Both contaminant families arrive in soil through the same gateways. Biosolids and compost spread on farmland, landfill leachate, irrigation with reclaimed wastewater, degrading agricultural films, atmospheric deposition, and fibers shed by fluorinated textiles all deliver PFAS and plastic particles to the same terrestrial reservoirs. Wastewater treatment plants are especially important crossroads because they concentrate both pollutants in the sludge later applied to land: a field study of linked landfill and wastewater systems detected microplastics and PFAS throughout the facilities, with both preferentially transferred to solid residuals such as biosolids. PFAS have even been measured on plastic materials in yard-waste compost, and at Taihu Lake in China the morphology and biofilm load of microplastics tracked the distribution of specific PFAS, including the emerging compound HFPO-DA. Yet the review’s authors separate co-occurrence from causation. Showing that two pollutants share a pathway does not prove that plastic transports the chemicals; that requires measurements of particle-associated PFAS, adsorption and desorption kinetics, changes in mobility, or altered biological uptake — data that remain rare in the field. Plastics can even act as secondary PFAS sources: side-chain fluorinated polymer coatings on textiles shed precursors that transform into mobile perfluoroalkyl acids.

The chemistry explains why no two PFAS–plastic pairings behave alike. PFAS owe their persistence to carbon–fluorine bonds among the strongest in organic chemistry, but their soil behavior comes from an amphiphilic architecture: a water-repelling fluorinated chain welded to a polar, usually negatively charged head group. At typical environmental pH, perfluoroalkyl carboxylic and sulfonic acids exist as anions, so the fluorinated tail seeks hydrophobic and dispersion-driven contacts while the charged head responds to electrostatic forces. Chain length sets the baseline: short-chain compounds stay dissolved and mobile in pore water, whereas long-chain homologs partition strongly onto organic matter, proteins, air–water interfaces and plastic surfaces. Functional groups add nuance. In one cited experiment, neutral perfluorooctanesulfonamide partitioned onto polyethylene, polystyrene and PVC far more readily than anionic PFOS of similar chain length, and sulfonates and sulfonamides generally out-adsorbed many carboxylates. Polymer identity matters just as much. Nonpolar polyethylene, polypropylene and polystyrene interact mainly through hydrophobic partitioning and dispersion forces, whereas polyamide’s amide groups enable hydrogen bonding and stronger polar interactions: a comparison of 18 polymers with eight PFAS found polyamide to be the strongest sorbent, with computed interaction energies matching the experimental trend. Modeling of PFOS on polypropylene nanoplastics likewise identified dispersion forces as dominant, with PFOS occupying both external and internal polymer domains.

The trouble is that such tidy results rarely survive contact with a real soil. Much of the literature rests on pristine, monodisperse beads suspended in water or packed into homogeneous quartz-sand columns at contaminant concentrations far above environmental levels. Real plastic debris is irregular, additive-laden and mineral-filled — commercial plastics containing talc and glass fibers sorbed more pollutants than pure polymers — and plastics incubated in a lake accumulated substantially more PFAS than clean particles exposed in laboratory water. The review’s most striking counterexample comes from the only core study that compared isolated particles, soil and a soil–plastic mixture directly. Polyamide microplastics initially showed enormous affinity for PFOA, PFHxS, 6:2 fluorotelomer sulfonic acid and GenX, but mixing them into paddy soil cut their adsorption capacity roughly sevenfold to 108-fold — an inhibition of 22 to 97 percent — as scanning electron microscopy revealed mineral grains and biofilm-like material smothering the polymer surface. Equilibrium adsorption stretched from about 24 hours for the isolated polymer to roughly 96 hours in the mixture, betraying mass-transfer bottlenecks absent from simplified systems. The authors frame the resulting behavior in three regimes — plastic-dominated, matrix-dominated and coating-mediated — between which any given particle may drift as it weathers, ages and acquires coatings.

