The difficulty stems from the sheer breadth of scales involved in soot formation. At the molecular level, small hydrocarbon radicals a few angstroms across react to build polycyclic aromatic hydrocarbons, the well-established precursors of soot. These molecular clusters evolve through physical or chemical nucleation into soot nuclei smaller than two nanometers, which then grow by surface reactions into primary particles. Within microseconds to milliseconds, those primary particles coagulate into fractal-like aggregates exceeding one hundred nanometers. Meanwhile, the reactive flow inside a combustor spans dimensions on the order of meters, with particle residence times measured in seconds. Capturing all of this in a single simulation has traditionally forced researchers to choose between fidelity and feasibility.

The most accurate traditional approach relies on sectional population balance models, which divide the soot particle population into many size bins and solve transport equations for each one. These models deliver excellent agreement with experiments but carry a computational cost that grows with the number of sections, rendering them impractical for multidimensional simulations of realistic engine geometries. The method of moments offers a cheaper alternative by tracking only a few statistical properties of the distribution, but it depends on closure assumptions that lose accuracy when the true particle size distribution deviates from the assumed shape. Semi-empirical correlations, such as the widely used Moss-Brookes model, are fast and simple but were calibrated for specific flame conditions and do not generalize reliably across different fuels and operating regimes.

The Melbourne team’s framework threads this needle by adopting a monodisperse particle dynamics formulation, which represents the soot population at any location by a single average particle size. Rather than resolving the full spread of sizes, the model tracks only three quantities: the total particle number density, the total carbon molar concentration contained in soot, and a surface area reconstructed from those two using a morphological relationship governed by a surface fractal dimension. Although this simplification can introduce deviations of up to roughly thirty percent relative to fully polydisperse treatments, the soot population rapidly approaches a self-preserving size distribution with an asymptotic collision frequency, so the average aggregate characteristics remain reasonably reliable at a fraction of the computational cost.

The truly novel ingredient, however, is where the model gets its chemistry. Instead of deriving nucleation and surface growth rates from detailed gas-phase kinetic mechanisms, which require calculating the concentrations of many polycyclic aromatic hydrocarbon species through large reaction networks, the researchers plugged in rate expressions extracted directly from reactive molecular dynamics simulations. The nucleation rate was obtained from simulations of isothermal acetylene pyrolysis at temperatures of 1200 to 1800 kelvin and high pressures of 189 to 568 atmospheres, monitoring the emergence of incipient soot clusters across a range of initial acetylene concentrations. The surface growth rate was similarly derived from simulations of incipient soot nanoparticles growing in acetylene pyrolysis, yielding an Arrhenius-type temperature dependence. Because these lumped rates depend only on acetylene concentration and temperature, they demand no prior knowledge of specific reaction intermediates or precursor species.

To benchmark the framework, the team simulated a premixed ethylene burner-stabilized stagnation flame, a well-characterized configuration featuring a porous plug burner, nitrogen shielding, and a cooled stagnation plate. The computational domain, reduced to a two-dimensional axisymmetric slice, was discretized into 46,400 quadrilateral cells with resolution down to about fifty micrometers near the inlet and plate. Combustion chemistry was handled with the reduced DRM22 mechanism comprising 22 species and 104 reactions, which the authors first validated against experimental species profiles in a comparable ethylene flame, confirming accurate predictions of oxygen and ethylene consumption and of water, carbon monoxide, hydrogen, and carbon dioxide formation. The predicted centerline temperature profile, peaking near 1700 kelvin a few millimeters above the burner, matched measurements closely.

The validation against soot measurements revealed a nuanced picture. In the baseline configuration, combining the semi-empirical Moss-Brookes nucleation rate with the classical hydrogen abstraction-carbon addition surface growth mechanism, the model predicted a soot number density within about six percent of a detailed sectional model from the literature and captured the measured soot volume fraction to within an order of magnitude. Notably, swapping in the molecular dynamics-derived nucleation rate while retaining the classical surface growth treatment improved the volume fraction prediction by roughly fifty-seven percent, bringing it within four percent of the sectional model’s result. Among all the reduced models tested, this molecular dynamics-informed nucleation variant reproduced the measured soot mobility diameters most accurately, particularly in the post-flame region beyond six millimeters above the burner.

