A local-to-global emissions inventory of macroplastic pollution

The study addresses the need for a clear, high-resolution baseline of macroplastic waste emissions to support an internationally legally binding Plastics Treaty. Plastic pollution persists in the environment and impacts ecosystems, infrastructure, society and the economy. Prior global efforts and UN processes underscore the value of emissions inventories in guiding policy and monitoring progress. The authors clarify the concept of ‘emissions’ for plastics as material moving from controlled (managed or mismanaged) systems to uncontrolled environments, distinguishing debris (>5 mm) from open burning (mass combusted in uncontrolled fires). The research aims to quantify and localize macroplastic emissions worldwide, identify hotspots and sources, and provide actionable data to inform policy and interventions.

Previous global models typically report plastics reaching aquatic environments (oceans, rivers) or use aggregate terms like ‘mismanaged plastic waste’ and ‘riverine outflows’. They often employ top-down approaches using regional averages, lack sub-national resolution, and do not explicitly represent emission mechanisms or open burning. Some comparable studies (e.g., Ryberg et al., Lau et al.) report differing totals owing to scope differences (inclusion of sea-based, industrial, construction sources, and microplastics) and methodological assumptions (expert-derived emission factors). Emission inventories have been foundational in climate and mercury agreements, highlighting the need for robust, standardized baselines for plastics. There is growing recognition that sub-national heterogeneity, behavioral factors (e.g., littering), and local waste management performance must be incorporated for credible estimates.

The authors developed a global, bottom-up emissions inventory at municipal scale (50,702 municipalities; baseline year 2020) within the SPOT model framework. Steps: (1) obtain, clean and harmonize municipal-level solid waste management data (generation, collection coverage, system performance, littering, street sweeping, treatment pathways including recycling—formal/informal—controlled and uncontrolled disposal, incineration, sorting/reprocessing, rejects); (2) predict missing waste management variables for all municipalities using machine learning; (3) apply probabilistic material flow analysis (MFA) to quantify plastic waste flows per municipality; (4) use conceptual sub-models to quantify emissions from five land-based sources—uncollected waste; littering (uncollected litter); collection system; uncontrolled disposal (dumpsites); rejects from sorting and reprocessing—considering behavior, waste system performance, and key phenomena (e.g., meteorological mobilization, surface runoff, dumpsite dynamics, street sweeping mitigation); (5) aggregate results bottom-up to country, regional and global levels, including corrections for rurality and settlement typology. Emissions are defined as transitions from contained to uncontained states and are categorized as debris (>5 mm) or open burning (mass of waste burned in uncontrolled conditions). The model excludes plastic retained within dumpsite masses (buried) and sea-based/industrial/construction sources; only dumpsite ‘working faces’ are treated as potential debris emitters. Open burning estimates rely on census/survey activity data from 44 countries rather than solely on generic emission factors. Uncertainty is handled with probabilistic inputs (reporting means with 5th–95th percentiles), propagating through municipal-level distributions and aggregated summaries.

