Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Reproducibility”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 757 records · Page 42

Searching for Phase Transitions in Artificial Kagome Spin Ice

Artificial Kagome spin ice is a metasurface consisting of small islands of polarizable magnets. These islands can be oriented by an external magnetic field applied in the same plane as the metasurface. By subsequently applying a relatively weaker field in the opposite direction of the initial field, a portion of the magnets in the lattice can be repolarized or “flipped” the other way. These flips cause instabilities in a previously uniformly aligned lattice, causing the system to interact with itself and induce changes in the overall magnetization of the metasurface. This Summer, my goal was to use this phenomenon to discern whether or not these systems could undergo a phase transition caused only by the initial change of the external magnetic field. Based on previous datasets that indicated that such a transition was observable, my task was to find and reproduce this potential phenomena. This process was conducted by repeating measurements on the system by using an external magnet to align the spin ice and then using Magnetic Force Microscopy (MFM) to probe the surface of the material to understand the changes in the magnetization over time. This report will discuss the characteristics of artificial Kagome spin ice, the methodology used to induce phase transitions, and the image processing methods used to understand the results. Although no reproducible evidence of a phase transition was found during the Summer, we were able to successfully reproduce short-term responses in the spin ice, as well as plot a hysteresis curve of the material.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

MLCommons Science Benchmarks

Benchmarks are a cornerstone of modern machine learning practice, providing standardized eval- uations that enable reproducibility, comparison, and scientific progress. Yet, as AI systems particularly deep learning models become increasingly dynamic, traditional static benchmarking approaches are losing their relevance. Models rapidly evolve in architecture, scale, and capability; datasets shift; and deployment contexts continuously change, creating a moving target for evaluation. Without adaptive benchmarking frame- works, both scientific assessment and real-world de- ployment risk becoming misaligned with actual system behavior. Drawing on our experience from MLCommons, educa- tional initiatives, and government programs such as the DOE s Million Parameter Consortium, we identify key barriers that hinder the broader adoption and utility of benchmarking in AI. These include substantial resource demands, limited access to specialized hardware, lack of expertise in benchmark design, and uncertainty among practitioners about how to relate benchmark results to their own application domains. Moreover, current benchmarks often emphasize peak performance on leadership-class hardware, offering limited guidance for more diverse, real-world deployment scenarios. We argue that benchmarking itself must become dy- namic in order to incorporate evolving models, updated data, and heterogeneous computational platforms while maintaining transparency, reproducibility, and inter- pretability. Democratizing this process requires not only technical innovation, but also systematic educational efforts spanning undergraduate to professional levels to develop sustained expertise in benchmark design and use. Finally, benchmarks should be framed and com- municated to support application-relevant comparisons, enabling both developers and users to make informed, context-sensitive decisions. Advancing dynamic and inclusive benchmarking practices will be essential to ensure that evaluation keeps pace with the evolving AI landscape and supports responsible, reproducible, and accessible AI deployment.

Hawks, Benjamin G. [Fermilab]↗

Development of SAM Code Capabilities for Safety Analysis of GCR Air-ingress Events

Air-ingress following a depressurized loss-of-forced-cooling (DLOFC) event is a challenging, multiphysics safety scenario for High-Temperature Gas-Cooled Reactors (HTGRs), involving coupled gas composition transport, buoyancy-driven flow redistribution, graphite oxidation, and structural heat-up. Despite its importance — air ingress is a key scenario identified in the PIRT process for the HTGRs — existing system-level safety codes have lacked the integrated capability to simulate the complete event sequence with high confidence. This report documents the development, validation, and demonstration of three new capabilities in the SAM code to address this gap: (1) a multi-component gas mixture flow model with binary diffusion to track the helium-air composition and its effect on system density and flow; (2) a 0-D graphite oxidation model based on the Roes correlation, including oxygen consumption and exothermic heat release; and (3) an isentropic critical flow model for accurate representation of primary system depressurization through a break. These capabilities are validated against two benchmark experiments. The NSTF heavy-gas ingress experiment validates the multi-component flow model: SAM correctly reproduces the rapid buoyancydriven flow stagnation and subsequent natural circulation recovery driven by composition-dependent density changes. The NACOK graphite oxidation experiment validates the oxidation model: SAM predicts a bottom-level graphite weight loss of 25%, in close agreement with the measured 24%, and reproduces the strong axial nonuniformity and block-geometry dependence of oxidation, at a level comparable to the SPECTRA and TINTE codes. The validated capabilities are then exercised together in an integrated, reactor-scale simulation of a DLOFC air-ingress transient in a simplified HTR-PM pebble-bed reactor. In a single calculation spanning approximately 8 days, SAM reproduces the complete accident sequence: rapid depressurization, densityand diffusion-driven air ingress over ˜15 hours, onset of buoyancy-driven natural circulation, exothermic graphite oxidation with a peak fuel temperature at ˜62 hours, and eventual passive cooldown. These results demonstrate that SAM now provides the nuclear community with a preliminarily validated, modern systemlevel tool for HTGR air-ingress safety analysis, filling a recognized capability gap. Future extensions to broaden species tracking, improve oxidation chemistry, and refine the reactor model are discussed.