The study also exposed instructive failures. Models employing the molecular dynamics-derived surface growth rate underpredicted soot loading by about two orders of magnitude, because the effective surface growth constant they carry is approximately three times lower than that of the hydrogen abstraction-carbon addition correlation, throttling carbon accumulation on particle surfaces. Conversely, the standard Moss-Brookes model as implemented in commercial software overpredicted the soot volume fraction by nearly three orders of magnitude, owing to an excessive surface growth contribution and an assumption that particles coagulate only by full coalescence into spheres, ignoring the fractal aggregates that real soot forms. Near the burner, at a plate separation of four millimeters, all reduced models overpredicted measurements by two to three orders of magnitude, a discrepancy the authors attribute partly to simplified nucleation chemistry and partly to sub-two-nanometer soot nuclei that standard condensation particle counters cannot detect.

For the authors, the significance lies less in any single number than in the demonstration that molecular dynamics-derived rate constants can be systematically assessed and deployed inside a CFD-coupled soot model, something they describe as the first such systematic evaluation. The framework captures soot volume fraction and particle size with accuracy comparable to detailed sectional modeling in the post-nucleation flame region, at dramatically lower computational cost, and without relying on reaction kinetic modeling of the precursor chemistry. That combination makes it a candidate tool for design-oriented simulations of engines, industrial burners, and aerosol synthesis reactors, where engineers need fast, trustworthy estimates of particulate emissions. The team cautions that the molecular dynamics rate constants were derived within a specific temperature and pressure window, and that additional benchmarking against other flame configurations will be necessary before the framework can be applied more broadly. Still, as regulators tighten limits on black carbon and the aviation industry confronts its soot footprint, a model that trades exhaustive chemistry for molecular-level insight at a fraction of the cost may prove exactly what practical combustion design has been waiting for.

Subject of Research: Multiscale simulation of soot particle nucleation, surface growth, and coagulation in flames using molecular dynamics-informed particle dynamics coupled with computational fluid dynamics

Article Title: Soot growth by monodisperse particle dynamics coupled with computational fluid dynamics

Article References: Fakharnezhad, A., Berry, J. D., & Goudeli, E. (2026). Soot growth by monodisperse particle dynamics coupled with computational fluid dynamics. Aerosol Research, 4(2), 279-291. https://doi.org/10.5194/ar-4-279-2026

Image Credits: AI Generated

DOI: 10.5194/ar-4-279-2026

Keywords: soot, combustion, computational fluid dynamics, molecular dynamics, particle dynamics, aerosol research, nucleation, surface growth, coagulation, ethylene flame, population balance, emissions modeling

The research, led by Johannes Heuser and Claudia Di Biagio of the Laboratoire Interuniversitaire des Systèmes Atmosphériques (LISA) in France, together with colleagues from institutions across France and Italy, was carried out in a large atmospheric simulation chamber. The team generated soot aerosols under carefully controlled combustion conditions, producing particles that ranged from nearly pure black carbon to organic-rich material dominated by brown carbon. By systematically varying the maturity and physico-chemical character of the particles, they created a family of laboratory soots spanning the diversity found in real emissions, from diesel-style flames to smouldering biomass.

The quantity at the heart of the study is written m = n − ik. The real part, n, governs scattering: how much light bends and deflects when it encounters a particle. The imaginary part, k, governs absorption: how much light energy is converted to heat. Both matter enormously for climate, because the balance between scattering and absorption determines whether an aerosol cools or warms the planet. For soot, k is among the largest of any atmospheric aerosol, which is why black carbon is generally considered the most important light-absorbing particle in the atmosphere — and why getting n and k right is essential for satellite retrievals, radiative transfer models and climate projections alike.

Retrieving these values from measurements is not straightforward. The team measured an extensive suite of particle properties inside the chamber: size distributions, morphological parameters, density, and the particles’ scattering, absorption and extinction coefficients across wavelengths from 450 to 630 nanometres. They then fed these inputs into two very different optical theories. The first, Mie theory, is the classical workhorse of aerosol optics. It computes exactly how light interacts with homogeneous spheres — a convenient but physically inaccurate picture of soot, which in reality forms branched, fractal aggregates of tiny spherules, like microscopic bunches of grapes.

