Faulty environmental data is not merely an academic problem. It can lead to flawed policies, wasted resources and regulatory gaps that put human health at risk. The fight against pollution and the effort to mitigate climate change depend on data that reflect the physical world. When official reporting misses what is happening on the ground, policy risks addressing only part of the problem.

Policymakers, investors and other decision-makers rely on the accuracy, completeness, consistency and timeliness of international data. Sustainability policies, scrutiny of international agreements and the allocation of billions of dollars in green finance draw on statistical reports produced by institutions and organizations such as the U.N., the United Nations Development Programme (UNDP), Eurostat, the European Environment Agency (EEA), the Food and Agriculture Organization (FAO) and Plastics Europe. Discrepancies between reported figures and conditions on the ground therefore deserve close attention.

Frameworks such as the International Monetary Fund’s (IMF) Data Quality Assessment Framework (DQAF) and the U.N.’s National Quality Assurance Frameworks (NQAF) place emphasis on whether statistical outputs are fit for use. Yet the complexity of pollution data, methodological differences between national reporting systems and corporate greenwashing can undermine their reliability. Fragmented or incomplete observational records also limit their usefulness in decision-making.

The stakes extend from greenhouse gas emissions and the trade in plastic waste to air-quality measurement, environmental protection expenditure and corporate environmental, social and governance (ESG) reporting. Data quality cannot be reduced to a mathematical measure. It is multidimensional, shaped by the interaction between data acquisition, user expectations, conformance and plausibility.

The System of Environmental-Economic Accounting (SEEA), adopted by the U.N. Statistical Commission, and the FAOSTAT Data Quality Framework provide important reference points for environmental statistics. A multidimensional approach to data quality is particularly important in monitoring agri-environmental indicators, including methane and nitrous oxide emissions.

More data, but not enough progress

Monitoring the Sustainable Development Goals (SDGs) provides a substantial test of environmental data quality. Within the framework’s 17 goals and 169 targets, the United Nations Environment Programme (UNEP) has assessed 92 environment-related indicators. Its "Measuring Progress" series examines both gaps in the available data and the environmental trends those data reveal.

Data availability and environmental progress are not the same thing. UNEP reported that the share of environment-related SDG indicators with sufficient data for analysis rose from 34% in 2018 to 42% in 2020 and 59% in 2022. For 2022, 38% of all 92 environment-related indicators showed positive trends indicating environmental improvement. This measures progress in environmental outcomes, not an improvement in the quality of the statistics themselves. UNEP’s assessment indicated that the world was not on track to meet the environmental ambitions of the 2030 Agenda for Sustainable Development.

Traditional statistical approaches can struggle to capture complex ecological interactions. Differences in national data quality assessment frameworks, outdated information, inconsistent measurement methods and inadequate institutional infrastructure can further weaken SDG reporting.

Environmental variables should not be assessed in isolation. Examining relationships among recycling rates (SDG 12.5.1), air quality (SDG 11.6.2) and marine pollution (SDG 14) requires sufficiently complete, disaggregated datasets. These relationships need to be investigated, rather than treated as an established sequence of cause and effect. To address data gaps, traditional methods should be complemented by remote sensing, in situ sensors and citizen science, with AI supporting the analysis of robust data.

What methane inventories miss

Methane reporting illustrates the consequences of gaps in environmental data. Methane (CH4) is a potent greenhouse gas with a much greater heat-trapping effect than carbon dioxide over a 20-year period. Studies have identified substantial differences between bottom-up emission inventories, including those submitted to the United Nations Framework Convention on Climate Change (UNFCCC), and top-down estimates derived from atmospheric observations.

Traditional bottom-up methods use emission factors to estimate emissions from industrial activities and equipment. These methods can underrepresent the large releases associated with equipment failures, leaks, isolation-valve leakage and abnormal operating conditions.

A study by Alvarez and colleagues, published in Science, demonstrated this problem in the U.S. oil and gas supply chain. It estimated methane emissions in 2015 at 13 ± 2 teragrams (Tg) a year, approximately 60% above the inventory estimate of the U.S. Environmental Protection Agency (EPA). Production, gathering and processing accounted for about 85% of the study’s national bottom-up estimate.

The researchers identified emissions released under abnormal operating conditions as a likely explanation for much of the gap. Such "tail-heavy” emission distributions mean that a small minority of "superemitters” can account for a large share of emissions. Inventories that fail to capture these events risk understating the scale of the problem.

Satellite observations offer an independent perspective. The Tropospheric Monitoring Instrument (TROPOMI) aboard the European Space Agency’s (ESA) Sentinel-5P satellite uses shortwave infrared (SWIR) measurements to observe atmospheric methane. Observations at spatial resolutions such as 5.5 times 7 kilometers can support emission estimates and comparisons with inventories.

