the test set was set aside for last model assessment. Data on hyperparameters are in the appendix as follows:

Appendix

Table xxx. Boosting Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	20 Sample % of all data
Data Split Preferences	Training and Validation Test	20% for validation data
Training Parameters	Shrinkage	0.1
	Interaction Depth	1
	Min Observation in node	10
	Training data used per tree	50
	Loss Function	Gaussian
	Scale Features	Yes
	Optimized max trees	100
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Decision Tree Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Min Observations of Split	20
	Min Observations in terminal	7
	Max Interaction Depth	30
	Scale Features	Yes
	Optimized	Max complexity penalty 1
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. K-Nearest Neighbors Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Weights	Rectangular
	Distance	Euclidian
	Scale Features	1
	Optimized	Max Nearest Neighbors 10
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Linear Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Training Parameters	Include Intercept	Yes
Training Parameters	Scale Features	Yes
Data Split	Train	1544
Data Split	Test	386

Table xxx. Neural Network Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and validation data	Sample 20% for validation data
Training Parameters	Activation Function	Logistic Sigmoid
	Algorithm	rprop+
	Stopping criteria loss function	1
	Max training repetitions	100000
	Scale Features	Yes
	Population size	20
	Generation	10
	Max number of layers	10
	Max nodes in each layer	10
	Parent selection	Roulette wheel
	Crossover method	Uniform
	Mutations	Reset
	Probability	10%
	Survival Method	Fitness based
	Elitism	10%
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Random Forest Regression Hyper parameter

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Training data used per tree	50%
	Features per split	Auto
	Scale Features	Yes
	Max Trees	100
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Regularized Linear Regression Hyper parameter

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Penalty	Lasso
	Include Intercept	Yes
	Scale Features	Yes
	Optimized	Yes
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Support Vector Machine Regression Hyper parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Weights	Linear
	Tolerance of termination criterion	0.001
	Epsilon	0.01
	Scale Features	Yes
	Max violation cost	5
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Density based clustering Hyper parameters

Training Parameters
Epsilon Neighborhood size	2
Min. core points	5
Distance	Normal
Scale Features	Yes

Table xxx. Fuzzy c-Means clustering Hyper parameters

Training Parameters
Max Iterations	25
Fuziness parameter	2
Scale Features	Yes
Optimized according to	BIG
Max clusters	10

Table xxx. Hierarchical Clustering Hyper parameters

Training Parameters
Distance	Euclidean
Linkage	Average
Scale features	Yes
Optimized According to	BIC
Max Clusters	10

Table xxx. Model based clustering Hyper parameters

Training Parameters
Model	Auto
Max Iterations	25
Scale features	Yes
Optimized According to	BIC
Max Clusters	10

Table xxx. Neighborhood-Based Clustering Hyper parameters

Training Parameters
Center type	Means
Algorithm	Hartigan-Wong
Max Iterations	25
Random sets	25
Scale features	Yes
Optimized according to	BIC
Max clusters	10

Table xxx. Random Forest Clustering Hyper parameters

Training Parameters
Trees	1000
Scale features	Yes
Optimized according to	BIC
Max Clusters	10

Q5. Overfitting Problem: Though the data has a high dimensionality (190 countries across 10 years), the models—especially machine learning algorithms—are capable of generalising, but they are also prone to overfitting. Not even talked about, let alone handled.

A5. As in the previous answer, you can check that the appendix indicates the strategies to combat overfitting, i.e. data split and automatic optimization of hyperparameters.

Table xxx. Boosting Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	20 Sample % of all data
Data Split Preferences	Training and Validation Test	20% for validation data
Training Parameters	Shrinkage	0.1
	Interaction Depth	1
	Min Observation in node	10
	Training data used per tree	50
	Loss Function	Gaussian
	Scale Features	Yes
	Optimized max trees	100
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Decision Tree Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Min Observations of Split	20
	Min Observations in terminal	7
	Max Interaction Depth	30
	Scale Features	Yes
	Optimized	Max complexity penalty 1
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. K-Nearest Neighbors Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Weights	Rectangular
	Distance	Euclidian
	Scale Features	1
	Optimized	Max Nearest Neighbors 10
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Linear Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Training Parameters	Include Intercept	Yes
Training Parameters	Scale Features	Yes
Data Split	Train	1544
Data Split	Test	386

Table xxx. Neural Network Regression Hyper-Parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and validation data	Sample 20% for validation data
Training Parameters	Activation Function	Logistic Sigmoid
	Algorithm	rprop+
	Stopping criteria loss function	1
	Max training repetitions	100000
	Scale Features	Yes
	Population size	20
	Generation	10
	Max number of layers	10
	Max nodes in each layer	10
	Parent selection	Roulette wheel
	Crossover method	Uniform
	Mutations	Reset
	Probability	10%
	Survival Method	Fitness based
	Elitism	10%
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Random Forest Regression Hyper parameter

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Training data used per tree	50%
	Features per split	Auto
	Scale Features	Yes
	Max Trees	100
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Regularized Linear Regression Hyper parameter

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Penalty	Lasso
	Include Intercept	Yes
	Scale Features	Yes
	Optimized	Yes
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Support Vector Machine Regression Hyper parameters

Data Split Preferences	Holdout Test Data	Sample 20% of all data
Data Split Preferences	Training and Validation Data	Sample 20% for validation data
Training Parameters	Weights	Linear
	Tolerance of termination criterion	0.001
	Epsilon	0.01
	Scale Features	Yes
	Max violation cost	5
Data Split	Train	1235
	Validation	309
	Test	386

Table xxx. Density based clustering Hyper parameters

Training Parameters
Epsilon Neighborhood size	2
Min. core points	5
Distance	Normal
Scale Features	Yes

Table xxx. Fuzzy c-Means clustering Hyper parameters

Training Parameters
Max Iterations	25
Fuziness parameter	2
Scale Features	Yes
Optimized according to	BIG
Max clusters	10

Table xxx. Hierarchical Clustering Hyper parameters

Training Parameters
Distance	Euclidean
Linkage	Average
Scale features	Yes
Optimized According to	BIC
Max Clusters	10

Table xxx. Model based clustering Hyper parameters

Training Parameters
Model	Auto
Max Iterations	25
Scale features	Yes
Optimized According to	BIC
Max Clusters	10

Table xxx. Neighborhood-Based Clustering Hyper parameters

Training Parameters
Center type	Means
Algorithm	Hartigan-Wong
Max Iterations	25
Random sets	25
Scale features	Yes
Optimized according to	BIC
Max clusters	10

Table xxx. Random Forest Clustering Hyper parameters

Training Parameters
Trees	1000
Scale features	Yes
Optimized according to	BIC
Max Clusters	10

Q6. Unjustified Model Choices: Models (random forest, k-NN, support vector machines, etc.) seem to be picked at random with little justification for why they should work on this dataset. The lack of any comparative framework or rationale means it feels like a series of descriptive sections without a great deal by way of methodological rigour.

A6. We have added the following sentences to the text in the data and methodological section:

Scientifically relevant to the research question is the use of a varied collection of machine learning regression and clustering algorithms since the aim is both explanatory and predictive: to reveal the complex, nonlinear, and possibly latent interactions between ESG-related macro-variables and N₂O emissions at the global level. Particularly appropriate for high-dimensional, non-linear data settings are regression algorithms like boosting, random forest, support vector machine, and K-nearest neighbors, which let the model catch complex patterns that conventional linear econometric models could overlook. Furthermore, the addition of regularized linear models and neural networks enables strong comparisons across various complexity levels and bias-variance tradeoffs, therefore guaranteeing that results are not the result of overfitting or model-specific artifacts. From the clustering perspective, algorithms such as hierarchical, density-based, and fuzzy C-means clustering allow for the grouping of nations according to comparable ESG profiles and emission patterns, therefore exposing latent structures hidden in regression-only settings. Clustering helps to achieve the goal of the study by helping to create policy-relevant groupings of nations beneficial for differentiated sustainability strategies. Aiming method choice to the multidimensional character of ESG and the multifactorial beginnings of N₂O emissions, this multi-method framework increases the empirical depth and theoretical consistency of the research. Model interpretability—through decision trees and feature importance scores—and predictive performance—through support vector machines and ensemble methods—guarantee that both scientific knowledge and practical relevance are covered.

