Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “sparse factorization”

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 361 records · Page 20
All sourcesNASA NTRSDOE OSTI
NASA NTRS2011

Sparse Solutions for Single Class SVMs: A Bi-Criterion Approach

In this paper we propose an innovative learning algorithm - a variation of One-class nu Support Vector Machines (SVMs) learning algorithm to produce sparser solutions with much reduced computational complexities. The proposed technique returns an approximate solution, nearly as good as the solution set obtained by the classical approach, by minimizing the original risk function along with a regularization term. We introduce a bi-criterion optimization that helps guide the search towards the optimal set in much reduced time. The outcome of the proposed learning technique was compared with the benchmark one-class Support Vector machines algorithm which more often leads to solutions with redundant support vectors. Through out the analysis, the problem size for both optimization routines was kept consistent. We have tested the proposed algorithm on a variety of data sources under different conditions to demonstrate the effectiveness. In all cases the proposed algorithm closely preserves the accuracy of standard one-class nu SVMs while reducing both training time and test time by several factors.

Das, Santanu↗
NASA NTRS1996

Immunohistochemical evidence of rapid extracellular matrix remodeling after iron-particle irradiation of mouse mammary gland

High-LET radiation has unique physical and biological properties compared to sparsely ionizing radiation. Recent studies demonstrate that sparsely ionizing radiation rapidly alters the pattern of extracellular matrix expression in several tissues, but little is known about the effect of heavy-ion radiation. This study investigates densely ionizing radiation-induced changes in extracellular matrix localization in the mammary glands of adult female BALB/c mice after whole-body irradiation with 0.8 Gy 600 MeV iron particles. The basement membrane and interstitial extracellular matrix proteins of the mammary gland stroma were mapped with respect to time postirradiation using immunofluorescence. Collagen III was induced in the adipose stroma within 1 day, continued to increase through day 9 and was resolved by day 14. Immunoreactive tenascin was induced in the epithelium by day 1, was evident at the epithelial-stromal interface by day 5-9 and persisted as a condensed layer beneath the basement membrane through day 14. These findings parallel similar changes induced by gamma irradiation but demonstrate different onset and chronicity. In contrast, the integrity of epithelial basement membrane, which was unaffected by sparsely ionizing radiation, was disrupted by iron-particle irradiation. Laminin immunoreactivity was mildly irregular at 1 h postirradiation and showed discontinuities and thickening from days 1 to 9. Continuity was restored by day 14. Thus high-LET radiation, like sparsely ionizing radiation, induces rapid-remodeling of the stromal extracellular matrix but also appears to alter the integrity of the epithelial basement membrane, which is an important regulator of epithelial cell proliferation and differentiation.

NASA Discipline Radiation Health↗
NASA NTRS1993

A Spectral Algorithm for Envelope Reduction of Sparse Matrices

The problem of reordering a sparse symmetric matrix to reduce its envelope size is considered. A new spectral algorithm for computing an envelope-reducing reordering is obtained by associating a Laplacian matrix with the given matrix and then sorting the components of a specified eigenvector of the Laplacian. This Laplacian eigenvector solves a continuous relaxation of a discrete problem related to envelope minimization called the minimum 2-sum problem. The permutation vector computed by the spectral algorithm is a closest permutation vector to the specified Laplacian eigenvector. Numerical results show that the new reordering algorithm usually computes smaller envelope sizes than those obtained from the current standard algorithms such as Gibbs-Poole-Stockmeyer (GPS) or SPARSPAK reverse Cuthill-McKee (RCM), in some cases reducing the envelope by more than a factor of two.