Yang, Gang↗

Studies of Bystander Effects in 3-D Tissue Systems Using a Low-LET Microbeam

It is now accepted that biological effects may occur in cells that were not themselves traversed by ionizing radiation but are close to those that were. Little is known about the mechanism underlying such a bystander effect, although cell-to-cell communication is thought to be important. Previous work demonstrated a significant bystander effect for clonogenic survival and oncogenic transformation in C3H 10T(1/2) cells. Additional studies were undertaken to assess the importance of the degree of cell-to-cell contact at the time of irradiation on the magnitude of this bystander effect by varying the cell density. When 10% of cells were exposed to a range of 2-12 alpha particles, a significantly greater number of cells were inactivated when cells were irradiated at high density than at low density. In addition, the oncogenic transformation frequency was significantly higher in high-density cultures. These results suggest that when a cell is hit by radiation, the transmission of the bystander signal through cell-to-cell contact is an important mediator of the effect, implicating the involvement of intracellular communication through gap junctions. Additional studies to address the relationship between the bystander effect and the adaptive response were undertaken. A novel apparatus, where targeted and non-targeted cells were grown in close proximity, was used to investigate these. It was further examined whether a bystander effect or an adaptive response could be induced by a factor(s) present in the supernatants of cells exposed to a high or low dose of X-rays, respectively. When non-hit cells were co-cultured for 24 h with cells irradiated with 5 Gy alpha-particles, a significant increase in both cell killing and oncogenic transformation frequency was observed. If these cells were treated with 2 cGy X-rays 5 h before co-culture with irradiated cells, approximately 95% of the bystander effect was cancelled out. A 2.5-fold decrease in the oncogenic transformation frequency was also observed. When cells were cultured in medium donated from cells exposed to 5 Gy X-rays, a significant bystander effect was observed for clonogenic survival. When cells were cultured for 5 h with supernatant from donor cells exposed to 2 cGy and were then irradiated with 4 Gy X-rays, they failed to show an increase in survival compared with cells directly irradiated with 4 Gy. However, a twofold reduction in the oncogenic transformation frequency was seen. An adaptive dose of X-rays cancelled out the majority of the bystander effect produced by alpha-particles. For oncogenic transformation, but not cell survival, radioadaption can occur in unirradiated cells via a transmissible factor(s). A pilot study was undertaken to observe the bystander effect in a realistic multicellular three-dimensional morphology. We found bystander responses in a three-dimensional, normal human-tissue system. Endpoints were induction of micronucleated and apoptotic cells. A charged-particle microbeam was used, allowing irradiation of cells in defined locations in the tissue yet guaranteeing that no cells located more than a few micrometers away receive any radiation exposure. Unirradiated cells up to 1 mm distant from irradiated cells showed a significant enhancement in effect over background, with an average increase in effect of 1.7-fold for micronuclei and 2.8-fold for apoptosis. The surprisingly long range of bystander signals in human tissue suggests that bystander responses may be important in extrapolating radiation risk estimates from epidemiologically accessible doses down to very low doses where nonhit bystander cells will predominate. Finally, it would be of great benefit to develop a reproducible tissue system suitable for critical radiobiological assays. We have developed a reliable protocol to harvest cells from tissue samples and to investigate the damage induced on a single cell basis. In order to result in a valid tool for bystander experiments, the method focuses on processing and analyzing radiation damage in individual cells as a function of their relative position in the tissue. We have investigated the micronucleus formation following partial irradiation with 3.5 MeV protons in an artificial human skin construct. Following the optimization of the Cytochalasin-B concentration and incubation time necessary to obtain a reproducible and suitable number of binucleated cells, the induction of micronuclei across the samples is assessed for 3 dose. The reproducible and low background frequency of micronuclei measured in this system allowed us to detect small increases following the irradiation exposure. The effect is statistically significant at doses as low as 0.1 Gy and it shows evidence of a spatial dependency as it decreases in the cells further away from the directly exposed area. This experimental protocol represents the initial steps in the development of an in vivo-like assay for complex radiation damage in human tissues.