The second approach, Rayleigh–Debye–Gans theory for fractal aggregates, or RDG-FA, was designed with that reality in mind. Instead of pretending each particle is a sphere, it treats the aggregate as a collection of small interacting units whose collective scattering and absorption depend on the fractal geometry of the whole structure. In principle, RDG-FA should describe soot more faithfully. In practice, it requires assumptions about particle morphology — the fractal dimension, the number and size of the primary spherules — and those assumptions carry their own uncertainties. The French-led team wanted to know how much each theory’s internal limitations would distort the refractive indices it returned.

To answer that question quantitatively, the researchers turned to Monte Carlo simulations. Rather than computing a single refractive index from a single set of inputs, they repeatedly perturbed the measured input parameters within their experimental uncertainties and recalculated the retrieval thousands of times. The resulting spread of output values reveals how sensitive each method is to errors in size, morphology, density and the optical coefficients themselves. This kind of uncertainty propagation is rare and laborious, but it transforms the retrieved numbers from bare point estimates into statements with defensible error bars — exactly what modellers need when deciding which laboratory values to adopt.

The headline finding is striking: both theories can reproduce the measured absorption and scattering of the chamber soot, yet the refractive indices they retrieve differ substantially. For soot whose elemental-to-total carbon ratio ranged from 0.79, representing black-carbon-dominated particles, down to 0, representing brown-carbon-dominated material, the Mie-based retrievals spanned real parts n of 1.35 to 2.68 and imaginary parts k of 0.38 to 0.31. The RDG-FA retrievals gave a narrower and generally different picture: n between 1.68 and 2.07 and k between 0.20 and 0.09. In other words, the same particles, measured in the same chamber, yield meaningfully different optical fingerprints depending on the theory used to interpret the data.

The Monte Carlo analysis explains why. Mie-based retrievals proved only weakly sensitive to ordinary measurement uncertainties in the input parameters, but they were strongly affected by particle coagulation and size growth inside the chamber. As soot aggregates collide and stick together, the spherical approximation underlying Mie theory degrades, and the retrieved refractive index absorbs the error. The RDG-FA retrievals, by contrast, remained stable even as the particles coagulated, because the theory naturally accommodates growing aggregates — but they depended strongly on the uncertainty of the assumed particle morphology. Each method, it turns out, has a characteristic weak point: Mie stumbles on shape, RDG-FA on assumed structure.

The practical implication is a call for consistency. Climate models, remote-sensing algorithms and laboratory databases often quote a single refractive index for soot without specifying which optical theory was used to derive it. This study shows that such a number cannot be treated as theory-independent. A refractive index retrieved with Mie theory should only be applied in models that also represent soot as spheres; an RDG-FA-derived value belongs with models that explicitly treat fractal aggregates. Mixing the two — plugging a Mie-retrieved k into an aggregate-based radiative calculation, for example — risks compounding errors in the estimated warming effect of black carbon, one of the largest single uncertainties in the human perturbation of the climate system.

The work also carries a broader lesson about how laboratory science handles imperfect models. Rather than asking which theory is simply correct, the researchers mapped where each one breaks down and quantified the consequences. The data sets and software underlying the retrievals have been released with persistent identifiers, allowing other groups to test alternative morphological assumptions or extend the analysis to other wavelengths and soot types. As the preprint moves through peer review, its central message is already clear: the light-absorbing power of soot is not just a property of the particle, but of the partnership between the particle and the physics used to describe it. For a substance with an outsized role in atmospheric warming, that partnership deserves the closest possible scrutiny.

Subject of Research: Retrieval of the complex refractive index of flame-generated soot aerosols using Mie and RDG-FA optical theories in an atmospheric simulation chamber

Article Title: Complex refractive indices of flame soot from different combustion conditions retrieved using Mie and RDG-FA theories: a simulation chamber study

Article References: Heuser, J., Di Biagio, C., Yon, J., Cazaunau, M., Bergé, A., Pangui, E., Zanatta, M., Renzi, L., Marinoni, A., Yu, C., Chevaillier, S., Ferry, D., Maillé, M., Formenti, P., Picquet-Varrault, B., & Doussin, J.-F. (2026). Complex refractive indices of flame soot from different combustion conditions retrieved using Mie and RDG-FA theories: a simulation chamber study. https://doi.org/10.5194/ar-2026-31

Image Credits: AI Generated

DOI: 10.5194/ar-2026-31

Keywords: soot, black carbon, refractive index, Mie theory, RDG-FA, fractal aggregates, aerosol optics, simulation chamber, light absorption, Monte Carlo uncertainty, combustion aerosol, climate modelling