Q7. Absence of Structural Controls: The analysis does not control for important structural heterogeneity, like region, income level, or economic development category. Such omission probably biases parameter estimates and compromises generalizability.

A7. We appreciate the reviewer’s observation concerning the omission of structural controls such as regional classification, income level, and economic development status. We fully recognize that these macro-structural dimensions can influence both the ESG profiles and the emissions dynamics of countries, potentially introducing omitted variable bias into parameter estimates and affecting generalizability. To address this, we clarify that our dataset includes a wide cross-section of countries representing various geographic regions and development stages, and the clustering algorithms applied in the study were designed to uncover latent groupings that, to some extent, reflect these structural differences. For example, the unsupervised clustering of countries based on ESG and N₂O indicators led to differentiated profiles that are interpretable in terms of economic development and governance performance, even if these groupings were not explicitly defined by region or income. Nonetheless, we acknowledge that including structural fixed effects—such as World Bank income classifications or regional dummies—could have helped isolate the effect of ESG indicators from broader macroeconomic patterns. In the revised version, we now include a brief discussion of this limitation in the “Limitations and Future Research” section and recommend that future work incorporate these controls directly into econometric models, or stratify models by income groups to improve causal inference and cross-group comparability. This refinement will further enhance the robustness and external validity of ESG-based emission modelling frameworks. In effect the “Limitations and Future Research” has been renewed has followed:

Limitations and Future Research

Though the study offers a novel mix of econometric and machine learning methods to examine N₂O emissions within an ESG framework, certain shortcomings merit consideration. First, especially for developing countries with poorer statistical infrastructures, the reliance on publicly available macro-level datasets, such as those from the World Bank, could raise questions about data completeness, granularity, and temporal alignment. The precision of cross-country comparisons and the robustness of drawn conclusions could be affected by this. Second, while the econometric models use fixed and random effects to control for unobserved heterogeneity, causal identification stays restricted and possible endogeneity between ESG variables and emissions results cannot be completely excluded. Although the machine learning models are good in predictive performance, they lack causal justifications. Third, the study does not explicitly control for structural country-level traits including geographic area, income classification, or development category. Excluding these macro-structural factors could create omitted variable bias and reduce the generalizability of parameter estimates across several socio-economic settings. Although the clustering algorithms show country groupings that sometimes fit development levels, future studies might improve robustness by adding such controls in econometric models or stratifying analyses appropriately.

Future research could look at the mix of subnational or corporate-level ESG policies to more precisely reflect emission behavior dynamics. Including temporal deep learning models or hybrid causal inference systems would help to more clearly distinguish the dynamic interactions between N₂O emissions and ESG indicators. Examining life-cycle assessment data and Scope 3 emissions would further make ESG frameworks for sustainability accounting more relevant. Furthermore, more study is required to assess the influence of particular ESG policy tools—such as carbon pricing, green bonds, or disclosure obligations—on N²O reduction under several government systems. Such extensions would increase the policy relevance and explanatory power of ESG models in addressing global environmental issues.

Q8. Restrained Interaction Effects: The authors do not investigate interactions between the ESG variables, thus losing potential insight on whether associations are synergistic or antagonistic.

A8. We have also added these considerations to Limitations and Future Research as follows:

Though the research presents a unique combination of econometric and machine learning techniques to investigate N₂O emissions under an ESG framework, some limitations deserve attention. First, especially for developing nations with weaker statistical infrastructures, the reliance on publicly available macro-level datasets, such as those from the World Bank, could call into question data completeness, granularity, and temporal alignment. This could influence the accuracy of cross-country comparisons as well as the strength of drawn conclusions. Second, although the econometric models control for unobserved heterogeneity using fixed and random effects, causal identification stays limited and complete exclusion of possible endogeneity between ESG variables and emissions outcomes is not possible. Though the machine learning models provide good predictive performance, they lack causal justifications and cannot completely clarify the fundamental processes propelling emissions patterns. Third, the research does not specifically control for structural country-level characteristics such geographic area, income classification, or development category. Ignoring these macro-structural elements could create missing variable bias and limit the generalizability of parameter estimates across different socio-economic contexts. Though the clustering algorithms show country groupings that occasionally correspond with development levels, future research could increase robustness by including such controls in econometric models or stratifying analyses appropriately.

Furthermore, although the thoroughness of this study, one significant drawback is the lack of interaction terms among ESG variables in both econometric and machine learning models. The ESG framework is naturally multidimensional, and its components—environmental, social, and governance indicators—often show synergistic or opposing interactions that could enhance or cancel their separate effects on N₂O emissions. The study might miss compound effects that could provide more profound explanatory power or policy-relevant insights by treating ESG factors in isolation. Strong governance, for example, might improve the efficacy of environmental policy; social equity could offset the environmental impact of economic expansion. Future studies should investigate the inclusion of interaction effects, perhaps via structural equation modeling, interaction terms in fixed effects regressions, or sophisticated machine learning models able to capture non-linear feature interactions. Considering these interaction dynamics would enhance the explanatory strength of ESG models and help to clarify the channels by which ESG aspects together affect emission results.

Future studies could also investigate the integration of subnational or corporate-level ESG policies to more precisely reflect emission behavior dynamics. Including hybrid causal inference systems or temporal deep learning models would help to more clearly separate the dynamic interactions between N²O emissions and ESG indicators. Looking at Scope 3 emissions and life-cycle assessment data would help to make ESG models more relevant for sustainable accounting even more. Moreover, more study is needed to evaluate how certain ESG policy tools—such as carbon pricing, green bonds, or disclosure obligations—affect N₂O reduction under several governance systems. Such extensions would enhance the policy relevance and explanatory power of ESG models in tackling global environmental challenges.

Q9. Weak Cluster Validation: The clustering analysis does not provide a sensitivity or robustness check. The distance metrics and parameter choices are not sufficiently justified. A glaring cluster size imbalance also deserve attention.

A9. We have also added the critique to the section limitations and future research as follows

Although the study offers a novel mix of econometric and machine learning methods to analyze N₂O emissions within an ESG framework, certain drawbacks merit consideration. First, especially for developing countries with less robust statistical systems, the reliance on publicly available macro-level datasets, such as those from the World Bank, could raise doubts about data completeness, granularity, and temporal alignment. This could affect the strength of drawn conclusions as well as the correctness of cross-country comparisons. Second, while the econometric models use fixed and random effects to control for unobserved heterogeneity, causal identification stays restricted and complete exclusion of potential endogeneity between ESG variables and emissions results is not feasible. Although the machine learning models have good predictive performance, they cannot completely capture the structural mechanisms driving N₂O emission trends and they do not provide causal interpretations. Third, the study does not explicitly control for structural country-level traits including geographic area, income group, or development category. Ignoring these macro-structural aspects could create missing variable bias and limit the generalizability of parameter estimates across different socio-economic settings. Though the clustering techniques show country groupings that somewhat reflect development levels, future research should either use stratified analysis or add such controls straight into econometric models to improve robustness.

Moreover, a major drawback of this thorough research is the lack of interaction terms among ESG variables in both econometric and machine learning models. Inherently multidimensional, the ESG framework's pillars—environmental, social, and governance—can show synergistic or opposing interactions that affect emissions results in complex ways. Examining ESG factors in isolation could cause the study to miss higher-order consequences that could improve policy relevance and explanatory richness. For instance, strong institutions might amplify the influence of environmental policy; social fairness could offset the environmental consequences of economic growth. Future studies should investigate these interactive dynamics using structural equation modeling, fixed-effects interaction terms, or sophisticated machine learning methods able to seize non-linear relationships. Considering these relationships would greatly increase the theoretical consistency and explanatory power of ESG-based emission models.