Barnard, Stephen T.↗
DOE OSTI2013

Characterizing Walk Trips in communities by Using Data from 2009 National Household Travel Survey, American Community Survey, and Other Sources

Non-motorized travel (i.e. walking and bicycling) are of increasing interest to the transportation profession, especially in context with energy consumption, reducing vehicular congestion, urban development patterns, and promotion of healthier life styles. This research project aimed to identify factors impacting the amount of travel for both walk and bike trips at the Census block group or tract level, using several public and private data sources. The key survey of travel behavior is the 2009 National Household Travel Survey (NHTS) which had over 87,000 walk trips for persons 16 and over, and over 6000 bike trips for persons 16 and over. The NHTS, in conjunction with the Census Bureau s American Community Survey, street density measures using Census Bureau TIGER, WalkScore , Nielsen Claritas employment estimates, and several other sources were used for this study. Stepwise Logistic Regression modeling techniques as well as Discriminant Analysis were applied using the integrated data set. While the models performed reasonably well for walk trips, travel by bike was abandoned due to sparseness of data. This paper discusses data sources utilized and modeling processes conducted under this study. It also presents a summary of findings and addresses data challenges and lesson-learned from this research effort.

NHTS↗
DOE OSTI2023

Predicting Seafloor Seismic Noise from Sea Surface Wave Heights and Vice Versa

Seismic hazard monitoring and global seismic tomography efforts are improved by recording seismic signals close to sources and at a variety of azimuths surrounding the source. While seismic networks have been deployed at many locations on land, seafloor stations are still sparse. Inclusion of ocean-based data would greatly improve the global coverage of seismic networks. However, the use of seafloor seismic data to complement land-based detection and characterization of events is complicated by the fact that ambient noise on the seafloor is in general much higher than noise levels on land. This noise is driven primarily by sea surface waves and tides. In this project, I define an empirical relationship between wave height and seafloor seismic displacement, and tie this relationship to the microseism noise theory of Longuet-Higgins and Ursell (1948). I find that while there appears to be a linear relationship between wave height and seismic displacement, wave height values can only be used with low confidence to predict seismic displacement and vice versa, and furthermore, factors other than wave height are significant in determining the level of seismic noise at a particular site.

58 GEOSCIENCES↗
DOE OSTI2015

Russian anthropogenic black carbon: Emission reconstruction and Arctic black carbon simulation

Development of reliable source emission inventories is needed to advance the understanding of the origin of Arctic haze using chemical transport modeling. This paper develops a regional anthropogenic black carbon (BC) emission inventory for the Russian Federation, the largest country by land area in the Arctic Council. Activity data from combination of local Russia information and international resources, emission factors based on either Russian documents or adjusted values for local conditions, and other emission source data are used to approximate the BC emissions. Emissions are gridded at a resolution of 0.1° × 0.1° and developed into a monthly temporal profile. Total anthropogenic BC emission of Russia in 2010 is estimated to be around 224 Gg. Gas flaring, a commonly ignored black carbon source, contributes a significant fraction of 36.2% to Russia's total anthropogenic BC emissions. Other sectors, i.e., residential, transportation, industry, and power plants, contribute 25.0%, 20.3%, 13.1%, and 5.4%, respectively. Three major BC hot spot regions are identified: the European part of Russia, the southern central part of Russia where human population densities are relatively high, and the Urals Federal District where Russia's major oil and gas fields are located but with sparse human population. BC simulations are conducted using the hemispheric version of Community Multi-scale Air Quality Model with emission inputs from a global emission database EDGAR (Emissions Database for Global Atmospheric Research)-HTAPv2 (Hemispheric Transport of Air Pollution) and EDGAR-HTAPv2 with its Russian part replaced by the newly developed Russian BC emissions, respectively. The simulation using the new Russian BC emission inventory could improve 30–65% of absorption aerosol optical depth measured at the AERONET sites in Russia throughout the whole year as compared to that using the default HTAPv2 emissions. At the four ground monitoring sites (Zeppelin, Barrow, Alert, and Tiksi) in the Arctic Circle, surface BC simulations are improved the most during the Arctic haze periods (October–March). The poor performance of Arctic BC simulations in previous studies may be partly ascribed to the Russian BC emissions built on out-of-date and/or missing information, which could result in biases to both emission rates and the spatial distribution of emissions. Finally, this study highlights that the impact of Russian emissions on the Arctic haze has likely been underestimated, and its role in the Arctic climate system needs to be reassessed. The Russian black carbon emission source data generated in this study can be obtained via http://abci.ornl.gov/download.shtml or http://acs.engr.utk.edu/Data.php.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1997