The motivation behind the work stems from a fundamental tension in the race toward net-zero emissions. Electrification is advancing rapidly, but internal combustion engines remain indispensable for heavy-duty transport, from long-haul trucking to shipping and agriculture, where batteries struggle to deliver the energy density these applications demand. Renewable synthetic fuels, produced by combining renewable electricity with captured carbon dioxide through processes such as thermal and electrocatalysis, offer a compelling alternative because they are compatible with existing fuel infrastructure and can store energy over long periods. Yet if these fuels are simply dropped into engines designed around fossil diesel, much of their potential is wasted. The researchers argue that fuels and engines must be co-optimized, with the fuel’s properties deliberately shaped to match the combustion system that will burn them.

Conventional fuel design falls into two camps, and both have limits. Passive fuel design involves blending components in fixed ratios before the engine runs, while active fuel design dynamically adjusts injection ratios during operation. Both are composition-driven: they can only optimize within the limited palette of candidate components on hand. Renewable synthetic fuels such as Fischer-Tropsch synfuels break through this constraint because their composition can be tuned at the synthesis stage itself, through adjustments in feedstock, catalyst selection, and reaction conditions. But this expanded freedom creates a combinatorial nightmare. Synthetic fuels can contain dozens to hundreds of components, and the number of possible mixtures explodes beyond what trial-and-error experimentation or brute-force simulation could ever explore.

The team’s solution draws on an idea known as the central fuel property hypothesis, developed under the US Department of Energy’s Co-Optima program. The hypothesis holds that the macroscopic combustion behavior of a fuel is dictated primarily by a small set of key physicochemical properties rather than by its exact molecular makeup. If that is true, the vast space of possible fuel compositions collapses into a compact set of property variables, and those properties can serve as a bridge between what a fuel is made of and how an engine performs. The researchers selected five properties for conventional diesel combustion: density, which governs spray penetration and mixture formation; cetane number, which controls autoignition reactivity and ignition delay; lower heating value, which determines the energy released per kilogram of fuel; oxygen content, which influences oxidation rates; and the yield sooting index, a measure of a fuel’s intrinsic tendency to produce soot.

This property-based representation also solves a second, equally stubborn problem: data scarcity. Unlike biomedical science, materials science, or geophysics, engine research has no large open-source datasets, and fuel composition data from different studies cannot be meaningfully merged because the fuels themselves differ so wildly, from gasoline and diesel to alcohols, ethers, and biodiesel with various additives. By converting every fuel into the same five physicochemical properties, the framework makes heterogeneous data from different laboratories comparable. The team mined published experimental studies on conventional diesel combustion engines, standardized units, applied domain knowledge to derive missing variables, and assembled a standardized dataset of 986 samples spanning diverse engine platforms, fuel types, and operating conditions. Because the resulting distributions were skewed and imbalanced, the researchers used a structured splitting strategy combining k-means clustering with label-stratified sampling to ensure that training, validation, and test folds all faithfully represented the full dataset.

The predictive model at the heart of the framework is architecturally inventive. Engine experiments in the literature fall into three intake modes: simulated boosting, naturally aspirated operation, and exhaust gas turbocharging, and intake conditions are frequently not reported for the latter two. To handle this variability, the team built a conditional expert architecture inspired by the mixture-of-experts concept. A Transformer encoder serves as a shared backbone, learning latent representations of the common input variables, while intake-mode-specific multilayer perceptron expert modules are dynamically activated depending on which intake mode applies. Crucially, the model embeds physics directly into its learning. The work produced per cycle equals the mean effective pressure multiplied by cylinder displacement, so thermal efficiency can be independently inferred from the predicted pressure. The deviation between the model’s directly predicted efficiency and this physics-derived value is penalized during training and used during inference to select, among an ensemble of five sub-models, the prediction with the smallest physical inconsistency.

The model’s performance on an independent test set was impressive: a coefficient of determination of 0.994 for mean effective pressure and 0.921 for thermal efficiency. With this predictive engine in hand, the researchers deployed the NSGA-II multi-objective optimization algorithm to search for fuel property combinations that would maximize efficiency in a naturally aspirated, four-cylinder diesel engine running at 1,800 rpm under a specified load. The optimization produced a Pareto front of solutions, from which the team retained those with physical consistency deviations below one percent and selected the most efficient. The winning combination differed from diesel in telling ways: slightly lower density, a modestly reduced cetane number, a lower heating value, and most dramatically, an oxygen content of 14.55 percent by weight and a yield sooting index of just 13.06, compared with diesel’s 125.82.