Transport experiments deliver the same conditional verdict. For a particle to act as a PFAS vector, the chemical must stay bound while the particle itself migrates; adsorption onto an immobile particle does the opposite, retarding movement. In saturated quartz-sand columns, PFOA adsorption dampened the negative surface charge of negatively charged polystyrene microplastics, increasing their retention, while it softened the positive charge of aminated particles, boosting their mobility — yet both particle types reduced PFOA transport, because the plastics were less mobile than the dissolved acid. Ionic strength and cation type can flip the outcome: divalent calcium retained polystyrene particles far more effectively than sodium through charge neutralization and possible cation bridging, and at pH 5, competition between PFOA and microplastics for attachment sites freed the chemical to break through faster. Natural sediments complicate matters further. There, tiny 0.13-micrometer polystyrene particles sailed through while larger fragments were strained out, and the sediment buffered acidic influent toward neutrality, erasing pH effects engineered at the inlet. In another sediment study, highly mobile 0.1-micrometer nanoplastics left PFOA transport essentially unchanged, with more than 91 percent of the chemical remaining unbound, while larger 5-micrometer microplastics produced only a small increase in retention. The same polymer–PFAS pair can therefore facilitate, retard, or ignore transport.

Biology obeys the same law of contingency, and the terrestrial evidence remains startlingly thin. The sole core soil-organism study found that PVC microplastics raised the bioaccumulation factors of PFOA and PFOS in earthworms by up to 200 percent and, at doses of 500 and 1,000 milligrams per kilogram, suppressed juvenile production — direct evidence that ingested plastic can open an additional exposure route. But the particles were pristine, doses high, and only one soil and species were tested. Plant studies point in both directions. In hydroponic soybean sprouts, PFOS adsorbed onto polystyrene particles lowered the freely dissolved fraction and blunted acute phytotoxicity, even as the PFOS coating encouraged plant tissues to internalize the nanoplastics; in water hyacinth, polystyrene microplastics raised whole-plant bioconcentration of PFOA, PFOS, GenX and F-53B, yet the combined toxicity was antagonistic for the legacy compounds and synergistic for the alternatives. Aquatic models fill mechanistic gaps with warnings about particle design: smaller 20-nanometer particles drove roughly 3.2-fold increases in PFOA uptake by Pacific oysters versus 2.3-fold for 500-nanometer particles, positively charged nanoplastics caused greater bioaccumulation of the alternative PFAS F-53B in aquatic insect larvae than negative ones, and natural organic matter coating nanoplastics intensified chronic liver, gut and microbiome damage in adult zebrafish.

Environmental aging cuts both ways, which is precisely why the review refuses to generalize. Naturally weathered polystyrene showed higher partition coefficients for legacy and emerging PFAS than virgin particles, with some compounds detectable only on aged surfaces, and biofilm-covered polyethylene, polypropylene, polyester and polyamide microfibers adsorbed more PFOS while releasing it more slowly — evidence that extracellular polymeric substances create new binding domains that prolong retention. Yet mineral coatings and competitive adsorption can mask those same domains, as the polyamide experiment showed, and biodegradation of polylactic acid increased PFOA uptake by decorating its surface with hydroxyl, carbonyl and carboxyl groups. The review also flags an underappreciated twist: some PFAS found clinging to environmental plastics may never have been adsorbed at all. Fluorinated textile coatings, water-repellent fabrics and multilayer materials shed fibers that already contain residual or transformation-derived PFAS, and decades-scale degradation studies show side-chain fluorotelomer polymers slowly releasing perfluoroalkyl acids in soils and water. A plastic particle can therefore be a sorbent, a temporary reservoir, or a source, and telling these roles apart requires polymer analysis, precursor measurements and controlled desorption tests rather than simple detection.