Two Improved Algorithms for Envelope and Wavefront Reduction

Two algorithms for reordering sparse, symmetric matrices or undirected graphs to reduce envelope and wavefront are considered. The first is a combinatorial algorithm introduced by Sloan and further developed by Duff, Reid, and Scott; we describe enhancements to the Sloan algorithm that improve its quality and reduce its run time. Our test problems fall into two classes with differing asymptotic behavior of their envelope parameters as a function of the weights in the Sloan algorithm. We describe an efficient 0(nlogn + m) time implementation of the Sloan algorithm, where n is the number of rows (vertices), and m is the number of nonzeros (edges). On a collection of test problems, the improved Sloan algorithm required, on the average, only twice the time required by the simpler Reverse Cuthill-Mckee algorithm while improving the mean square wavefront by a factor of three. The second algorithm is a hybrid that combines a spectral algorithm for envelope and wavefront reduction with a refinement step that uses a modified Sloan algorithm. The hybrid algorithm reduces the envelope size and mean square wavefront obtained from the Sloan algorithm at the cost of greater running times. We illustrate how these reductions translate into tangible benefits for frontal Cholesky factorization and incomplete factorization preconditioning.

Kumfert, Gary↗
DOE OSTI2020

Conservative Estimation of Tail Probabilities from Limited Sample Data

Several sparse-sample uncertainty quantification (UQ) methods are compared for conservative but not overly conservative estimation of small tail probabilities involving responses that lay beyond specified thresholds in the tails of probability distributions. Sixteen very differently shaped distributions (or probability density functions, PDFs) and tail probability magnitudes ranging from 10 -5 to 10 -1 are considered in order for the study to be relevant to a wide range of risk analysis and quantification of margins and uncertainty (QMU) problems. The emphasis of the study is on limited data regimes ranging from N = 2 to 20 samples, reflective of most experimental and some expensive computational situations. Relatively simple sparse-sample UQ methods tested for this regime involve statistical tolerance interval "Equivalent Normal and related "Ensemble of Normals" and "Superdistribution (SD) approaches. (The independently derived SD is effectively equivalent to the Bayesian posterior predictive distribution given the assumptions of the derivation.) The performance of the methods was generally improved for N ≥ 5 samples with a generalized Jackknife resampling technique, which determines a tail probability estimate by averaging estimates from smaller sub-samples. Several quantitative metrics for method conservatism and accuracy of tail probability estimation are used to assess and rank the methods' performance over many random trials for each test PDF and probability magnitude. A variant of Bootstrap resampling was also tried, but did not significantly improve tail probability estimates in most cases. Detailed results are presented from over 100-million tests over the above factors that provide useful granular information on which methods or combination of methods perform best in various areas of the factor space.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Structural Heterogeneity and Hydrodynamics of an Intrinsically Disordered Protein Condensate

Biology demonstrates precise control over the free-energy landscape through the selective partitioning of biomacromolecules into membraneless organelles, enabling essential functions such as biochemical transformations, signaling cascades, and mechanical reinforcement. Although the function of these condensates depends on their underlying structure and hydrodynamics, molecular-scale information on these systems remains sparse. Here, in this study, neutron scattering is used to probe the organization and dynamics of the intrinsically disordered N-terminal domain of Galectin-3, an extracellular lectin responsible for facilitating liquid–liquid phase separation on the cellular surface, in both dilute and condensed phases. Dilute solutions contain isolated protein chains in equilibrium with mesoscopic clusters, whereas the condensed phase adopts a bicontinuous, microemulsion-like morphology. The dilute phase behavior is quantitatively described by coarse-grained polymer models from soft-matter physics, demonstrating their predictive power for complex biological proteins. At elevated concentrations, the proteins self-assemble akin to block copolymers, microphase separating through the aggregation of hydrophobic domains along the protein contour. The resulting condensate remains fluid-like despite a 25-fold increase in concentration; its internal hydrodynamics slow by only a factor of 3 relative to dilute protein chains. These results provide a molecular-level framework for how disordered proteins achieve both the structural complexity and dynamic fluidity of biomolecular condensates.