Computational fluid dynamics simulations revealed why these properties work together so effectively. The lower cetane number lengthens the ignition delay, giving more time for the lower-density fuel to evaporate and mix with hot air, which increases the proportion of premixed combustion. The higher oxygen content accelerates oxidation, so combustion starts later but proceeds faster, concentrating heat release and shortening the combustion duration. The reduced heating value smooths the pressure rise that this faster combustion would otherwise produce. Meanwhile, the shrunken fuel-rich zones and enhanced oxidation slash soot formation, and the shorter persistence of high-temperature reaction zones curtails the thermal residence time that generates nitrogen oxides. The efficiency gain, the soot reduction, and the emissions cut all emerge from the same coordinated set of property changes, which is precisely the kind of synergistic design that composition-driven blending struggles to find.

The framework’s implications reach well beyond a single simulation. The optimized property combination can serve as a concrete target for the synthesis stage of renewable fuel production, guiding catalyst and feedstock choices rather than leaving fuel developers to grope through unguided trial and error. The researchers are candid about limitations: the dataset, though carefully constructed, remains modest; the demonstration covered one engine and one operating condition; and the optimized fuel was represented by conventional surrogate components rather than an actual synthetic fuel. Extending the framework to spark-ignition engines would require swapping cetane number for octane number, while other combustion modes and aviation fuels would demand additional property descriptors such as volatility, auto-ignition temperature, or aromatic content. Future work aims to validate the property targets in real engine experiments and to close the loop from design through synthesis to testing. If that loop can be closed, the era of fuels engineered atom-by-atom for the engines that burn them may be closer than it appears.

Subject of Research: AI-driven property-oriented design of renewable synthetic fuels for sustainable internal combustion engines

Article Title: An AI-driven framework for property-oriented fuel design toward sustainable engine combustion

Article References: Liu, Y., Li, A., Li, Y., Fan, Y., Qi, Y., Zhang, C., Zhu, L., & Huang, Z. (2026). An AI-driven framework for property-oriented fuel design toward sustainable engine combustion. iScience, 29(10), Article 117677. https://doi.org/10.1016/j.isci.2026.117677

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117677

Keywords: artificial intelligence, fuel design, internal combustion engines, renewable synthetic fuels, diesel combustion, machine learning, thermal efficiency, emissions reduction, Fischer-Tropsch, soot, nitrogen oxides, Co-Optima

Black carbon, the fine particulate matter produced by incomplete combustion of fossil fuels, biomass and other carbonaceous fuels, is one of the most important short-lived climate forcers in the atmosphere. Unlike carbon dioxide, which persists for centuries, black carbon warms the atmosphere intensely but briefly, absorbing sunlight and heating the air around it. Yet global climate assessments have long treated black carbon as a single, uniform compound. The new study argues that this simplification conceals a critical truth: black carbon exists as a continuum of particles with sharply different physical, chemical and optical properties, and those differences matter enormously for climate.

To resolve this heterogeneity, an international team led by researchers at the Southern University of Science and Technology in Shenzhen, together with colleagues at the Max Planck Institute for Chemistry and other institutions, assembled approximately 2,500 thermal–optical measurements of carbon fractions across a wide range of emission sources. Thermal–optical analysis separates elemental carbon into operationally defined fractions, allowing the researchers to distinguish between two fundamentally different subtypes of black carbon: char and soot. Char is the less light-absorbing, lower-maturity form, typically produced at lower combustion temperatures, while soot is the highly graphitized, strongly light-absorbing form generated by hot, efficient combustion processes.

Using this measurement database, the team constructed a global inventory that separately tracks char and soot emissions from 1750 to 2019, spanning the full arc of industrialization. The results reveal a striking transformation. In 1750, soot accounted for only about 16 percent of total anthropogenic black carbon emissions. By 2019, that fraction had climbed to roughly 35 percent. The driver of this shift is unmistakable: the relentless growth of fossil fuel consumption. Coal combustion, diesel engines and other high-temperature fossil fuel sources produce proportionally far more soot than the smoldering biomass burning and traditional domestic fires that dominated the pre-industrial world.