• Global macroplastic emissions in 2020 are estimated at 52.1 Mt year−1 [48.3–56.3], about 21% of municipal plastic waste generated (251.7 Mt [233.1–272.4]).
• By type: 57% open burning (29.9 Mt [27.6–32.4]) and 43% debris (22.2 Mt [20.6–24.0]).
• Absolute country hotspots: India 9.3 Mt [6.5–12.7]; Nigeria 3.5 Mt [2.6–4.6]; Indonesia 3.4 Mt [2.5–4.3]; China 2.8 Mt [2.1–3.7]. China ranks fourth, reflecting recent expansion of controlled landfill and incineration.
• In India: modelled collection coverage 81% [80–82]; 53% wt. [51–56] of emissions arise from uncollected waste (30% wt. debris; 23% wt. open burning). Estimated open burning: 56.8 Mt MSW [40.0–77.7], including 5.8 Mt plastic [4.1–7.9].
• Regional patterns: Southern Asia emits 15.1 Mt [12.1–18.7]; Sub-Saharan Africa 13.3 Mt [12.0–14.7]. Urban and semi-urban settlements dominate emissions due to higher populations and waste generation.
• Income groups: HICs generate more plastic waste per capita (0.17 kg cap−1 day−1 [0.15–0.20]) but contribute only 0.3% (0.16 Mt [0.14–0.19]) of global emissions owing to near-universal collection and controlled disposal; none are in the top 90 emitters. Plastic waste export emissions were excluded; estimated effect is negligible (~0.03 Mt year−1 from top OECD exporters to non-OECD + Turkey in 2022).
• Per-capita examples: China 1.97 kg cap−1 year−1 [1.48–2.61] (rank 153); India 6.64 [4.66–9.08] (rank 127); Russia 11.71 [7.80–16.17]. Sub-Saharan Africa average 12.01 [10.83–13.25], indicating potential future dominance in absolute terms as populations grow.
• Municipal distributions show wide uncertainty but stark differences: Hamburg median 0.02 kg cap−1 year−1 [0.01–0.06] vs Mogadishu 13.63 [4.05–36.70]; Agra 0.77–11.87 (5th–95th) and Maracaibo 0.11–4.72.
• Source attribution: Uncollected waste is the largest source globally—68% of total (35.6 Mt), and 85% (18.7 Mt) of debris emissions. In LMCs (mean collection 74% [72–75]), uncollected waste drives 56% (19.9 Mt [17.8–22.3]) of uncollected-waste emissions, accounting for 38% of the global total and 51% (11.3 Mt) of debris.
• Uncontrolled disposal (dumpsites) contributes 25% (12.8 Mt [11.5–14.3]) of global emissions, 98% of which is open burning; only ~0.25 Mt debris (~0.4% of plastics deposited) is emitted from dumpsites.
• In HICs, littering is the dominant debris source, accounting for 53% of debris emissions and 49% (0.08 Mt; ~0.06 kg cap−1 year−1) of all macroplastic emissions in the Global North; ~0.03 Mt each in Northern America and the Rest of Europe.
• Material type: In the Global South, flexible plastics constitute a higher share of debris emissions (56:44 flexible:rigid), whereas in the Global North rigid plastics dominate (33:67), reflecting differences in waste composition and drivers (littering vs meteorological mobilization).
• Concentration: 69% (35.7 Mt) of global emissions originate from 20 countries (4 LICs, 9 LMCs, 7 UMCs).

The inventory delivers a high-resolution, bottom-up understanding of macroplastic emissions, directly addressing the need to identify where, how, and why plastics leave managed systems. By disaggregating five emission sources and distinguishing debris from open burning, it clarifies mechanisms and prioritizes interventions. Findings show uncollected waste is the primary global driver—especially in LMCs and LICs—implying that expanding and improving collection services would yield the greatest reductions, more than mitigating all other land-based macroplastic sources combined. Open burning emerges as a larger mass pathway than debris globally, underscoring overlooked health and climate risks and the necessity to explicitly include open burning in Plastics Treaty negotiations. High per-capita burdens in Sub-Saharan Africa highlight where rapid investments and policy reforms could preempt large future absolute emissions. The inventory’s municipal resolution allows policymakers to distinguish extreme hotspots from well-performing areas, guiding targeted action plans, resource allocation, and monitoring of progress over time.

This work introduces a global, municipal-scale, bottom-up macroplastic emissions inventory that conceptualizes and quantifies five emission sources and two emission types for 2020. It advances beyond prior models by: employing sub-national data with machine learning gap-filling; using probabilistic MFA; explicitly modelling open burning and dumpsite emission mechanisms; and focusing on contained-to-uncontained transitions. Key insights include the dominance of uncollected waste and the substantial role of open burning. The inventory provides a practical baseline to support Plastics Treaty implementation, enabling national and sub-national action plans and evidence-based prioritization. Future work should expand high-quality municipal data collection (including rural areas), refine behavioral and system performance parameters, improve representation of informal sector activities, and continuously update the model structure and inputs. Integrating upstream measures (material and product redesign, source reduction) with systemic improvements in collection and disposal will be critical to achieving rapid emissions reductions.

• Data scarcity and quality: Robust, quality-controlled municipal activity data are limited, especially for rural areas, leading to broader uncertainty ranges and reliance on probabilistic assumptions and sub-models.
• Scope exclusions: Sea-based, industrial, and construction sources are outside scope; microplastics are not included; only the dumpsite working face is treated as a debris emitter, potentially differing from other modelling assumptions.
• Exports: Plastic waste export-related emissions were deliberately excluded, potentially underestimating emissions for some HICs, though the global impact is assessed as negligible in recent years (~0.03 Mt year−1 for 2022).
• Open burning estimation: While based on census/survey activity data from 44 countries, gaps remain in direct measurements; regional behaviors and practices may vary.
• Uncertainty propagation: Conservative probability density functions for inputs create wide municipal-level uncertainty distributions; some model parameters (e.g., littering behavior, informal dumping/burning practices) are difficult to measure directly.
• Temporal baseline: Estimates reflect year 2020 conditions; changes in policy, infrastructure, and behaviors since then are not captured.