59 BASIC BIOLOGICAL SCIENCES↗

Development of Coarse-Grained Models for Poly(4-vinylphenol) and Poly(2-vinylpyridine): Polymer Chemistries with Hydrogen Bonding

In this paper, we identify the modifications needed in a recently developed generic coarse-grained (CG) model that captured directional interactions in polymers to specifically represent two exemplary hydrogen bonding polymer chemistries—poly(4-vinylphenol) and poly(2-vinylpyridine). We use atomistically observed monomer-level structures (e.g., bond, angle and torsion distribution) and chain structures (e.g., end-to-end distance distribution and persistence length) of poly(4-vinylphenol) and poly(2-vinylpyridine) in an explicitly represented good solvent (tetrahydrofuran) to identify the appropriate modifications in the generic CG model in implicit solvent. For both chemistries, the modified CG model is developed based on atomistic simulations of a single 24-mer chain. This modified CG model is then used to simulate longer (36-mer) and shorter (18-mer and 12-mer) chain lengths and compared against the corresponding atomistic simulation results. We find that with one to two simple modifications (e.g., incorporating intra-chain attraction, torsional constraint) to the generic CG model, we are able to reproduce atomistically observed bond, angle and torsion distributions, persistence length, and end-to-end distance distribution for chain lengths ranging from 12 to 36 monomers. We also show that this modified CG model, meant to reproduce atomistic structure, does not reproduce atomistically observed chain relaxation and hydrogen bond dynamics, as expected. Simulations with the modified CG model have significantly faster chain relaxation than atomistic simulations and slower decorrelation of formed hydrogen bonds than in atomistic simulations, with no apparent dependence on chain length.

36 MATERIALS SCIENCE↗

INVESTIGATING THE CONNECTION BETWEEN QUASI-PERIODIC OSCILLATIONS AND SPECTRAL COMPONENTS WITH NuSTAR DATA OF GRS 1915+105

Low-frequency quasi-periodic oscillations (QPOs) are commonly observed during the hard states of black hole binaries. Several studies have established various observational/empirical correlations between spectral parameters and QPO properties, indicating a close link between the two. However, the exact mechanism of generation of QPOs is not yet well understood. In this paper, we present our attempts to comprehend the connection between the spectral components and the low-frequency QPO (LFQPO) observed in GRS 1915+105 using the data from NuSTAR. Detailed spectral modeling as well as the presence of the LFQPO and its energy dependence during this observation have been reported by Miller et al. and Zhang et al., respectively. We investigate the compatibility of the spectral model and the energy dependence of the QPO by simulating light curves in various energy bands for small variation of the spectral parameters. The basic concept here is to establish the connection, if any, between the QPO and the variation of either a spectral component or a specific parameter, which in turn can shed some light on the origin of the QPO. We begin with the best-fit spectral model of Miller et al. and simulate the light curve by varying the spectral parameters at frequencies close to the observed QPO frequency in order to generate the simulated QPO. Furthermore we simulate similar light curves in various energy bands in order to reproduce the observed energy dependence of the rms amplitude of the QPO. We find that the observed trend of increasing rms amplitude with energy can be reproduced qualitatively if the spectral index is assumed to be varying with the phases of the QPO. Variation of any other spectral parameter does not reproduce the observed energy dependence.

