Investigation of the spatial relationship between CO2 emissions and a hazardous waste management policy: a Geographically Weighted Regression analysis in Texas, USA. = 지리적 가중회귀모형(Geographic Weighted Regression model)을 활용한 유해 폐기물 정책의 환경적 영향 분석

다국어 초록 (Multilingual Abstract)

This dissertation presents an in-depth analysis of the effects of 25 distinct features on carbon dioxide equivalent (CO2eq) emissions at the ZIP code level across the state of Texas, USA. Utilizing the advanced analytical capabilities of Geographically Weighted Regression (GWR) and Multiscale Geographically Weighted Regression (MGWR), the study aims to uncover the spatial relationships between various independent variables and CO2eq emissions and understand the spatial heterogeneities present within the ZIP codes and regions of Texas. The research methodically examines an array of factors, including demographic, economic, land use, and environmental variables. Among these, the study finds a significant positive relationship between the proportion of the Black African American population, the extent of Agricultural and Forest land, and the levels of CO2eq emissions. Intriguingly, the analysis also reveals a significant negative relationship between higher poverty levels and CO2eq emissions. This counterintuitive finding suggests a complex interplay between socio-economic status and environmental impact.
Spatial analysis further highlights regional disparities within Texas. Notably, the Eastern region of Texas exhibits a more pronounced response to the studied features in terms of their impact on CO2eq emissions. This variation is indicative of the diverse levels of industrialization, urbanization, and socio-economic conditions prevailing across the state. The thesis extends beyond mere identification of relationships and delves into policy implications. The findings illuminate the critical role of demographic and land-use factors in influencing CO2eq emissions. Based on these insights, the study proposes a set of policy recommendations aimed at mitigating CO2eq emissions. These recommendations are tailored to address the specific needs and characteristics of different regions within Texas, highlighting the importance of context-specific approaches in environmental policy-making. In conclusion, this thesis contributes significantly to the understanding of the spatial dynamics of CO2eq emissions in Texas. It underscores the multifaceted nature of environmental challenges and the need for nuanced, region-specific policy interventions. By integrating demographic, economic, and land-use factors into the analysis of CO2eq emissions, the study offers a comprehensive framework for policymakers to devise more effective and targeted environmental strategies. This dissertation is essential in academic research as it provides a comprehensive analysis of the spatial dynamics of CO2eq emissions in Texas, revealing critical relationships between demographic, economic, and land-use factors, and offering targeted, region-specific policy recommendations for more effective environmental strategies.

국문 초록 (Abstract)

본 논문은 탄소배출량(이산화탄소 등가물, CO2equivalent)과 사회경제적 및 환경 요인에 대한 예측 모형을 구축하여, 탄소배출량에 영향을 미치는 요인들을 계량화하고 정책 대안을 제시하였다. 다중 규모 지리적 가중 회귀(MGWR)을 활용하여 텍사스 주의 우편 번호 및 지역 내에 존재하는 공간적 이질성을 통제하고 독립 변수와 이산화탄소 배출량 사이의 관계를 연구하였다. 독립변수로는 인구 통계학적, 경제적, 토지 사용 및 환경 요인이 활용되었다. 모형 분석 결과를 살펴보면, 흑인 인구의 비율과 농업 및 산림 토지의 범위 및 이산화탄소 배출량 수준 사이에 양의 관계가 있는 것을 확인하였다. 반면, 빈곤 수준과 이산화탄소 배출량 사이에는 음의 관계가 있음을 확인하였다. 이러한 결과를 통해 사회경제지표와 환경 영향 사이의 복잡한 상호 작용이 있다는 것이 확인되었으며, 고유한 지리적 특성에 따라 그 영향이 상이한 것을 확인할 수 있었다. 본 논문은 모형을 통한 인구 통계학적, 경제적, 토지 사용 및 환경 요인과 환경적 영향과의 관계성을 파악한 것을 바탕으로 정책적 방향성도 제시하였다는 점에서 큰 의의가 있다. 본 연구를 통해 텍사스 탄소배출량의 공간적인 이질성을 이해하고 이에 영향을 미칠 수 있는 요인들을 예측하여 환경 문제의 상호의존적인 특성과 지역 특성이 반영된 정책 수립의 필요성을 강조하였다. 본 연구 결과는 정책 결정자 외에도 도시 환경을 연구하는 연구자들의 연구 및 의사결정을 지원할 수 있는 프레임워크를 제공했다는 점에서 큰 의의가 있다.