Another issue has to do with the clustering technique. Using density-based, fuzzy C-means, hierarchical, model-based, neighborhood-based, and random forest clustering, the paper takes a comparative, multi-algorithm approach. This decision shows the diversity of the data set and the intricate structure of ESG-emissions links across nations. We evaluated every clustering solution using a consistent set of performance measures—including silhouette score, Calinski–Harabasz index, Dunn index, entropy, and Pearson's γ—to guarantee internal validation. These indices helped to assess cluster cohesion as well as separation, therefore guiding the choice of the best-performing algorithm. Still, we admit that more sensitivity testing—such as parameter perturbation or distance metric variation—would improve the strength of findings. We also understand that certain clustering results showed notable cluster size imbalance, a characteristic that could arise from the underlying distribution of ESG attributes. Future studies should handle this by means of parameter tuning, external validation employing income or regional classifications, and visualisation tools to enhance interpretability and generalizability.

At last, more research could combine subnational or company-level ESG data to more accurately reflect emission behaviour at more detailed levels. Including hybrid causal inference systems or temporal deep learning models would help to better identify dynamic links between N₂O emissions and ESG indicators. Including life-cycle assessment data and Scope 3 emissions in the study would help to increase the relevance of ESG-based systems for sustainability accounting even more. Examining the efficacy of particular ESG policy tools—such as carbon pricing, green bonds, or disclosure obligations—across various governance systems would also help to improve the empirical strength and policy use of ESG-climate models.

Q10. The results reaffirm the conclusions in the manuscript but overstate the robustness and explanatory power of the models. Although the econometric and machine learning analyses do find associations between N₂O emissions and several ESG indicators in pooled data (such as forest depletion, energy intensity, income inequality), this is undermined by a lack of hypothesis testing or model validation, and control for confounders.

A10. We appreciate the reviewer's significant remark and share their view that validating empirical assertions depends on model robustness and control for confounders. The study, let me be clear, does have formal hypothesis testing in the econometric framework. In the fixed and random effects regressions (see Section 4.1 and 4.2), joint significance tests and Hausman specification tests were used to support the relevance and consistency of the ESG variables in accounting for N₂O emissions. With respect to model validation, the paper integrates several econometric and machine learning techniques, so enabling triangulation of findings across analytical tools. As described in Tables 3, 4, and 9, the clustering algorithms were assessed using internal validation criteria including the Dunn index, Calinski-Harabasz index, and silhouette coefficients, so providing robustness.
Still, we recognize the need for more control of possible confounders, particularly structural heterogeneity including area, development level, and income group, which could generate omitted variable bias. The updated Limitations and Future Research part has handled this restriction. Future work will combine hierarchical modeling, stratified analysis, and region-specific interaction terms to better account for latent structural effects and increase generalizability. Although the present findings are noteworthy and directionally consistent with theory, we believe that adding more granular control variables and cross-validation methods would strengthen the explanatory claims even more.

Q11. While this conclusion implies substantial support for ESG-based policies and ESG-finance mechanisms, they do not test the impact of specific ESG instruments empirically (e.g., green bonds, regulatory enforcement). That said, some of the conclusions are more speculative than data-driven.

A11. These considerations have been integrated in the policy recommendations as follows:

The results of this study underline the pressing need to enhance the integration of nitrous oxide (N²O) issues into climate mitigation plans and ESG-based environmental governance systems. Policymakers should aim toward including clear N₂O reporting criteria into ESG disclosure standards since N₂O emissions in high-impact industries including agriculture, energy, and wastewater management contribute much. Doing so would improve the completeness of sustainability reporting and match national climate pledges with a more whole accounting of greenhouse gas emissions. ESG criteria have to be customized to fit sector-specific emission dynamics as well, so encouraging the use of low-N₂O technologies including process optimization in wastewater treatment, better soil management, and precision agriculture.

In sectors capital-intensive, financial instruments such as green bonds and ESG-aligned investment portfolios show promise as catalysts for emission reduction. Although this paper emphasizes their importance within a larger ESG-finance approach, it does not scientifically assess the causal influence of particular tools including bond issuance or regulatory enforcement, it does not scientifically assess the causal influence of particular tools including bond issuance or regulatory enforcement. Support for these tools is therefore conceptual and based on agreement with observed macro-level correlations between N₂O emissions and ESG indicators. Using granular financial or firm-level data, quasi-experimental techniques, and policy-specific evaluation frameworks, future studies should investigate these mechanisms more directly.

Moreover, the noted link between regulatory quality and higher N₂O emissions emphasizes the need of including sustainability into the institutional framework of economic growth. Supported by capacity building and policy coordination, regulatory frameworks have to develop beyond compliance toward enabling systematic transformation. While investments in digital infrastructure and machine learning tools should be scaled to enable predictive monitoring and real-time emissions tracking, public-private partnerships can help to speed the deployment of mitigation technologies.

Policy design has to include social equity at last so that countries with limited access to infrastructure and services do not suffer more from mitigation activities. Differentiated policy design and focused international collaboration are well-founded on ESG-informed clustering of national profiles. Thus, addressing the multi-dimensional problems of N²O reduction within the larger sustainable development goal requires a more integrated, evidence-based ESG policy framework.

Q12. Additionally, not every core questions brought up in the introduction is fully discussed. It does not delve into how ESG elements correlate with each other or their different potential impact across a country. Likewise, although clustering is discussed as a way to identify groups of countries, the policy implications of these clusters are underdeveloped.

A12. These considerations have been integrated in the policy recommendations as follows:

The findings of this study emphasize the urgent need to improve the integration of nitrous oxide (N²O) concerns into climate mitigation strategies and ESG-based environmental governance systems. Given the major N₂O emissions in high-impact industries like agriculture, energy, and wastewater management, policymakers should strive for the inclusion of clear N₂O reporting criteria within ESG disclosure standards. Including these emissions into uniform sustainability reporting would improve the thoroughness of greenhouse gas accounting and more closely match national climate promises with real environmental performance. ESG benchmarks should also be changed to reflect sector-specific emission dynamics, so promoting the use of low-NâO technologies including process optimization in wastewater treatment, better soil and fertilizer management, and precision agriculture.

In industries capital-intensive, financial instruments such as green bonds and ESG-aligned investment portfolios offer interesting ways to motivate emission reduction. Although this paper highlights their possibilities as part of a more general ESG-finance strategy, it does not scientifically evaluate the causal influence of particular instruments including green bond issuance or regulatory enforcement. The support shown in this paper is therefore conceptual, based on observed macro-level correlations between ESG factors and N₂O results. Future studies should directly look at how well these tools work by using firm-level or transaction-level data and applying causal inference methods including quasi-experimental designs or policy evaluation frameworks.

Moreover, the noted link between better regulatory quality and higher N₂O emissions emphasizes the need of including sustainability objectives early into the institutional framework of economic development. Supported by robust institutions, policy consistency, and capacity building, regulatory frameworks have to change from enforcement and compliance toward enabling systematic sustainability transitions. Investments in digital monitoring systems—including machine learning for predictive analytics and real-time tracking—should be scaled to enhance both governance and transparency. Public-private partnerships can be quite important in hastening the large-scale deployment of these inventions.

From a policy design standpoint, it is also vital to include ideas of social equity, therefore guaranteeing that efforts to reduce emissions do not disproportionately affect low-income people or nations with restricted access to infrastructure and services. Furthermore, although the clustering analysis done in this study revealed different groups of nations depending on shared ESG-emission profiles, the relevant policy consequences should be more operationalized. For instance, while those in high-governance, high-emission clusters may need to concentrate on sectoral reform and technological innovation, countries in high-emission, low-governance clusters might gain from targeted international support and institutional strengthening. ESG-informed clustering therefore offers a data-driven basis for varied policy responses and international cooperation customized to particular national settings.

Eventually, tackling the several N²O reduction issues inside the larger context of sustainable development will call for a more integrated, evidence-based ESG policy framework that not only directs reporting and finance but also organizes governance, innovation, and worldwide cooperation.

Q13. The number of references in this manuscript is vast, quite few of them relevant and up to date. There has been little critical engagement with foundational literature including but not limited to the environmental economics literature and transferability metrics currently in use, and which relate to potential and actual machine learning applications for environmental modeling.