Carrick, Brian R. [Massachusetts Inst. of Technolo↗
NASA NTRS2022

Drill-Site Selection for Cosmogenic-Nuclide Exposure Dating of the Bed of the Greenland Ice Sheet

Direct observations of the size of the Greenland Ice Sheet during Quaternary interglaciations are sparse yet valuable for testing numerical models of ice-sheet history and sea level contribution. Recent measurements of cosmogenic nuclides in bedrock from beneath the Greenland Ice Sheet collected during past deep-drilling campaigns reveal that the ice sheet was significantly smaller, and perhaps largely absent, sometime during the past 1.1 million years. These discoveries from decades-old basal samples motivate new, targeted sampling for cosmogenic-nuclide analysis beneath the ice sheet. Current drills available for retrieving bed material from the US Ice Drilling Program require < 700 m ice thickness and a frozen bed, while quartz-bearing bedrock lithologies are required for measuring a large suite of cosmogenic nuclides. We find that these and other requirements yield only ∼ 3.4 % of the Greenland Ice Sheet bed as a suitable drilling target using presently available technology. Additional factors related to scientific questions of interest are the following: which areas of the present ice sheet are the most sensitive to warming, where would a retreating ice sheet expose bare ground rather than leave a remnant ice cap, and which areas are most likely to remain frozen bedded throughout glacial cycles and thus best preserve cosmogenic nuclides? Here we identify locations beneath the Greenland Ice Sheet that are best suited for potential future drilling and analysis. These include sites bordering Inglefield Land in northwestern Greenland, near Victoria Fjord and Mylius-Erichsen Land in northern Greenland, and inland from the alpine topography along the ice margin in eastern and northeastern Greenland. Results from cosmogenic-nuclide analysis in new sub-ice bedrock cores from these areas would help to constrain dimensions of the Greenland Ice Sheet in the past.

Greenland↗
DOE OSTI2024

Cloud condensation nuclei concentrations derived from the CAMS reanalysis

Determining number concentrations of cloud condensation nuclei (CCN) is one of the first steps in the chain in analysis of cloud droplet formation, the direct microphysical link between aerosols and cloud droplets, and a process key for aerosol–cloud interactions (ACI). However, due to sparse coverage of in situ measurements and difficulties associated with retrievals from satellites, a global exploration of their magnitude, source as well as temporal and spatial distribution cannot be easily obtained. Thus, a better representation of CCN numbers is one of the goals for quantifying ACI processes and achieving uncertainty-reduced estimates of their associated radiative forcing. Here, we introduce a new CCN dataset which is derived based on aerosol mass mixing ratios from the latest Copernicus Atmosphere Monitoring Service reanalysis (CAMSRA) in a diagnostic model that uses CAMSRA aerosol properties and a simplified kappa-Köhler framework suitable for global models. The emitted aerosols in CAMSRA are not only based on input from emission inventories using aerosol observations, they also have a strong tie to satellite-retrieved aerosol optical depth (AOD) as this is assimilated as a constraining factor in the reanalysis. Furthermore, the reanalysis interpolates for cases of poor or missing retrievals and thus allows for a full spatiotemporal quantification of CCN numbers. The derived CCN dataset captures the general trend and spatial and temporal distribution of total CCN number concentrations and CCN from different aerosol species. A brief evaluation with ground-based in situ measurements demonstrates the improvement of the modelled CCN over the sole use of AOD as a proxy for CCN as the overall correlation coefficient improved from 0.37 to 0.71. However, we find the modelled CCN from CAMSRA to be generally high biased and find a particular erroneous overestimation at one heavily polluted site which emphasises the need for further validation. The CCN dataset (https://doi.org/10.26050/WDCC/QUAERERE_CCNCAMS_v1, Block, 2023), which is now freely available to users, features 3-D CCN number concentrations of global coverage for various supersaturations and aerosol species covering the years 2003–2021 with daily frequency. This dataset is one of its kind as it offers lots of opportunities to be used for evaluation in models and in ACI studies.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1993