79 ASTRONOMY AND ASTROPHYSICS↗

A MODEL FOR THE DUST ENVELOPE OF THE SILICATE CARBON STAR IRAS 09425-6040

IRAS 09425-6040 (I09425) is a silicate carbon star with conspicuous crystalline silicate and water-ice features and emission excesses in the far-infrared and millimeter (mm) wavelength ranges. To understand properties of the dust envelope of I09425, we propose a physical model based on the observations and known properties of asymptotic giant branch stars and dust. We perform radiative transfer model calculations using multiple dust shells and disks with various dust species. We compare the model results with the observed spectral energy distribution (SED) acquired with different telescopes. We find that the physical model for I09425 using multiple shells of carbon and silicate dust and multiple disks of amorphous and crystalline silicates reproduces the observed SED fairly well. This object looks to have detached cold O-rich (silicate and water-ice) dust shells, which could be remnants of the recent chemical transition from O to C and an inner C-rich dust shell. A long-lived thin disk of very large silicate grains can reproduce the emission excess in the mm wavelength band and a recently formed thick disk of crystalline silicates can reproduce the prominent emission features in the spectral range 8–45 μm. The highly crystallized silicates could be recently formed by high temperature annealing due to the last O-rich superwind just before the chemical transition of the central star. I09425 could be a rare object that has the remnants of past O-rich stellar winds in the outer shells as well as in the circumbinary disks.

79 ASTRONOMY AND ASTROPHYSICS↗

SOLAR MODELS WITH NEW LOW METAL ABUNDANCES

In the past decade, the photospheric abundances of the Sun had been revised several times by many observers. The standard solar models constructed with the new low-metal abundances disagree with helioseismic results and detected neutrino fluxes. The solar model problem has puzzled some stellar physicists for more than 10 years. Rotation, enhanced diffusion, convection overshoot, and magnetic fields are used to reconcile the new abundances with helioseismology. The too low helium subsurface abundance in enhanced diffusion models can be improved by the mixing caused by rotation and magnetic fields. The problem of the depth of the convective zone in rotating models can be resolved by convection overshoot. Consequently, the Asplund–Grevesse–Sauval rotation model including overshooting (AGSR) reproduces the seismically inferred sound-speed and density profiles and the convection zone depth as well as the Grevesse and Sauval model computed before. But this model fails to reproduce the surface helium abundance, which is 0.2393 (2.6 σ away from the seismic value), and neutrino fluxes. The magnetic model called AGSM keeps the agreement of the AGSR and improves the prediction of the surface helium abundance. The observed separation ratios r {sub 02} and r {sub 13} are reasonably reproduced by AGSM. Moreover, neutrino fluxes calculated by this model are not far from the detected neutrino fluxes and the predictions of previous works.

79 ASTRONOMY AND ASTROPHYSICS↗

A Parameter Study for Modeling Mg ii h and k Emission during Solar Flares

Solar flares show highly unusual spectra in which the thermodynamic conditions of the solar atmosphere are encoded. Current models are unable to fully reproduce the spectroscopic flare observations, especially the single-peaked spectral profiles of the Mg ii h and k lines. We aim to understand the formation of the chromospheric and optically thick Mg ii h and k lines in flares through radiative transfer calculations. We take a flare atmosphere obtained from a simulation with the radiative hydrodynamic code RADYN as input for a radiative transfer modeling with the RH code. By iteratively changing this model atmosphere and varying thermodynamic parameters such as temperature, electron density, and velocity, we study their effects on the emergent intensity spectra. We reproduce the typical single-peaked Mg ii h and k flare spectral shape and approximate the intensity ratios to the subordinate Mg ii lines by increasing either densities, temperatures, or velocities at the line core formation height range. Additionally, by combining unresolved upflows and downflows up to ∼250 km s{sup −1} within one resolution element, we reproduce the widely broadened line wings. While we cannot unambiguously determine which mechanism dominates in flares, future modeling efforts should investigate unresolved components, additional heat dissipation, larger velocities, and higher densities and combine the analysis of multiple spectral lines.