In Canada’s Montney basin, a TROPOMI analysis of natural gas operations estimated methane emissions of 2.6 ± 2.2 kilotonnes a day, above the inventory estimate used in the study, though with substantial uncertainty. Pixels containing gas facilities showed methane concentrations averaging 11 parts per billion (ppb) above the background. These findings raise questions about how well inventories capture emissions in the area examined.

In Queensland, Australia, a TROPOMI analysis estimated that six coal mines emitted 570 ± 98 gigagrams (Gg) of methane a year in 2018-2019. Although these mines accounted for just 7% of national coal production, their estimated emissions were equivalent to about 55% of all coal-mining methane emissions Australia reported to the UNFCCC.

In South America, satellite-based inversion models provide another way to assess inventories compiled using Intergovernmental Panel on Climate Change (IPCC) guidelines. One analysis estimated annual anthropogenic methane emissions of 19 Tg for Brazil, 9.2 Tg for Argentina and 7.0 Tg for Venezuela. Such analyses also allow closer examination of individual sectors, including livestock and enteric fermentation, rather than relying solely on national totals.

These studies show why climate targets and net-zero projections need a sound baseline. Where inventories underestimate emissions, they can distort assessments of mitigation needs and weaken the evidence supporting carbon pricing mechanisms and emissions trading systems. Satellite-supported monitoring should be integrated into inventory verification, with clear regulatory requirements for reconciling estimates and investigating discrepancies.

When plastics accounting misleads

Plastic pollution presents a different set of measurement challenges. Differences in how countries calculate packaging waste and recycling rates can complicate comparisons even within the EU. Eurostat figures for 2023 put packaging waste across the EU-27 at 177.8 kg per person, with plastic accounting for 19.8% of the total, or 15.8 million tonnes.

Separately, a comparison of packaging-waste reporting using 2019 data put Ireland at 65 kilograms of plastic packaging waste per person, 87% above the EU average. That difference cannot simply be read as a difference in consumption. Ireland used Waste Analysis (WA), based on waste measurements, while most EU countries used Placed on the Market (PoM). The choice of method can affect the quantities reported.

PoM relies on industry declarations submitted under Extended Producer Responsibility (EPR) schemes. These declarations can contain gaps arising from under-reporting, freeriding and the exclusion of producers below specified thresholds under the de minimis rule. Differences in coverage can therefore contribute to lower reported quantities under PoM than under WA.

Such methodological differences limit the comparability of circular economy indicators and need to be addressed when interpreting European waste statistics, including reporting under WStatR. Without that scrutiny, incomplete reporting can obscure the scale of plastic waste.

Chemical recycling raises further questions about how recycling performance is measured. Industry promotes these processes for materials that are difficult to recycle mechanically, such as contaminated or multilayer plastics. Pyrolysis and gasification break down plastic waste through energy-intensive processes. Their inclusion in recycling statistics, particularly through the "Mass Balance Approach,” requires careful scrutiny of what the resulting figures represent.

In a petrochemical steam cracker, pyrolysis oil derived from plastic waste may be mixed with a much larger quantity of virgin fossil feedstock. Once the inputs are mixed and processed, tracing their respective contributions to an individual product is difficult. An accounting allocation is not the same as a direct measurement of that product’s recycled content.

Depending on the allocation rules, the recycled share of the input can be assigned to selected outputs, such as consumer packaging. This creates a risk of misleading claims when the distinction between an allocated share and physical recycled content is not made clear.

Allocation rules, including fuel-use exempt allocation, therefore need to be assessed against what they actually count and exclude. The absence of molecule-level traceability does not establish that a particular package contains no recycled material. Claims about recycled content should make their accounting basis clear, rather than imply a degree of physical certainty that the method cannot demonstrate.

Environmental organizations, including Zero Waste Europe, Rethink Plastic Alliance and NRDC, have challenged claims about the circularity and environmental benefits of chemical recycling. The calculation of carbon savings reported as avoided emissions in life cycle assessments (LCA) also warrants scrutiny. The question is whether the reported benefits are supported by the underlying methods and evidence.

Accounting methods that overstate recycling performance risk disadvantaging mechanical recycling and obscuring the distinction between material recycling and incineration. Clear definitions and transparent reporting are needed to prevent a recycling success story on paper from being mistaken for a demonstrated environmental benefit.

At the Geneva meeting of the Intergovernmental Negotiating Committee’s fifth session (INC-5.2), negotiations on the Global Plastics Treaty ended without agreement on a text. Clear definitions and comparable data remain essential to assessing policies on plastic production and recycling.

Environmental data should support scrutiny, not provide cover for greenwashing. Bottom-up inventories and industry declarations should be subject to cross-validation against independent top-down estimates and sensor-network measurements, with AI supporting the analysis. That requires scrutiny of how data are collected, processed and reported, not simply an increase in the volume collected.

Accurate data are more than a statistical output: they help determine the direction of environmental policy. A sound route is of little use when the instruments guiding it are unreliable. Transparency, accountability and physical traceability must become routine parts of environmental reporting. Progress against pollution should be demonstrated in the physical world, not merely recorded on paper.