A13. The writers emphasize the need of anchoring the study in both basic and modern literature from environmental economics and environmental modeling. Although the paper already contains a sizable and varied reference list—many of which directly address nitrous oxide (N₂O) emissions, ESG frameworks, and recent developments in machine learning for sustainability modeling—the addition of more classical and methodologically rigorous sources from environmental economics and statistical modeling would surely enhance the theoretical scaffolding of the work. Understanding the policy relevance of ESG metrics depends on seminal works and methodological frameworks in environmental economics, such as the role of externalities, cost-benefit analysis in pollution control, and the valuation of ecosystem services. The revised manuscript makes efforts to more clearly engage with these works. The methodological decisions of the study have also been contextualized by including more literature on transferability metrics—such as model portability across sectors or regions—and foundational works in climate econometrics (e.g., Castle and Hendry, 2020) and econometric identification techniques. Furthermore, by citing research that has used ML to GHG emission forecasting, ecosystem behavior, and agri-environmental systems (e.g., Dhaliwal et al., 2025; Vasilaki et al., 2020), the machine learning element has been more appropriately placed within the larger field of environmental modeling. These changes tackle the issue of low theoretical depth by bridging the gap between predictive modeling and environmental science. The writers value the reviewer's recommendation, which has inspired a more focused integration of multidisciplinary frameworks supporting the empirical and theoretical contributions of the work. This guarantees the paper is not only methodologically robust but also well-aligned with accepted research practices in environmental policy assessment and sustainability science.

Q14. These citations are repetitive (e.g., multiple mentions of the same authors without clear separation) and some references mentioned are only implicitly related to N₂O, but only form tangential basis to the papers involved empirical core. Furthermore there is a strong reliance on very recent and sometimes obscure sources and little integration of well-established studies or review papers in the field.

A14. We appreciate the reviewer's careful comments on citation quality, repetition, and the balance between foundational and recent sources. Though contextually relevant to ESG or climate governance, we admit the original paper had many occurrences of repeated citations and references that were only indirectly related to nitrous oxide (N₂O) emissions. The updated version of the paper has now addressed these problems carefully. We systematically refined the literature review to (1) remove redundant mentions of the same authors unless substantively different contributions were cited, (2) give priority to references with direct empirical or theoretical engagement with N₂O emissions within environmental economics, ESG governance, or machine learning applications, and (3) balance recent empirical studies with more established, peer-reviewed literature. Where relevant, we have substituted niche or narrowly focused references with more general review papers and foundational works in environmental modeling (e.g., IPCC assessment reports, OECD policy frameworks, and classical environmental economics studies), so ensuring that the conceptual and empirical scaffolding of the paper is in line with mainstream academic conversation.
Moreover, we increased interaction with important review papers in environmental governance, emissions modeling, and ESG finance to support theoretical foundation. Aiming to improve academic rigor, thematic focus, and the credibility of the paper's empirical contributions, these changes show up all over the updated manuscript—especially in the Introduction and Literature Review sections.

Q15. In addition, important methodological references have been omitted regarding machine learning models validation, econometric restrictions in panel data, or clustering methods. More specifically, most works on the reference list could use curating, sharpening its focus and critical scouring in their selection.

A15. The following references have been added to the methodological section:

Addington, O., Zeng, Z. C., Pongetti, T., Shia, R. L., Gurney, K. R., Liang, J., ... & Sander, S. P. (2021). Estimating nitrous oxide (N2O) emissions for the Los Angeles Megacity using mountaintop remote sensing observations. Remote Sensing of Environment, 259, 112351.

Best, R., Nazifi, F., & Cheng, H. (2024). Carbon Pricing Impacts on Four Pollutants: A Cross-Country Analysis. Energies, 17(11), 2596.

Bourel, M., Cugliari, J., Goude, Y., & Poggi, J. M. (2024). Boosting diversity in regression ensembles. Statistical Analysis and Data Mining: The ASA Data Science Journal, 17(1), e11654.

Cen, X., Müller, C., Kang, X., Zhou, X., Zhang, J., Yu, G., & He, N. (2024). Nitrogen deposition contributed to a global increase in nitrous oxide emissions from forest soils. Communications Earth & Environment, 5(1), 532.

Dhanoa, M. S., Sanderson, R., Lister, S. J., Cardenas, L. M., Ellis, J. L., López, S., & France, J. (2024). Decision tree learning with random forest models using agricultural and ecological field data incorporating multi-factor studies and covariate structure. CABI Reviews, 19(1).

Dradra, Z., & Abdennadher, C. (2024). Modeling the contribution of energy consumption to climate change: A panel cointegration analysis for mediterranean countries. Journal of the Knowledge Economy, 15(1), 1142-1158.

Kalra, S., Lamba, R., & Sharma, M. (2020). Machine learning based analysis for relation between global temperature and concentrations of greenhouse gases. Journal of Information and Optimization Sciences, 41(1), 73-84.

Kim, J., Yoon, S., & Hong, S. (2024). Exploring Influencing Factors at Student and Teacher/School levels on Science Self-efficacy Using Machine Learning and Multilevel Latent Profile Analysis. SAGE Open, 14(4), 21582440241284915.

Ko, J., Leung, C. K., Chen, X., & Palmer, D. A. (2024). From emissions to emotions: Exploring the impact of climate change on happiness across 140 countries. Global Transitions, 6, 231-240.

Liao, J., Zheng, W., Liao, Q., & Lu, S. (2024). Global latitudinal patterns in forest ecosystem nitrous oxide emissions are related to hydroclimate. npj Climate and Atmospheric Science, 7(1), 187.

Marzadri, A., Amatulli, G., Tonina, D., Bellin, A., Shen, L. Q., Allen, G. H., & Raymond, P. A. (2021). Global riverine nitrous oxide emissions: The role of small streams and large rivers. Science of The Total Environment, 776, 145148.

Peng, Y., Wang, T., Li, J., Li, N., Bai, X., Liu, X., ... & Chang, R. (2024). Temporal-scale-dependent mechanisms of forest soil nitrous oxide emissions under nitrogen addition. Communications Earth & Environment, 5(1), 512.

Saha, D., Basso, B., & Robertson, G. P. (2021). Machine learning improves predictions of agricultural nitrous oxide (N2O) emissions from intensively managed cropping systems. Environmental Research Letters, 16(2), 024004.

Sengupta, A., & Ismail, F. N. (2024, December). Modelling methane emissions from rice paddies using machine learning. In 2024 39th International Conference on Image and Vision Computing New Zealand (IVCNZ) (pp. 1-6). IEEE.

Sivakumar, S., & Venkataraman, S. (2025). Evaluating Machine Learning Approaches: A Comparative Study of Random Forest and Neural Networks in Grade Classification. Indonesian Journal of Data and Science, 6(1), 74-81.

Song, H., Peng, C., Zhang, K., Li, T., Yang, M., Liu, Q., & Zhu, Q. (2023). Quantifying patterns, sources and uncertainty of nitrous oxide emissions from global grazing lands: Nitrogen forms are the determinant factors for estimation and mitigation. Global and Planetary Change, 223, 104080.

Yu, L., Zhang, Q., Tian, Y., Sun, W., Scheer, C., Li, T., & Zhang, W. (2022). Global variations and drivers of nitrous oxide emissions from forests and grasslands. Frontiers in Soil Science, 2, 1094177.

Q16. Tables: There are just a ton of tables. In many cases — most glaringly in the results of the regression analyses (see e.g., Table 2 and Table 8) no important statistical indicators (e.g., adjusted R², VIF — multicollinearity — or post-estimation tests) are provided or are only weakly explained.

A16. Tests and statistics have been added in each table containing the results of the econometric analysis.

Q17. Figures: Clustering figures references (Figures 1 and 2) as-enumerated without interpretative depth. For example, cluster visualization can be aided with clearer labeling, titles on the axes and a better explanation of what the visual pattern actually indicates either for a policy action or empirical interpretation.