Behind these conclusions lies a sobering map of where the science has and has not looked. Of the 148 publications that survived full-text screening, 45.9 percent fell into the Ecosystem domain and 33.1 percent into Water, while Health accounted for 10.8 percent and Air for 2.7 percent — leaving Soil with just 7.4 percent of the literature. Scientific output, which surged after 2020, concentrates heavily in China and the United States, limiting how well findings transfer to regions with different soils, climates and waste practices. The eight core studies reduce to five saturated column experiments, one batch soil adsorption test, one earthworm bioaccumulation study and one compost survey; polystyrene and PFOA dominate the particle–chemical matrix far beyond their real-world prevalence, and quartz sand, pristine spheres, short exposures and elevated concentrations remain the default. Analytical obstacles compound the problem: extracting microplastics from organic-rich soils without altering them, distinguishing submicron particles, and quantifying PFAS amid laboratory background contamination — much of it from PTFE labware — and severe LC–MS/MS matrix effects all push current methods to their limits.

The road forward, the authors argue, runs through field reality. Priority work includes monitoring PFAS and particle-associated fractions together across soil profiles and land uses — biosolid-amended fields, firefighting-foam sites, landfill surroundings and reclaimed-water-irrigated cropland — alongside intact soil cores that preserve macropores, aggregates and preferential flow, unsaturated and transient moisture conditions in which PFAS gather at air–water interfaces, particles conditioned in real soils before testing, environmentally realistic concentration gradients, and complete mass balances that track dissolved, soil-bound, plastic-associated and lost PFAS rather than nominal doses. Desorption deserves the attention adsorption has hoarded, because a particle only vectors contamination if the chemical stays aboard during transport and disembarks somewhere exposure occurs — including inside a gut. Broader panels of polymers and PFAS chemistries, direct tests of trophic transfer through terrestrial food webs, and models that treat PFAS and plastics as interacting but distinct phases round out the agenda. Until such data exist, the team advises treating the plastic vector role as a site-specific hypothesis rather than an inherent property of plastic pollution — because in most soils, organic matter and minerals, not plastic, still rule the fate of forever chemicals.

The investigation was underpinned by a meticulous and systematic literature review adhering to rigorous standards akin to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). By interrogating a dataset of over 35,000 publications sourced primarily from the Scopus database across a fifteen-year period, the team honed in on 385 meticulously selected studies—each integrating satellite-borne elevation data within coastal hazard contexts. This comprehensive approach targeted the most up-to-date research coinciding with the availability of global digital elevation models (DEMs) and mean dynamic topography (MDT) datasets, ensuring a consistent data horizon for meaningful comparison.

Central to the analysis was the often-overlooked aspect of vertical datum conversion, a process critical to aligning elevation data with mean sea level (MSL). The team found that many studies failed to properly document or apply the necessary conversion procedures, leading to significant underestimations of coastal elevations and, by extension, flood susceptibility. Vertical datum conversion entails not only recognizing the original vertical reference system of a DEM but also integrating auxiliary datasets such as geoid models and local sea-level indicators to calibrate elevation data precisely.

To empirically quantify these discrepancies, the researchers processed four state-of-the-art global DEMs—CoastalDEM v2.1, FABDEM v1.0, GLL-DTM v.2, and DeltaDTM v1—each originally referenced to different geoid models (EGM96 or EGM2008). By rigorously translating these DEMs into a common vertical framework using the latest hybrid MDT data (HYBRID-CNES-CLS2022), they reconciled variations in geoid heights and sea surface levels derived from satellite altimetry and gravitational field measurements. This conversion employed advanced spatial interpolation and smoothing techniques to extrapolate sea-level data over land, a necessary step given the coastal focus of the study.

The implications of aligning elevation datasets vertically to MSL were profound. When comparing unconverted and converted DEMs, stark differences emerged in estimates of land areas and populations at risk from a hypothetical one-meter relative sea-level rise (RSLR). The uncorrected models consistently underestimated exposure, sometimes by substantial margins, due to misaligned vertical references. The research extended beyond area metrics, incorporating an ensemble of three global population datasets—WorldPop, LandScan 2020, and LandScan 2023—to robustly quantify how many people currently reside below or near local MSL thresholds.

Further scrutiny uncovered a widespread and systematic issue: many influential scientific publications, some referenced extensively in the latest IPCC Assessment Reports, failed to incorporate precise sea-level datum referencing. Out of studies incorporated into the IPCC AR6 cycle, less than 15% demonstrated proper vertical datum integration. This revelation suggests that even well-recognized climate assessments may underrepresent coastal vulnerability globally, potentially skewing policy guidance and resource allocation.