79 ASTRONOMY AND ASTROPHYSICS↗

Constraints on the IBEX Ribbon’s Origin from Its Evolution over a Solar Cycle

Abstract In 2009, the Interstellar Boundary Explorer (IBEX) discovered a narrow “ribbon” of energetic neutral atom emissions across the sky with properties correlated with the solar wind latitudinal structure and the interstellar magnetic field draped around the heliosphere. It is widely believed that the ribbon is formed from the escape of heliospheric ENAs into the local interstellar medium and their eventual return as secondary ENAs. However, there is no consensus on the rate of pitch angle scattering of these PUIs before they become secondary ENAs. We test two opposing limits of scattering rates (“weak” versus “strong”) by solving a time-dependent model of the ribbon that evolves with the solar cycle, and we compare them to IBEX observations over 2009–2019. First, we find that both models qualitatively reproduce the evolution of IBEX fluxes for most of the data set, with a few exceptions, although the strong (or “spatial retention”) scattering model greatly underestimates the observed fluxes. Regardless, time dependence of fluxes cannot distinguish these models. Second, the ribbon’s geometric properties, i.e., its center and radius, are significantly different between the models. The spatial retention model reproduces the observed ribbon centers as a function of energy and time slightly better than the weak scattering model, and the spatial retention model reproduces the observed ribbon radius over energy and time almost perfectly, whereas the weak scattering model compares poorly. Our analysis favors the spatial retention mechanism as the source of the IBEX ribbon, but it requires modification to increase the flux of ENAs observed at 1 au.

Zirnstein, E. J. (ORCID:0000000172400618)↗

The DESI One-Percent Survey: Constructing Galaxy–Halo Connections for ELGs and LRGs Using Auto and Cross Correlations

In the current Dark Energy Spectroscopic Instrument (DESI) survey, emission line galaxies (ELGs) and luminous red galaxies (LRGs) are essential for mapping the dark matter distribution at z ~ 1. We measure the auto and cross correlation functions of ELGs and LRGs at 0.8 < z ≤ 1.0 from the DESI One-Percent survey. Following Gao et al., we construct the galaxy–halo connections for ELGs and LRGs simultaneously. With the stellar–halo mass relation for the whole galaxy population (i.e., normal galaxies), LRGs can be selected directly by stellar mass, while ELGs can also be selected randomly based on the observed number density of each stellar mass, once the probability P sat of a satellite galaxy becoming an ELG is determined. We demonstrate that the observed small scale clustering prefers a halo mass-dependent P sat model rather than a constant. With this model, we can well reproduce the auto correlations of LRGs and the cross correlations between LRGs and ELGs at r p > 0.1 Mpc h –1 . We can also reproduce the auto correlations of ELGs at r p > 0.3 Mpc h –1 (s > 1 Mpc h –1 ) in real (redshift) space. Although our model has only seven parameters, we show that it can be extended to higher redshifts and reproduces the observed auto correlations of ELGs in the whole range of 0.8 < z ≤ 1.6, which enables us to generate a lightcone ELG mock for DESI. With the above model, we further derive halo occupation distributions for ELGs, which can be used to produce ELG mocks in coarse simulations without resolving subhalos.

79 ASTRONOMY AND ASTROPHYSICS↗

Was There a 3.5 keV Line?

Abstract The 3.5 keV line is a purported emission line observed in galaxies, galaxy clusters, and the Milky Way whose origin is inconsistent with known atomic transitions and has previously been suggested to arise from dark matter decay. We systematically reexamine the bulk of the evidence for the 3.5 keV line, attempting to reproduce six previous analyses that found evidence for the line. We only reproduce one of the analyses; in the other five, we find no significant evidence for a 3.5 keV line when following the described analysis procedures on the original data sets. For example, previous results claimed 4 σ evidence for a 3.5 keV line from the Perseus cluster; we dispute this claim, finding no evidence for a 3.5 keV line. We find evidence for background mismodeling in multiple analyses. We show that analyzing these data in narrower energy windows diminishes the effects of mismodeling but returns no evidence for a 3.5 keV line. We conclude that there is little robust evidence for the existence of the 3.5 keV line. Some of the discrepancy of our results from those of the original works may be due to the earlier reliance on local optimizers, which we demonstrate can lead to incorrect results. For ease of reproducibility, all code and data are publicly available.