A17. Figure 1 and 2 have been commented as follows

Figure 1. Clusters and noisepoint. Visualized using dimensionality reduction—e.g., t-SNE or PCA—this figure shows the output of a density-based clustering algorithm, probably DBSCAN. The result reveals a dominant single cluster (Cluster 1, in pink), indicating a high concentration of observations with similar ESG-N₂O profiles. There are also a very small secondary cluster (Cluster 2, in green) and some noise points (in blue). The notable disparity between clusters points to a very homogeneous data structure or conservative parameter tuning (e.g., a high eps or low minPts threshold), which compromises granularity. Although the algorithm effectively identifies outliers, the small creation of separate clusters might compromise the possibility for detailed policy segmentation unless feature engineering or parameter optimization is more fine-tuned.

Figure 2. Cluster mean. With special emphasis on N²O emissions, Figure 2 shows the average values of chosen ESG-related variables across three clusters. The most remarkable finding is the stark difference in N₂O emission levels between clusters. With a significantly higher mean for N₂O emissions, Cluster 0 (pink) suggests this group is made up of countries or units with high N₂O emission intensity. By comparison, Cluster 2 (blue) has significantly less N₂O emissions, therefore qualifying as a low-emission cluster. This trend verifies that the clustering result is mostly driven by N2O emissions. The variable ASNFD, probably linked to environmental deterioration (e.g., forest depletion), also displays higher values in Cluster 2, indicating a complicated environmental profile where forest stress does not always correspond with high N₂O production. Structural variations in emissions sources—such as low agricultural intensity or efficient Cluster 2 mitigation efforts—could explain this one. On the other hand, Cluster 0 shows more environmental risk by combining high N2O emissions with moderate forest depletion. The other indicators—EIPE and FA—show little variation across clusters and so help to explain the observed differentiation less. These findings draw attention to the need of N₂O emissions and ASNFD in clarifying ESG-related diversity and offer a foundation for focused mitigation measures reflecting the unique environmental and governance profile of every cluster.

Q18. Data Quality: Although the data source (World Bank ESG database and others) is reliable, the manuscript lacks a detailed description of data cleaning methods, missing data treatment, and transformations performed. Also, no standard diagnostics of panel data (stationarity test, multicolinearity check, control of heteroscedasticity, etc.) are analyzed.

A18. We have added the following proposition in the data and methodology:

Missing data have been replaced with the mean of observations.

We also have added tests and statistics in each econometric model as follows:

Q19. Cluster Evaluation Metrics: Clustering evaluation metrics such as silhouette scores or Dunn indices are appropriate, however, I would have liked to see a bit more discussion on what exactly the poor performance of some algorithms means for the robustness of the analysis.

A19. We have added this critique in the section “Limitations and Future Research”

Though the research presents a unique combination of econometric and machine learning techniques to examine N₂O emissions in an ESG framework, some shortcomings deserve attention. First, especially for developing nations with weaker statistical systems, the reliance on publicly available macro-level datasets, such as those from the World Bank, could raise questions about data completeness, granularity, and temporal consistency. These constraints could weaken the force of the conclusions reached and lower the accuracy of cross-country comparisons. Second, although the econometric models use fixed and random effects to compensate for unobserved heterogeneity, causal identification stays limited and possible endogeneity between ESG indicators and emission results cannot be completely excluded. Though the machine learning models show good predictive performance, they lack explanatory clarity since they do not reflect the structural causal mechanisms behind N₂O emissions. Third, the study does not explicitly control for important structural country-level characteristics including geographic area, income group, or stage of economic development. Lacking such controls could create omitted variable bias and reduce the generalizability of the results across several socio-economic settings. Although the clustering algorithms generate country groupings that roughly correspond with development status, future work should improve robustness by adding stratified models or using macro-structural variables as controls.

Another drawback is the handling of ESG variables in isolation. Inherently multidimensional, the ESG framework comprises environmental, social, and governance pillars that sometimes interact synergistically or oppositely. Ignoring these interdependencies could cloud intricate mediating routes whereby ESG elements together affect emissions. Strong governance, for instance, could improve the implementation of environmental policies; social inclusion could reduce the environmental effect of economic growth. Future studies should use structural equation modeling, interaction terms in fixed-effects regressions, or sophisticated machine learning architectures able of capturing non-linear dependencies to explore such interaction effects. This would enhance the theoretical strength as well as the practical relevance of ESG-based emissions modeling.

Important factors also relate to the clustering framework used. To capture the diversity and multidimensionality of ESG-emissions interactions across nations, this paper uses a comparative, multi-algorithm approach—testing density-based, fuzzy C-means, hierarchical, model-based, neighborhood-based, and random forest clustering techniques. A consistent set of internal validation metrics—silhouette scores, Calinski–Harabasz indices, Dunn indices, entropy, and Pearson's γ—was used to assess every algorithm. Although this improves methodological openness, we acknowledge that certain algorithms, especially fuzzy C-means and neighborhood-based clustering, showed rather bad performance in terms of intra-cluster cohesion and separation. Low silhouette and Dunn values, as well as high entropy, suggest inaccurate cluster boundaries and potential data overlap, which undermines interpretability. By comparison, hierarchical clustering and density-based clustering revealed more obvious group structures and better validation results. A significant cluster size disparity still raises questions about overgeneralization and the representativeness of smaller clusters, particularly in the density-based model. Data-driven concentration or poor hyperparameter choice—for example, epsilon, minPts in DBSCAN—could cause this imbalance. Future research should include more robustness tests—sensitivity analysis across parameter settings, changes in distance metrics, and maybe ensemble clustering techniques—to enhance the interpretability and generalizability of the findings. Furthermore, including external validation using regional or income classifications and using sophisticated visualisation tools could improve the clarity of clustering results and their applicability for policy creation.

Future studies could eventually gain from including subnational or firm-level ESG data to more precisely reflect emissions behavior at smaller spatial or organizational levels. Using hybrid causal inference systems or temporal deep learning models could help to untangle dynamic interactions between N₂O emissions and ESG components over time. Including life-cycle assessment data and Scope 3 emissions, therefore, would increase the practical relevance of ESG measures for sustainability accounting even more. At last, empirical testing of the efficacy of particular ESG policy tools—such as carbon pricing, green bonds, or disclosure obligations—across several governance settings would offer much-needed data to steer climate policy and increase the relevance of ESG frameworks in worldwide mitigation efforts.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Q5. The paper focuses on global N₂O emissions. At what scale were the variable data collected by the authors? Is the data volume sufficient for cluster and machine learning analysis?

The fifth question has not been answered completely. The authors use not only N2O emissions data, which casts doubt on the reliability of the data

Author Response

Q5. The paper focuses on global N₂O emissions. At what scale were the variable data collected by the authors? Is the data volume sufficient for cluster and machine learning analysis? The fifth question has not been answered completely. The authors use not only N2O emissions data, which casts doubt on the reliability of the data.

A5. We have added a fourth appendix with the following descriptive statistics to describe the data. The appendix also presents the hyperparameters of the algorithms including the learning rate partitions.