Cognitive factors associated with immersion in virtual environments

Immersion into the dataspace provided by a computer, and the feeling of really being there or 'presence', are commonly acknowledged as the uniquely important features of virtual reality environments. How immersed one feels appears to be determined by a complex set of physical components and affordances of the environment, and as yet poorly understood psychological processes. Pimentel and Teixeira say that the experience of being immersed in a computer-generated world involves the same mental shift of 'suspending your disbelief for a period of time' as 'when you get wrapped up in a good novel or become absorbed in playing a computer game'. That sounds as if it could be right, but it would be good to get some evidence for these important conclusions. It might be even better to try to connect these statements with theoretical positions that try to do justice to complex cognitive processes. The basic precondition for understanding Virtual Reality (VR) is understanding the spatial representation systems that localize our bodies or egocenters in space. The effort to understand these cognitive processes is being driven with new energy by the pragmatic demands of successful virtual reality environments, but the literature is largely sparse and anecdotal.

Psotka, Joseph↗
DOE OSTI2015

WE-G-BRD-02: Characterizing Information Loss in a Sparse-Sampling-Based Dynamic MRI Sequence (k-T BLAST) for Lung Motion Monitoring

Purpose: Rapid MRI is an attractive, non-ionizing tool for soft-tissue-based monitoring of respiratory motion in thoracic and abdominal radiotherapy. One big challenge is to achieve high temporal resolution while maintaining adequate spatial resolution. K-t BLAST, sparse-sampling and reconstruction sequence based on a-priori information represents a potential solution. In this work, we investigated how much “true” motion information is lost as a-priori information is progressively added for faster imaging. Methods: Lung tumor motions in superior-inferior direction obtained from ten individuals were replayed into an in-house, MRI-compatible, programmable motion platform (50Hz refresh and 100microns precision). Six water-filled 1.5ml tubes were placed on it as fiducial markers. Dynamic marker motion within a coronal slice (FOV: 32×32cm{sup 2}, resolution: 0.67×0.67mm{sup 2}, slice-thickness: 5mm) was collected on 3.0T body scanner (Ingenia, Philips). Balanced-FFE (TE/TR: 1.3ms/2.5ms, flip-angle: 40degrees) was used in conjunction with k-t BLAST. Each motion was repeated four times as four k-t acceleration factors 1, 2, 5, and 16 (corresponding frame rates were 2.5, 4.7, 9.8, and 19.1Hz, respectively) were compared. For each image set, one average motion trajectory was computed from six marker displacements. Root mean square error (RMS) was used as a metric of spatial accuracy where measured trajectories were compared to original data. Results: Tumor motion was approximately 10mm. The mean(standard deviation) of respiratory rates over ten patients was 0.28(0.06)Hz. Cumulative distributions of tumor motion frequency spectra (0–25Hz) obtained from the patients showed that 90% of motion fell on 3.88Hz or less. Therefore, the frame rate must be a double or higher for accurate monitoring. The RMS errors over patients for k-t factors of 1, 2, 5, and 16 were.10(.04),.17(.04), .21(.06) and.26(.06)mm, respectively. Conclusions: K-t factor of 5 or higher can cover the high frequency component of tumor respiratory motion, while the estimated error of spatial accuracy was approximately.2mm.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2017