79 ASTRONOMY AND ASTROPHYSICS↗

Galaxy Size and Rotation Curve Diversity in ΛCDM with Baryons

The observed rotation curves of dwarf galaxies exhibit significant diversity at fixed halo mass, challenging galaxy formation within the cold dark matter (CDM) model. Previous cosmological galaxy formation simulations with baryonic physics fail to reproduce the full diversity of rotation curves, suggesting that there is a flaw in baryonic feedback models, observational bias, or that an alternative to CDM must be invoked. In this work, we use the Marvelous Massive Dwarf zoom-in simulations, a suite of high-resolution dwarf simulations with M 200 ∼ 10 10 –10 11 M ⊙ and M * ∼ 10 7 –10 9 M ⊙ , designed to target the mass range where the galaxy rotation curve diversity is maximized, i.e., between and 100 km s −1 . We add to this a set of low-mass galaxies from the Marvel Dwarf Zoom Volumes to extend the galaxy mass range to lower values. Our fiducial star formation and feedback models produce simulated dwarfs with a broader range of rotation curve shapes, similar to observations. These simulations both create dark matter cores via baryonic feedback, reproducing the slower-rising rotation curves, while also allowing for compact galaxies and steeply rising rotation curves. Our simulated dwarfs also reproduce the observed size–M * relation, including scatter, producing both extended and compact dwarfs for the first time in simulated field dwarfs. However, the slowly rising and high baryon mass fraction, as well as the steeply rising and low baryon mass fraction, remain missing. We explore star formation and feedback models and conclude that previous simulations may have had feedback that was too strong to produce compact dwarfs.

Cruz, Akaxia [Flatiron Institute, New York, NY (Un↗

A spectroscopically confirmed double source plane lens system in the Hyper Suprime-Cam Subaru Strategic Program

We report the serendipitous discovery of HSC J142449−005322, a double source plane lens system in the Hyper Suprime-Cam Subaru Strategic Program. We dub the system Eye of Horus. The lens galaxy is a very massive early-type galaxy with stellar mass of ∼7×10{sup 11} M{sub ⊙} located at z{sub L}=0.795. The system exhibits two arcs/rings with clearly different colors, including several knots. We have performed spectroscopic follow-up observations of the system with FIRE on Magellan. The outer ring is confirmed at z{sub S2}=1.988 with multiple emission lines, while the inner arc and counterimage is confirmed at z{sub S1}=1.302. This makes it the first double source plane system with spectroscopic redshifts of both sources. Interestingly, redshifts of two of the knots embedded in the outer ring are found to be offset by Δz=0.002 from the other knots, suggesting that the outer ring consists of at least two distinct components in the source plane. We perform lens modeling with two independent codes and successfully reproduce the main features of the system. However, two of the lensed sources separated by ∼0.7 arcsec cannot be reproduced by a smooth potential, and the addition of substructure to the lens potential is required to reproduce them. Higher-resolution imaging of the system will help decipher the origin of this lensing feature and potentially detect the substructure.

79 ASTRONOMY AND ASTROPHYSICS↗

Global dust model intercomparison in AeroCom phase I

This study presents the results of a broad intercomparison of a total of 15 global aerosol models within the AeroCom project. Each model is compared to observations related to desert dust aerosols, their direct radiative effect, and their impact on the biogeochemical cycle, i.e., aerosol optical depth (AOD) and dust deposition. Additional comparisons to Angström exponent (AE), coarse mode AOD and dust surface concentrations are included to extend the assessment of model performance and to identify common biases present in models. These data comprise a benchmark dataset that is proposed for model inspection and future dust model development. There are large differences among the global models that simulate the dust cycle and its impact on climate. In general, models simulate the climatology of vertically integrated parameters (AOD and AE) within a factor of two whereas the total deposition and surface concentration are reproduced within a factor of 10. In addition, smaller mean normalized bias and root mean square errors are obtained for the climatology of AOD and AE than for total deposition and surface concentration. Characteristics of the datasets used and their uncertainties may influence these differences. Large uncertainties still exist with respect to the deposition fluxes in the southern oceans. Further measurements and model studies are necessary to assess the general model performance to reproduce dust deposition in ocean regions sensible to iron contributions. Models overestimate the wet deposition in regions dominated by dry deposition. They generally simulate more realistic surface concentration at stations downwind of the main sources than at remote ones. Most models simulate the gradient in AOD and AE between the different dusty regions. However the seasonality and magnitude of both variables is better simulated at African stations than Middle East ones. The models simulate the offshore transport of West Africa throughout the year but they overestimate the AOD and they transport too fine particles. The models also reproduce the dust transport across the Atlantic in the summer in terms of both AOD and AE but not so well in winter-spring nor the southward displacement of the dust cloud that is responsible of the dust transport into South America. Based on the dependency of AOD on aerosol burden and size distribution we use model bias with respect to AOD and AE to infer the bias of the dust emissions in Africa and the Middle East. Finally, according to this analysis we suggest that a range of possible emissions for North Africa is 400 to 2200 Tg yr -1 and in the Middle East 26 to 526 Tg yr -1 .