Appendix 4 Descriptive statistics

Table 38. Descriptive Statistics

						95% Confidence Interval Mean							95% Confidence Interval Variance
	Valid	Missing	Median	Mean	Std. Error of Mean	Upper	Lower	Std. Deviation	Coefficient of variation	MAD	MAD robust	IQR	Upper	Lower
AGRI	1930	0	0.000	1.108×10⁺¹²	5.204×10⁺¹¹	2.129×10⁺¹²	8.761×10⁺¹⁰	2.286×10⁺¹³	20.630	0.000	0.000	0.000	5.574×10⁺²⁶	4.913×10⁺²⁶
FRT	1930	0	1.909.000	2.558×10⁺¹²	8.606×10⁺¹¹	4.245×10⁺¹²	8.697×10⁺¹¹	3.781×10⁺¹³	14.783	1.907.250	2.827.689	3.239.407	1.524×10⁺²⁷	1.343×10⁺²⁷
GI	1930	0	0.000	-7.876.310	2.611.503	-2.754.645	-12.997.975	114.727.926	-14.566	0.000	0.000	30.500	1.403×10⁺¹⁰	1.237×10⁺¹⁰
ISL20	1930	0	0.000	2.171	0.075	2.319	2.023	3.309	1.524	0.000	0.000	5.200	11.673	10.288
N2O	1930	0	0.284	1.666×10⁺¹³	1.581×10⁺¹²	1.976×10⁺¹³	1.356×10⁺¹³	6.945×10⁺¹³	4.170	0.200	0.296	0.420	5.143×10⁺²⁷	4.533×10⁺²⁷
WATER	1930	0	1.000×10⁺⁹	3.563×10⁺⁹	9.203×10⁺⁷	3.744×10⁺⁹	3.383×10⁺⁹	4.043×10⁺⁹	1.135	1.000×10⁺⁹	1.483×10⁺⁹	8.041×10⁺⁹	1.743×10⁺¹⁹	1.536×10⁺¹⁹
GDPG	1930	0	2.215×10⁺¹⁴	2.162×10⁺¹⁴	7.647×10⁺¹²	2.312×10⁺¹⁴	2.012×10⁺¹⁴	3.359×10⁺¹⁴	1.554	2.215×10⁺¹⁴	3.284×10⁺¹⁴	4.440×10⁺¹⁴	1.203×10⁺²⁹	1.061×10⁺²⁹
FMLP	1930	0	6.460×10⁺¹⁴	5.204×10⁺¹⁴	7.873×10⁺¹²	5.359×10⁺¹⁴	5.050×10⁺¹⁴	3.459×10⁺¹⁴	0.665	2.110×10⁺¹⁴	3.128×10⁺¹⁴	7.300×10⁺¹⁴	1.275×10⁺²⁹	1.124×10⁺²⁹
RQE	1930	0	-0.178	8.771×10⁺¹²	1.929×10⁺¹²	1.256×10⁺¹³	4.987×10⁺¹²	8.476×10⁺¹³	9.664	0.651	0.965	1.335	7.661×10⁺²⁷	6.752×10⁺²⁷
RDE	1930	0	0.000	2.277×10⁺¹²	6.394×10⁺¹¹	3.531×10⁺¹²	1.023×10⁺¹²	2.809×10⁺¹³	12.336	0.000	0.000	0.349	8.412×10⁺²⁶	7.414×10⁺²⁶
STJA	1930	0	103.530	9.141×10⁺¹¹	2.196×10⁺¹¹	1.345×10⁺¹²	4.834×10⁺¹¹	9.648×10⁺¹²	10.554	103.530	153.494	1.842.075	9.924×10⁺²⁵	8.747×10⁺²⁵
SLRI	1930	0	2.000	1.959×10⁺¹¹	1.130×10⁺¹¹	4.175×10⁺¹¹	-2.580×10⁺¹⁰	4.965×10⁺¹²	25.351	2.000	2.965	6.000	2.629×10⁺²⁵	2.317×10⁺²⁵
ASFND	1930	0	0.000	3.729×10⁺¹³	2.949×10⁺¹²	4.308×10⁺¹³	3.151×10⁺¹³	1.296×10⁺¹⁴	3.475	0.000	0.000	0.094	1.790×10⁺²⁸	1.578×10⁺²⁸
EIPE	1930	0	0.000	1.946×10⁺¹⁴	5.776×10⁺¹²	2.060×10⁺¹⁴	1.833×10⁺¹⁴	2.537×10⁺¹⁴	1.304	0.000	0.000	3.840×10⁺¹⁴	6.865×10⁺²⁸	6.051×10⁺²⁸
FA	1930	0	3.110×10⁺¹⁴	3.343×10⁺¹⁴	5.880×10⁺¹²	3.459×10⁺¹⁴	3.228×10⁺¹⁴	2.583×10⁺¹⁴	0.773	2.010×10⁺¹⁴	2.980×10⁺¹⁴	4.090×10⁺¹⁴	7.114×10⁺²⁸	6.271×10⁺²⁸
	Skewness	Kurtosis	Shapiro-Wilk	Range	Minimum	Maximum	25th percentile	50th percentile	75th percentile	25th percentile	50th percentile	75th percentile	Sum	Variance
AGRI	23.431	581.910	0.024	6.130×10⁺¹⁴	-9.230	6.130×10⁺¹⁴	0.000	0.000	0.000	0.000	0.000	0.000	2.139×10⁺¹⁵	5.228×10⁺²⁶
FRT	15.039	228.666	0.040	6.540×10⁺¹⁴	0.000	6.540×10⁺¹⁴	1.593	1.909.000	3.241.000	1.593	1.909.000	3.241.000	4.936×10⁺¹⁵	1.429×10⁺²⁷
GI	-14.773	218.983	0.041	1.911×10⁺⁶	-1.911×10⁺⁶	63.000	0.000	0.000	30.500	0.000	0.000	30.500	-1.520×10⁺⁷	1.316×10⁺¹⁰
ISL20	1.053	-0.549	0.668	10.500	0.000	10.500	0.000	0.000	5.200	0.000	0.000	5.200	4.189.930	10.947
N2O	5.755	40.252	0.260	7.080×10⁺¹⁴	-90.000.000	7.080×10⁺¹⁴	0.046	0.284	0.466	0.046	0.284	0.466	3.215×10⁺¹⁶	4.824×10⁺²⁷
WATER	0.526	-1.466	0.766	1.000×10⁺¹⁰	0.000	1.000×10⁺¹⁰	0.000	1.000×10⁺⁹	8.041×10⁺⁹	0.000	1.000×10⁺⁹	8.041×10⁺⁹	6.877×10⁺¹²	1.635×10⁺¹⁹
GDPG	-0.551	0.925	0.969	1.999×10⁺¹⁵	-9.990×10⁺¹⁴	1.000×10⁺¹⁵	0.682	2.215×10⁺¹⁴	4.440×10⁺¹⁴	0.682	2.215×10⁺¹⁴	4.440×10⁺¹⁴	4.172×10⁺¹⁷	1.129×10⁺²⁹
FMLP	-0.479	-1.389	0.845	9.980×10⁺¹⁴	-2.530×10⁺⁶	9.980×10⁺¹⁴	8.703×10⁺¹³	6.460×10⁺¹⁴	8.170×10⁺¹⁴	8.703×10⁺¹³	6.460×10⁺¹⁴	8.170×10⁺¹⁴	1.004×10⁺¹⁸	1.196×10⁺²⁹
RQE	9.733	93.385	0.076	8.910×10⁺¹⁴	-2.378×10⁺⁶	8.910×10⁺¹⁴	-0.771	-0.178	0.564	-0.771	-0.178	0.564	1.693×10⁺¹⁶	7.185×10⁺²⁷
RDE	13.597	192.082	0.054	4.450×10⁺¹⁴	-1.716×10⁺⁶	4.450×10⁺¹⁴	0.000	0.000	0.349	0.000	0.000	0.349	4.395×10⁺¹⁵	7.889×10⁺²⁶
STJA	10.520	108.898	0.066	1.060×10⁺¹⁴	0.000	1.060×10⁺¹⁴	3.592	103.530	1.845.668	3.592	103.530	1.845.668	1.764×10⁺¹⁵	9.308×10⁺²⁵
SLRI	25.325	639.995	0.017	1.260×10⁺¹⁴	0.000	1.260×10⁺¹⁴	0.000	2.000	6.000	0.000	2.000	6.000	3.780×10⁺¹⁴	2.465×10⁺²⁵
ASFND	4.211	18.872	0.326	9.900×10⁺¹⁴	0.000	9.900×10⁺¹⁴	0.000	0.000	0.094	0.000	0.000	0.094	7.197×10⁺¹⁶	1.679×10⁺²⁸
EIPE	1.034	-0.014	0.776	1.122×10⁺¹⁵	-1.230×10⁺¹⁴	9.990×10⁺¹⁴	0.000	0.000	3.840×10⁺¹⁴	0.000	0.000	3.840×10⁺¹⁴	3.757×10⁺¹⁷	6.438×10⁺²⁸
FA	0.528	-0.557	0.943	9.990×10⁺¹⁴	0.000	9.990×10⁺¹⁴	1.130×10⁺¹⁴	3.110×10⁺¹⁴	5.220×10⁺¹⁴	1.130×10⁺¹⁴	3.110×10⁺¹⁴	5.220×10⁺¹⁴	6.453×10⁺¹⁷	6.673×10⁺²⁸