The Formation of Uranus and Neptune: Fine-tuning in Core Accretion

Uranus and Neptune are ice giants with ∼15% atmospheres by mass, which places them in a category intermediate between rocky planets and gas giants. These atmospheres are too massive to have been primarily outgassed, but they never underwent runaway gas accretion. The ice giants never reached critical core mass (M{sub crit}) in a full gas disk, but their cores are ≳M{sub crit}, suggesting that their envelopes were mainly accreted at the end of the disk lifetime. Pebble accretion calls into question traditional slow atmospheric growth during this phase. We show that the full-sized ice giants predominantly accreted gas from a disk depleted by at least a factor of ∼100. Such a disk dissipates in ≲10{sup 5} years. Why would both cores stay subcritical for the entire ∼Myr disk lifetime, only to reach M{sub crit} in the final 10{sup 5} years? This is fine tuned. Ice giants in the outer disk have atmospheric mass fractions comparable to the disk gas-to-solid ratio during the bulk of their gas accretion. This point in disk evolution coincides with a dynamical upheaval: the gas loses its ability to efficiently damp the core random velocities, allowing them to be gravitationally excited by Jupiter and Saturn. We suggest that the ice giants’ cores began growing on closer-in orbits (staying subcritical), and migrated out during this dynamical instability. There, their orbits circularized after accreting much of their mass in solids. Finally, they accreted their envelopes from a depleted nebula, where the sparseness of feeding-zone gas prevented runaway.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Growth and Stability of Stratiform Carrollite (CuCo2S4) in the Tenke-Fungurume Ore District, Central African Copperbelt

Abstract Carrollite (CuCo2S4) is the main ore mineral of the critical battery metal cobalt, yet, surprisingly, detailed characterization of its thermodynamic properties and of its trace element contents remain sparse in the scientific literature. To fill these knowledge gaps, we generated the first thermodynamic data set for carrollite using well-characterized samples obtained from the stratiform sulfide mineralization style at Fungurume 88 deposit in the Democratic Republic of Congo. Detailed mineralogical investigation of these same samples suggests that carrollite has limited ability to incorporate other ‘sweetener’ elements (except Cu and Ni) by substitution mechanisms. Modeling using CHNOSZ 2.0.0. software and an integrated Cu-Co-S-O thermodynamic data set provides insights into the stability of carrollite relative to other Co-bearing minerals under physicochemical conditions relevant to the Central African Copperbelt. Further exploration of these models indicates that dropping redox potential in the mineralizing system is insufficient to explain the mineral-scale and deposit-scale zonation patterns commonly observed in Cu-Co deposits of the Central African Copperbelt. Instead, factors such as increasing pH, decreasing T, decreasing sulfur activity, or some combination of changing physicochemical parameters may need to be invoked to explain these zonation trends. At the Fungurume 88 deposit, the absence of linnaeite (Co3S4) suggests that ore precipitation occurred at temperatures cooler than 211 °C. From our pH-logfO2 diagrams, and relative to higher temperatures, such a lower T regime is marked by a greater degree of offset between the modeled Cu and Co solubility contours, thus facilitating a more differentiated and zoned distribution of these two metals at the deposit scale. Our models are offered to the geoscience community as exceptionally useful tools for modeling and visualizing the Cu-Co-S-O systematics and may be applied to a variety of geological and geometallurgical questions in the Central African Copperbelt and other deposits elsewhere.