54 ENVIRONMENTAL SCIENCES↗

Immersion freezing of water and aqueous ammonium sulfate droplets initiated by humic-like substances as a function of water activity

Immersion freezing of water and aqueous (NH 4 ) 2 SO 4 droplets containing leonardite (LEO) and Pahokee peat (PP) serving as surrogates for humic-like substances (HULIS) has been investigated. Organic aerosol containing HULIS are ubiquitous in the atmosphere; however, their potential for ice cloud formation is uncertain. Immersion freezing has been studied for temperatures as low as 215K and solution water activity, a w , from 0.85 to 1.0. The freezing temperatures of water and aqueous solution droplets containing LEO and PP are 5–15 K warmer than homogeneous ice nucleation temperatures. Heterogeneous freezing temperatures can be represented by a horizontal shift of the ice melting curve as a function of solution aw by Δ a w = 0.2703 and 0.2466, respectively. Corresponding heterogeneous ice nucleation rate coefficients, Jhet, are (9.6 ± 2.5) × 10 4 and (5.4 ± 1.4) × 10 4 cm -2 s -1 for LEO and PP containing droplets, respectively, and remain constant along freezing curves characterized by Δ a w . Consequently predictions of freezing temperatures and kinetics can be made without knowledge of the solute type when relative humidity and ice nuclei (IN) surface areas are known. The acquired ice nucleation data are applied to evaluate different approaches to fit and reproduce experimentally derived frozen fractions. In addition, we apply a basic formulation of classical nucleation theory (α( T )-model) to calculate contact angles and frozen fractions. Contact angles calculated for each ice nucleus as a function of temperature, α( T )-model, reproduce exactly experimentally derived frozen fractions without involving free-fit parameters. However, assigning the IN a single contact angle for the entire population (single-α model) is not suited to represent the frozen fractions. Application of α-PDF, active sites, and deterministic model approaches to measured frozen fractions yield similar good representations. Furthermore, when using a single parameterization of α-PDF or active sites distribution to fit all individual a w immersion freezing data simultaneously, frozen fraction curves are not reproduced. This implies that these fitting formulations cannot be applied to immersion freezing of aqueous solutions, and suggests that derived fit parameters do not represent independent particle properties. Thus, from fitting frozen fractions only, the underlying ice nucleation mechanism and nature of the ice nucleating sites cannot be inferred. In contrast to using fitted functions obtained to represent experimental conditions only, we suggest to use experimentally derived J het as a function of temperature and a w that can be applied to conditions outside of those probed in laboratory. Finally, this is because J het (T) is independent of time and IN surface areas in contrast to the fit parameters obtained by representation of experimentally derived frozen fractions.

54 ENVIRONMENTAL SCIENCES↗

Elemental ratio measurements of organic compounds using aerosol mass spectrometry: characterization, improved calibration, and implications