Table 39. Association matrix-Covariance

AGRI

FRT

ISL20

N2O

WATER

GDPG

FMLP

RQE

RDE

STJA

SLRI

ASFND

EIPE

AGRI

5.228×10⁺²⁶

2.725×10⁺²⁶

-9.277×10⁺¹⁶

1.526×10⁺¹²

-1.847×10⁺²⁵

-3.951×10⁺²¹

6.714×10⁺²⁵

-4.862×10⁺²⁶

3.841×10⁺²⁶

3.211×10⁺²⁶

1.007×10⁺²⁶

-2.172×10⁺²³

1.302×10⁺²⁶

-8.129×10⁺²⁵

-3.707×10⁺²⁶

FRT

2.725×10⁺²⁶

1.429×10⁺²⁷

2.015×10⁺¹⁶

8.761×10⁺¹²

-4.262×10⁺²⁵

-9.117×10⁺²¹

1.080×10⁺²⁷

-1.115×10⁺²⁷

1.956×10⁺²⁷

7.509×10⁺²⁶

2.403×10⁺²⁶

-5.012×10⁺²³

3.790×10⁺²⁶

1.824×10⁺²⁶

-8.555×10⁺²⁶

-9.277×10⁺¹⁶

2.015×10⁺¹⁶

1.316×10⁺¹⁰

17.182.406

1.315×10⁺¹⁷

2.814×10⁺¹³

1.707×10⁺¹⁸

4.109×10⁺¹⁸

-2.164×10⁺¹⁸

1.794×10⁺¹⁶

7.203×10⁺¹⁵

-3.168×10⁺¹⁷

-1.103×10⁺¹⁸

-4.159×10⁺¹⁷

2.607×10⁺¹⁷

ISL20

1.526×10⁺¹²

8.761×10⁺¹²

17.182.406

10.947

3.793×10⁺¹²

4.972×10⁺⁹

8.231×10⁺¹³

2.883×10⁺¹⁴

5.607×10⁺¹²

4.318×10⁺¹²

1.060×10⁺¹²

-4.254×10⁺¹¹

-3.975×10⁺¹³

1.315×10⁺¹⁴

-4.124×10⁺¹³

N2O

-1.847×10⁺²⁵

-4.262×10⁺²⁵

1.315×10⁺¹⁷

3.793×10⁺¹²

4.824×10⁺²⁷

5.979×10⁺²¹

8.793×10⁺²⁶

2.887×10⁺²⁷

-1.462×10⁺²⁶

-3.795×10⁺²⁵

-1.523×10⁺²⁵

-3.264×10⁺²⁴

-4.672×10⁺²⁶

1.364×10⁺²⁷

1.324×10⁺²⁶

WATER

-3.951×10⁺²¹

-9.117×10⁺²¹

2.814×10⁺¹³

4.972×10⁺⁹

5.979×10⁺²¹

1.635×10⁺¹⁹

-1.070×10⁺²³

1.408×10⁺²³

-3.127×10⁺²²

-8.117×10⁺²¹

-3.259×10⁺²¹

-6.982×10⁺²⁰

-8.215×10⁺²²

2.792×10⁺²²

-1.161×10⁺²²

GDPG

6.714×10⁺²⁵

1.080×10⁺²⁷

1.707×10⁺¹⁸

8.231×10⁺¹³

8.793×10⁺²⁶

-1.070×10⁺²³

1.129×10⁺²⁹

2.992×10⁺²⁸

6.727×10⁺²⁶

2.127×10⁺²⁶

7.909×10⁺²⁵

-4.236×10⁺²⁵

7.166×10⁺²⁷

1.258×10⁺²⁸

-9.573×10⁺²⁶

FMLP

-4.862×10⁺²⁶

-1.115×10⁺²⁷

4.109×10⁺¹⁸

2.883×10⁺¹⁴

2.887×10⁺²⁷

1.408×10⁺²³

2.992×10⁺²⁸

1.196×10⁺²⁹

-3.576×10⁺²⁷

-1.042×10⁺²⁷

-3.691×10⁺²⁶

-1.020×10⁺²⁶

4.612×10⁺²⁷

1.537×10⁺²⁸

6.531×10⁺²⁷

RQE

3.841×10⁺²⁶

1.956×10⁺²⁷

-2.164×10⁺¹⁸

5.607×10⁺¹²

-1.462×10⁺²⁶

-3.127×10⁺²²

6.727×10⁺²⁶

-3.576×10⁺²⁷

7.185×10⁺²⁷

1.630×10⁺²⁷

6.983×10⁺²⁶

-1.719×10⁺²⁴

2.112×10⁺²⁷

-3.834×10⁺²⁶

-2.213×10⁺²⁷

RDE

3.211×10⁺²⁶

7.509×10⁺²⁶

1.794×10⁺¹⁶

4.318×10⁺¹²

-3.795×10⁺²⁵

-8.117×10⁺²¹

2.127×10⁺²⁶

-1.042×10⁺²⁷

1.630×10⁺²⁷

7.889×10⁺²⁶

2.287×10⁺²⁶

-4.462×10⁺²³

2.235×10⁺²⁶

1.554×10⁺²⁵

-7.617×10⁺²⁶

STJA

1.007×10⁺²⁶

2.403×10⁺²⁶

7.203×10⁺¹⁵

1.060×10⁺¹²

-1.523×10⁺²⁵

-3.259×10⁺²¹

7.909×10⁺²⁵

-3.691×10⁺²⁶

6.983×10⁺²⁶

2.287×10⁺²⁶

9.308×10⁺²⁵

-1.791×10⁺²³

1.765×10⁺²⁶

-6.332×10⁺²⁵

-3.058×10⁺²⁶

SLRI

-2.172×10⁺²³

-5.012×10⁺²³

-3.168×10⁺¹⁷

-4.254×10⁺¹¹

-3.264×10⁺²⁴

-6.982×10⁺²⁰

-4.236×10⁺²⁵

-1.020×10⁺²⁶

-1.719×10⁺²⁴

-4.462×10⁺²³

-1.791×10⁺²³

2.465×10⁺²⁵

3.541×10⁺²⁵

-3.814×10⁺²⁵

-6.551×10⁺²⁵

ASFND

1.302×10⁺²⁶

3.790×10⁺²⁶

-1.103×10⁺¹⁸

-3.975×10⁺¹³

-4.672×10⁺²⁶

-8.215×10⁺²²

7.166×10⁺²⁷

4.612×10⁺²⁷

2.112×10⁺²⁷

2.235×10⁺²⁶

1.765×10⁺²⁶

3.541×10⁺²⁵

1.679×10⁺²⁸

-3.448×10⁺²⁶

-2.429×10⁺²⁷

EIPE

-8.129×10⁺²⁵

1.824×10⁺²⁶

-4.159×10⁺¹⁷

1.315×10⁺¹⁴

1.364×10⁺²⁷

2.792×10⁺²²

1.258×10⁺²⁸

1.537×10⁺²⁸

-3.834×10⁺²⁶

1.554×10⁺²⁵

-6.332×10⁺²⁵

-3.814×10⁺²⁵

-3.448×10⁺²⁶

6.438×10⁺²⁸

1.918×10⁺²⁷

-3.707×10⁺²⁶

-8.555×10⁺²⁶

2.607×10⁺¹⁷

-4.124×10⁺¹³

1.324×10⁺²⁶

-1.161×10⁺²²

-9.573×10⁺²⁶

6.531×10⁺²⁷

-2.213×10⁺²⁷

-7.617×10⁺²⁶

-3.058×10⁺²⁶

-6.551×10⁺²⁵

-2.429×10⁺²⁷

1.918×10⁺²⁷

6.673×10⁺²⁸

Table 40. Correlation

	AGRI	FRT	GI	ISL20	N2O	WATER	GDPG	FMLP	RQE	RDE	STJA	SLRI	ASFND	EIPE	FA
AGRI	1.000	0.315	-0.035	0.020	-0.012	-0.043	0.009	-0.061	0.198	0.500	0.457	-0.002	0.044	-0.014	-0.063
FRT	0.315	1.000	0.005	0.070	-0.016	-0.060	0.085	-0.085	0.610	0.707	0.659	-0.003	0.077	0.019	-0.088
GI	-0.035	0.005	1.000	0.045	0.016	0.061	0.044	0.104	-0.222	0.006	0.007	-0.556	-0.074	-0.014	0.009
ISL20	0.020	0.070	0.045	1.000	0.017	0.372	0.074	0.252	0.020	0.046	0.033	-0.026	-0.093	0.157	-0.048
N2O	-0.012	-0.016	0.016	0.017	1.000	0.021	0.038	0.120	-0.025	-0.019	-0.023	-0.009	-0.052	0.077	0.007
WATER	-0.043	-0.060	0.061	0.372	0.021	1.000	-0.079	0.101	-0.091	-0.071	-0.084	-0.035	-0.157	0.027	-0.011
GDPG	0.009	0.085	0.044	0.074	0.038	-0.079	1.000	0.258	0.024	0.023	0.024	-0.025	0.165	0.148	-0.011
FMLP	-0.061	-0.085	0.104	0.252	0.120	0.101	0.258	1.000	-0.122	-0.107	-0.111	-0.059	0.103	0.175	0.073
RQE	0.198	0.610	-0.222	0.020	-0.025	-0.091	0.024	-0.122	1.000	0.685	0.854	-0.004	0.192	-0.018	-0.101
RDE	0.500	0.707	0.006	0.046	-0.019	-0.071	0.023	-0.107	0.685	1.000	0.844	-0.003	0.061	0.002	-0.105
STJA	0.457	0.659	0.007	0.033	-0.023	-0.084	0.024	-0.111	0.854	0.844	1.000	-0.004	0.141	-0.026	-0.123
SLRI	-0.002	-0.003	-0.556	-0.026	-0.009	-0.035	-0.025	-0.059	-0.004	-0.003	-0.004	1.000	0.055	-0.030	-0.051
ASFND	0.044	0.077	-0.074	-0.093	-0.052	-0.157	0.165	0.103	0.192	0.061	0.141	0.055	1.000	-0.010	-0.073
EIPE	-0.014	0.019	-0.014	0.157	0.077	0.027	0.148	0.175	-0.018	0.002	-0.026	-0.030	-0.010	1.000	0.029
FA	-0.063	-0.088	0.009	-0.048	0.007	-0.011	-0.011	0.073	-0.101	-0.105	-0.123	-0.051	-0.073	0.029	1.000

Figure 9. Correlation plot E-Environment

Figure 10. Correlation plot S-Social

Figure 11. Correlation plot G-Governance

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The dpi of the image needs to be improved and is currently not very clear, such as in Figure 1. The academic standardization of the format needs to be strengthened, such as the 370 line formula representation and table 3. The academic innovation and contribution section of this article can be improved, but currently it is not very focused and a bit vague.

Author Response

Q1. The dpi of the image needs to be improved and is currently not very clear, such as in Figure 1.

A1. The level of dpi has been improved to 1000 as follows:

Q2. The academic standardization of the format needs to be strengthened, such as the 370 line formula representation and table 3.

A2. We have used the math tool to rewrite the equations as follows.

Tables have many information and it can be difficult to store them in a fashionable format.

Q3. The academic innovation and contribution section of this article can be improved, but currently it is not very focused and a bit vague.