Mineralogy↗
DOE OSTI2018

Improving signal strength in serial crystallography with DIALS geometry refinement

The DIALS diffraction-modeling software package has been applied to serial crystallography data. Diffraction modeling is an exercise in determining the experimental parameters, such as incident beam wavelength, crystal unit cell and orientation, and detector geometry, that are most consistent with the observed positions of Bragg spots. These parameters can be refined by nonlinear least-squares fitting. In previous work, it has been challenging to refine both the positions of the sensors (metrology) on multipanel imaging detectors such as the CSPAD and the orientations of all of the crystals studied. Since the optimal models for metrology and crystal orientation are interdependent, alternate cycles of panel refinement and crystal refinement have been required. To simplify the process, a sparse linear algebra technique for solving the normal equations was implemented, allowing the detector panels to be refined simultaneously against the diffraction from thousands of crystals with excellent computational performance. Separately, it is shown how to refine the metrology of a second CSPAD detector, positioned at a distance of 2.5 m from the crystal, used for recording low-angle reflections. With the ability to jointly refine the detector position against the ensemble of all crystals used for structure determination, it is shown that ensemble refinement greatly reduces the apparent nonisomorphism that is often observed in the unit-cell distributions from still-shot serial crystallography. In addition, it is shown that batching the images by timestamp and re-refining the detector position can realistically model small, time-dependent variations in detector position relative to the sample, and thereby improve the integrated structure-factor intensity signal and heavy-atom anomalous peak heights.

36 MATERIALS SCIENCE↗
DOE OSTI2016

Stability-driven nonnegative matrix factorization to interpret spatial gene expression and build local gene networks

Spatial gene expression patterns enable the detection of local covariability and are extremely useful for identifying local gene interactions during normal development. The abundance of spatial expression data in recent years has led to the modeling and analysis of regulatory networks. The inherent complexity of such data makes it a challenge to extract biological information. We developed staNMF, a method that combines a scalable implementation of nonnegative matrix factorization (NMF) with a new stability-driven model selection criterion. When applied to a set of Drosophila early embryonic spatial gene expression images, one of the largest datasets of its kind, staNMF identified 21 principal patterns (PP). Providing a compact yet biologically interpretable representation of Drosophila expression patterns, PP are comparable to a fate map generated experimentally by laser ablation and show exceptional promise as a data-driven alternative to manual annotations. Our analysis mapped genes to cell-fate programs and assigned putative biological roles to uncharacterized genes. Finally, we used the PP to generate local transcription factor regulatory networks. Spatially local correlation networks were constructed for six PP that span along the embryonic anterior-posterior axis. Using a two-tail 5% cutoff on correlation, we reproduced 10 of the 11 links in the well-studied gap gene network. In conclusion, the performance of PP with the Drosophila data suggests that staNMF provides informative decompositions and constitutes a useful computational lens through which to extract biological insight from complex and often noisy gene expression data.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

Physics informed deep neural network embedded in a chemical transport model for the Amazon rainforest

Secondary organic aerosols (SOA) are fine particles in the atmosphere, which interact with clouds, radiation and affect the Earth’s energy budget. SOA formation involves chemistry in gas phase, aqueous aerosols, and clouds. Simulating these chemical processes involve solving a stiff set of differential equations, which are computationally expensive steps for three-dimensional chemical transport models. Deep neural networks (DNNs) are universal function approximators that could be used to represent the complex nonlinear changes in aerosol physical and chemical processes; however, key challenges such as generalizability to extended time periods, preservation of mass balance, simulating sparse model outputs, and maintaining physical constraints have limited their use in atmospheric chemistry. Here, we develop an approach of using a physics-informed DNN that overcomes previous such challenges and demonstrates its applicability for the chemical formation processes of isoprene epoxydiol SOA (IEPOX-SOA) over the Amazon rainforest. The DNN is trained with data generated by simulating IEPOX-SOA over the entire atmospheric column, using the Weather Research and Forecasting Model coupled with Chemistry (WRF-Chem). The trained DNN is then embedded within WRF-Chem to replace the computationally expensive default solver of IEPOX-SOA formation. The trained DNN predictions generalizes well with the default model simulation of the IEPOX-SOA mass concentrations and its size distribution (20 size bins) over several days of simulations in both dry and wet seasons. The embedded DNN reduces the computational expense of WRF-Chem by a factor of 2. Our approach shows promise in terms of application to other computationally expensive chemistry solvers in climate models.

54 ENVIRONMENTAL SCIENCES↗