Methodologically, the researchers employed the ArcGIS Pro platform to standardize transformations, employing bilinear resampling and inverse distance weighting algorithms to meticulously adjust DEMs for datum biases. The use of robust geoid models from the GFZ Helmholtz Centre and hybrid MDT products enhanced the geographic fidelity and temporal relevance of the analyses. This methodological rigor establishes a replicable framework that future coastal hazard assessments can adopt to improve accuracy.

A critical aspect of this work was the recognition that accurate vertical referencing transcends mere technical detail; it is foundational for correctly estimating low-elevation coastal zones (LECZ), which harbor some of the world’s densest human populations and most economically vital infrastructure. By recalibrating elevation data consistently with sea-level datasets, risk assessments can better capture subtle but consequential degrees of inundation potential, especially in vulnerable deltaic and estuarine environments.

Beyond the technical elevation adjustments, the researchers highlight the limitations inherent to current approaches to exposure modeling. Notably, their exposure statistics are conservative, focusing only on relative elevation without applying hydrodynamic inundation models or factoring in dynamic population growth in coastal zones. Recognizing these constraints, the study nonetheless represents a critical correction to the baseline data inputs that underlie numerous impact projections.

This breakthrough recalibration of coastal elevation data aligns with emerging calls for more transparent, data-driven, and reproducible methodologies within the climate science community. It provides an indispensable corrective lens to reconcile disparate elevation datasets and sea-level assumptions, setting a higher standard for environmental impact assessments and infrastructure planning.

The findings underscore an urgent need for widespread adoption of consistent vertical datum conversions and up-to-date sea-level referencing in coastal research. Given global population trends concentrating billions in vulnerable coastal regions, ignoring these methodological imperatives risks underestimating future hazards and compromising adaptive responses.

Moreover, this paradigm shift offers a clarion call for the IPCC and other scientific consortia to revisit foundational datasets informing policy models. The systematic review uncovered that nearly 90% of coastal hazard studies neglected adequate sea-level reference integration, suggesting that global hazard aggregation may be fundamentally flawed. Rectifying this will improve the precision of projections guiding climate resilience investments.

This research thus represents a milestone in coastal risk science, marrying satellite geodesy, geospatial analysis, and rigorous bibliometric assessment to expose a critical blind spot in existing hazard appraisals. It charts a path forward for harmonizing global elevation data with real-world sea-level dynamics—an essential step in safeguarding vulnerable coastal populations in the Anthropocene.

As coastal impacts accelerate worldwide, the precision of elevation and sea-level datasets emerges as a linchpin in effective climate adaptation. By driving awareness of vertical datum discrepancies and establishing best-practice conversion protocols, this study empowers scientists, policymakers, and planners to craft more accurate and actionable hazard maps.

The broader implications extend to urban planning, disaster risk reduction, and insurance modeling, where marginal differences in elevation measurement can translate into huge shifts in exposure and vulnerability profiles. The refined datasets and conversion workflows pioneered here are poised to become standard tools for next-generation coastal assessments.

Ultimately, this comprehensive reevaluation challenges prevailing narratives of coastal safety and vulnerability, advocating a recalibration of risk perspectives grounded in rigorous geospatial fidelity. As the world grapples with rising seas, such heightened accuracy in elevation referencing represents a critical advancement in our collective capacity to anticipate and mitigate coastal hazards.

Subject of Research: Coastal elevation accuracy, vertical datum conversion, sea-level reference frames, global digital elevation models, coastal hazard assessment

Article Title: Sea level much higher than assumed in most coastal hazard assessments

Article References:
Seeger, K., Minderhoud, P.S.J. Sea level much higher than assumed in most coastal hazard assessments. Nature (2026). https://doi.org/10.1038/s41586-026-10196-1

DOI: https://doi.org/10.1038/s41586-026-10196-1