An extensive inventory of municipal open waste burning nationwide based on machine learning analysis
نویسندگان
1 Department of Environmental Engineering, Faculty of Engineering, Universitas Diponegoro, Semarang, Indonesia
2 Department of Electrical Engineering, Universitas Negeri Malang, Jl. Semarang No. 5, Malang, Indonesia
3 Environmental Sustainability Research Group, Department of Environmental Engineering, Faculty of Engineering, Universitas Diponegoro, Semarang, 50275 Indonesia
4 Carbon Neutrality Promotion Division, Institute of Environmental Science and Technology, The University of Kitakyushu, Kitakyushu City, 808-0135 Japan
5 Program Study of Engineering and Environmental Management, College of Vocational Studies, Institut Pertanian Bogor, 16151, Indonesia
6 Geographic Information Sciences and Technology Research Group, Faculty of Earth Sciences and Technology, Institut Teknologi Bandung, Bandung 40132, Indonesia
7 Sustainable Business School, London South Bank University, 103 Borough Road, London SE1 0AA, United Kingdom
8 Environmental Sustainability Research Group, Department of Environmental Engineering, Faculty of Engineering, Universitas Diponegoro, Semarang, 50275 Indonesia
doi
10.22034/gjesm.2026.02.14چکیده
BACKGROUND AND OBJECTIVES: Residential open waste burning remains a major environmental and climate challenge in Indonesia. Th study objectives were to provide a nationwide evaluation of open burning by quantifying its extent, identifying spatial patterns, analyzing material composition, and uncovering behavioral drivers. The ultimate objectives of this study were to generate evidence-based insights that can inform targeted policies and interventions to reduce open waste burning and its associated impacts. METHODS: The study integrates multiple approaches, including national datasets, a large-scale household survey, spatial analysis, and topic modelling of qualitative responses. Waste burning activities were estimated using an Intergovernmental for Climate Change-based activity and emission inventory framework, with a key refinement incorporating official Indonesian administrative village–city classifications to more accurately assign uncollected waste fractions at the subnational level. Primary data were collected via an online questionnaire, yielding 722 valid responses distributed across major regional clusters in Indonesia, including Java, Sumatra, Kalimantan, Sulawesi, Bali–Nusa Tenggara, and eastern regions, characterizing household waste generation, disposal practices, and behavioral drivers. FINDINGS: The results indicate that in the reference year, approximately 26.6 million tonnes of municipal waste were estimated to be openly burned in Indonesia, corresponding to about 23.3 million tonnes of carbon dioxide equivalent emissions. Residential burning accounts for about 70 percent of the total waste burned far exceeding contributions from landfill fires. Spatial analysis revealed strong clustering of open waste burning in densely populated provinces, particularly on Java Island, with notable intra-provincial variation at the district level. Material composition analysis showed that dry and combustible fractions, such as plastics, paper, wood, and garden waste, were disproportionately represented in burned waste. Latent Dirichlet Allocation (LDA) topic modelling identified inadequate waste collection infrastructure and household convenience as dominant drivers of open burning. CONCLUSION: This study demonstrates that open burning is a widespread and critical issue in Indonesia, with significant climate and environmental consequences. Given that residential burning accounts for about 70% of the 26.6 million tonnes burned annually, targeted improvements in household waste collection services offer strong potential to reduce emissions.