Elemental compositions of organic aerosol (OA) particles provide useful constraints on OA sources, chemical evolution, and effects. The Aerodyne high-resolution time-of-flight aerosol mass spectrometer (HR-ToF-AMS) is widely used to measure OA elemental composition. This study evaluates AMS measurements of atomic oxygen-to-carbon (O : C), hydrogen-to-carbon (H : C), and organic mass-to-organic carbon (OM : OC) ratios, and of carbon oxidation state ( OS C ) for a vastly expanded laboratory data set of multifunctional oxidized OA standards. For the expanded standard data set, the method introduced by Aiken et al. (2008), which uses experimentally measured ion intensities at all ions to determine elemental ratios (referred to here as "Aiken-Explicit"), reproduces known O : C and H : C ratio values within 20% (average absolute value of relative errors) and 12%, respectively. The more commonly used method, which uses empirically estimated H 2 O + and CO + ion intensities to avoid gas phase air interferences at these ions (referred to here as "Aiken-Ambient"), reproduces O : C and H : C of multifunctional oxidized species within 28 and 14% of known values. The values from the latter method are systematically biased low, however, with larger biases observed for alcohols and simple diacids. A detailed examination of the H 2 O + , CO + , and CO 2 + fragments in the high-resolution mass spectra of the standard compounds indicates that the Aiken-Ambient method underestimates the CO + and especially H 2 O + produced from many oxidized species. Combined AMS–vacuum ultraviolet (VUV) ionization measurements indicate that these ions are produced by dehydration and decarboxylation on the AMS vaporizer (usually operated at 600 °C). Thermal decomposition is observed to be efficient at vaporizer temperatures down to 200 °C. These results are used together to develop an "Improved-Ambient" elemental analysis method for AMS spectra measured in air. The Improved-Ambient method uses specific ion fragments as markers to correct for molecular functionality-dependent systematic biases and reproduces known O : C (H : C) ratios of individual oxidized standards within 28% (13%) of the known molecular values. The error in Improved-Ambient O : C (H : C) values is smaller for theoretical standard mixtures of the oxidized organic standards, which are more representative of the complex mix of species present in ambient OA. For ambient OA, the Improved-Ambient method produces O : C (H : C) values that are 27% (11%) larger than previously published Aiken-Ambient values; a corresponding increase of 9% is observed for OM : OC values. These results imply that ambient OA has a higher relative oxygen content than previously estimated. The OS C values calculated for ambient OA by the two methods agree well, however (average relative difference of 0.06 OS C units). This indicates that OS C is a more robust metric of oxidation than O : C, likely since OS C is not affected by hydration or dehydration, either in the atmosphere or during analysis.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Interpreting machine learning prediction of fire emissions and comparison with FireMIP process-based models

Annual burned areas in the United States have increased 2-fold during the past decades. With more large fires resulting in more emissions of fine particulate matter, an accurate prediction of fire emissions is critical for quantifying the impacts of fires on air quality, human health, and climate. This study aims to construct a machine learning (ML) model with game-theory interpretation to predict monthly fire emissions over the contiguous US (CONUS) and to understand the controlling factors of fire emissions. The optimized ML model is used to diagnose the process-based models in the Fire Modeling Intercomparison Project (FireMIP) to inform future development. Results show promising performance for the ML model, Community Land Model (CLM), and Joint UK Land Environment Simulator-Interactive Fire And Emission Algorithm For Natural Environments (JULES-INFERNO) in reproducing the spatial distributions, seasonality, and interannual variability of fire emissions over the CONUS. Regional analysis shows that only the ML model and CLM simulate the realistic interannual variability of fire emissions for most of the subregions (r >0.95 for ML and r =0.14~0.70 for CLM), except for Mediterranean California, where all the models perform poorly (r =0.74 for ML and r <0.30 for the FireMIP models). Regarding seasonality, most models capture the peak emission in July over the western US. However, all models except for the ML model fail to reproduce the bimodal peaks in July and October over Mediterranean California, which may be explained by the smaller wind speeds of the atmospheric forcing data during Santa Ana wind events and limitations in model parameterizations for capturing the effects of Santa Ana winds on fire activity. Furthermore, most models struggle to capture the spring peak in emissions in the southeastern US, probably due to underrepresentation of human effects and the influences of winter dryness on fires in the models. As for extreme events, both the ML model and CLM successfully reproduce the frequency map of extreme emission occurrence but overestimate the number of months with extremely large fire emissions. Comparing the fire PM 2.5 emissions from the ML model with process-based fire models highlights their strengths and uncertainties for regional analysis and prediction and provides useful insights into future directions for model improvements.

54 ENVIRONMENTAL SCIENCES↗