The optical consequences of this compositional change are profound. Soot absorbs light far more efficiently per unit mass than char, a property quantified by the mass absorption cross section. As the soot fraction of emissions has grown, the average mass absorption cross section of anthropogenic black carbon has risen accordingly, meaning that contemporary emissions trap more solar energy per kilogram than their historical counterparts. The researchers used radiative transfer calculations, including Lorenz–Mie theory and multiple-sphere T-matrix methods, to quantify how the changing char–soot mixture alters the absorption properties of the atmospheric black carbon burden, and coupled these estimates with simulations using the Community Earth System Model version 2.1.

Crucially, the inventory’s story is not built on models alone. The shift toward a higher soot fraction aligns with independent evidence preserved in sedimentary records. Lake sediments, such as the 150-year records from Lake Phayao in northern Thailand and from lakes on the Tibetan Plateau, archive the deposition of soot and char separately, and these archives corroborate the temporal trends reconstructed from the emission inventory. The agreement between bottom-up accounting and geological archives lends considerable weight to the central conclusion: the global warming potential of atmospheric black carbon, per unit of mass emitted, has increased since the mid-eighteenth century.

The implications ripple through climate science. Because climate models and assessments such as those of the Intergovernmental Panel on Climate Change generally treat black carbon as a single compound, they may misestimate its historical and present-day radiative forcing. If today’s black carbon is more absorbing than the black carbon of 1750, then the radiative forcing attributed to a given mass of emissions has grown over time, and the climate system’s response to soot reductions will differ from what single-compound models predict. Resolving black carbon subtypes in climate models, the authors argue, is essential for accurate estimation of black carbon’s contribution to observed warming and for projecting its future role.

The findings also carry a pointed message for mitigation policy. Not all black carbon sources are equal. Soot-rich sources, dominated by fossil fuel combustion in transport, industry and power generation, deliver far more warming per unit of emissions than char-dominated sources such as residential biomass burning. Prioritizing soot-rich sources for emission controls would therefore yield disproportionate climate benefits in the near term, a particularly attractive proposition given black carbon’s short atmospheric lifetime. Cutting soot-heavy emissions today produces almost immediate cooling, unlike reductions in long-lived greenhouse gases whose benefits unfold over decades to centuries.

The study also intersects with a broader and sometimes counterintuitive debate in aerosol–climate science. As societies clean up particulate pollution for public health reasons, some of the masking effect of reflective aerosols is lost, potentially accelerating near-term warming. Because black carbon is absorbing rather than reflecting, distinguishing its subtypes becomes even more important when weighing the climate consequences of air quality policies. A more absorbing black carbon burden means that the warming side of the aerosol ledger is heavier than previously assumed, sharpening the urgency of targeted controls on the most light-absorbing emissions.

By fusing thousands of laboratory and field measurements, a 270-year emission reconstruction, optical theory and Earth system modeling, the research delivers one of the most detailed portraits yet of how humanity has changed not just the quantity but the quality of the soot it injects into the atmosphere. The black carbon of the Anthropocene is darker, more absorbing and more warming than the black carbon of the age of sail and charcoal, and recognizing that evolution is a necessary step toward both sharper climate projections and smarter, faster-acting mitigation strategies.

Subject of Research: Historical increase in the global warming potential of atmospheric black carbon driven by a rising soot fraction from fossil fuel combustion since 1750

Article Title: Increase in global warming potential of atmospheric black carbon since 1750

Article References: Shen, H., Shen, H., Cheng, Y., Andreae, M. O., Jiao, X., Liang, X., Mai, Z., Guo, P., Zheng, L., Peng, X., Du, W., Wang, R., Zhang, G., Zhai, J., Zeng, Y., Sun, H. Z., Chen, Y., Wang, C., Huang, T., … Zhang, G. (2026). Increase in global warming potential of atmospheric black carbon since 1750. Nature Geoscience. https://doi.org/10.1038/s41561-026-02085-z

Image Credits: AI Generated

DOI: 10.1038/s41561-026-02085-z

Keywords: black carbon, soot, char, short-lived climate forcer, fossil fuel combustion, radiative forcing, mass absorption cross section, emission inventory, thermal-optical analysis, climate modeling, sediment records, Nature Geoscience