목차 (Table of Contents)

• ABSTRACT i
• Table of contents iii
• List of tables v
• List of figures vi
• 1. INTRODUCTION 1
• 1.1. History of waste management 1
• 1.2. Environmental policies 6
• 1.3. Research question and objective 8
• 1.4. Statistical analysis and Predictive modelling 8
• 1.5. Scope and area of study 18
• 1.6. Related work 19
• 1.7. Purpose, relevance and contribution of the study 20
• 2. THEORETICAL FRAMEWORK 21
• 2.1. Review of basic concepts 21
• 2.2. Previous works related to the study of policy impact 22
• 2.3. Previous works related to the study of environmental policy ś impact 22
• 2.4. Previous works related to the study of environmental systems through predictive modeling 23
• 2.5. Previous works that study the application of predictive modeling for policy impact 24
• 2.6. Previous works that study the application of predictive modeling for environmental systems 25
• 2.7. Previous works that study factors affecting CO2 emissions 25
• 2.8. Previous works that apply GWR technique for environmental impact investigation 36
• 2.9. Previous works that apply MGWR technique for environmental impact investigation 40
• 2.10. Advantages and limitations of GWR 43
• 3. METHODOLOGY 43
• 3.1. Research design and approach 43
• 3.2. Scope and area of study 44
• 3.3. Methods 46
• 3.4. Description of variables 49
• 3.4.1. Carbon Dioxide equivalent 49
• 3.4.2. Socioeconomic variables 50
• 3.4.3. Energy consumption and expenditure per sector 53
• 3.4.4. Land Cover 55
• 3.4.5. Hazardous waste 56
• 3.5. Data analysis and preprocessing 58
• 3.5.1. Data preparation and preliminary analysis 58
• 3.5.2. Spatial Interpolation of Total Releases 58
• 3.5.3. GWR Analysis 59
• 3.5.4. MGWR Analysis 59
• 4. RESULTS 61
• 4.1. Data preparation and preliminary analysis 61
• 4.1.1. Data preparation 61
• 4.1.2. GMW and MGWR analysis 61
• 4.2. Variable selection 62
• 4.2.1. Correlation Analysis 62
• 4.2.2. Stepwise Regression 62
• 4.2.3. Validation 63
• 4.2.4. Ensuring Robustness and Interpretability and Application in Environmental Science 63
• 4.3. Performance statistics 65
• 4.4. Characteristics of spatial distribution 66
• 4.5. Spatial heterogeneity of influencing factors of CO2 67
• 4.5.1. GWR 68
• 4.5.2. GWR Spatial distribution of regression coefficients by regions 74
• 4.5.3. MGWR 76
• 4.5.4. MGWR Spatial distribution of regression coefficients by regions 83
• 4.5.5. SHAP plot 84
• 5. DISCUSSION 85
• 5.1. Poverty and race features 85
• 5.2. CO2eq distribution 88
• 5.3. Spatial analysis of predictors and locations 90
• 5.3.1. GWR spatial distribution of regression coefficients by regions 93
• 5.3.2. MGWR spatial distribution of regression coefficients by regions 97
• 5.4. Big cities vs. small communities 98
• 5.5. Policy implications 103
• 6. CONCLUSION 105
• REFERENCES 107
• 국문초록 131
• ACKNOWLEDGMENTS 132