A3. We have added a sub-paragraph in the fifth section as follows:

Innovativeness of the contribution. The research provides a fresh and timely conceptual framework for assessing global N₂O emissions using Environmental, Social, and Governance (ESG) measures and thus responds to a significant research gap for sustainability studies and policy debate. Applying the reach of the models to N₂O—a highly impactful greenhouse gas much too frequently ignored by existing approaches—the research extends the conventional, carbon-focused view of the impacts of climate within the analysis framework of the ESG to an integrated view of environmental responsibility.

Methodologically, the study stands out for utilizing the combination of panel econometric methods and advanced machine learning methods, such as clustering and prediction modeling. Not only do explanations grow more rich using the hybrid approach, but predictive power is significantly improved, which captures subtle, non-linear patterns that might go undetected using more traditional approaches.

A new innovation would be the creation of data-driven country typologies where countries would be categorized based on their emission pathways and performance regarding their environment, social and governance factors. The categorization would provide evidence-based, nuanced observations for the creation of tailored policy interventions and for underpinning sustainable investing strategies.

By pointing to major structural drivers of N₂O emissions—specifically forest cover and energy intensity—the research offers policymakers a more effective platform from which to embark upon focused activities to curb. Using macro-sustainability analysis and advanced data science, the research makes a valuable contribution to the literature by filling an important gap between climate and ESG interactions and offering policymakers sound evidence-driven avenues for more effective and equitable climate policy.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Author/s addressed all the comments had been mentioned, and I agree for publication in this version

Author Response

Q1. Author/s addressed all the comments had been mentioned, and I agree for publication in this version

A1. Thanks.

Reviewer 5 Report

Comments and Suggestions for Authors

Overall, the revised manuscript makes an admirable effort to address the concerns from the previous round of peer review. The newly added hypthesis, enlarged literature review, and more detailed methodological justification greatly elevate the paper’s rigor and clarity. The incorporation of more econometric models and machine learning algorithms are well-argued, with the treatment of the limitations of each model type tending toward the appropriate level of reflection and Transparency. I especially admire your responses to the critiques surrounding the structural controls, model validation, and ESG interaction effects. The revised “Limitations and Future Research” section, which I perceive to be mostly new, is detailed and shows a genuine struggle with the task at hand. Improving your cluster analysis’s interpretation was especially useful, as was its supplement with internal validation statistics.

Comments on the Quality of English Language

Even though this manuscript is strong in terms of its organization and argumentation, it could benefit from a further language polish to enhance the language’s fluency and the distinguished length and complexity of some of its sentences. This would improve the text’s readability and efficiency. Overall, I believe that these findings are a valued addition to the scientific literature on environmental sustainability and ESG performance and greenhouse gases and encourage you to address these more minor concerns to finish improving the clarity and strength of your manuscript.

Author Response

Q1. Comments and Suggestions for Authors. Overall, the revised manuscript makes an admirable effort to address the concerns from the previous round of peer review. The newly added hypthesis, enlarged literature review, and more detailed methodological justification greatly elevate the paper’s rigor and clarity. The incorporation of more econometric models and machine learning algorithms are well-argued, with the treatment of the limitations of each model type tending toward the appropriate level of reflection and Transparency. I especially admire your responses to the critiques surrounding the structural controls, model validation, and ESG interaction effects. The revised “Limitations and Future Research” section, which I perceive to be mostly new, is detailed and shows a genuine struggle with the task at hand. Improving your cluster analysis’s interpretation was especially useful, as was its supplement with internal validation statistics.

A1. Thanks.

Q2. Comments on the Quality of English Language. Even though this manuscript is strong in terms of its organization and argumentation, it could benefit from a further language polish to enhance the language’s fluency and the distinguished length and complexity of some of its sentences. This would improve the text’s readability and efficiency. Overall, I believe that these findings are a valued addition to the scientific literature on environmental sustainability and ESG performance and greenhouse gases and encourage you to address these more minor concerns to finish improving the clarity and strength of your manuscript.

A2. We have tried to improve the quality of English. Thanks.

Author Response File: Author Response.pdf

Article Menu

A Machine Learning and Panel Data Analysis of N₂O Emissions in an ESG Framework

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

Article Menu

A Machine Learning and Panel Data Analysis of N2O Emissions in an ESG Framework

Further Information

Guidelines

MDPI Initiatives

Follow MDPI

A Machine Learning and Panel Data Analysis of N₂O Emissions in an ESG Framework