Search NASA⌕ Search

SEARCH · Search NASA

Results for “Approximate deconvolution”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Surface Charge Distribution Mapping Using an Electrostatic Probe

Undesired accumulation of charge on dielectric materials causing electrostatic discharges can be an issue in pulsed power systems and electrical systems in general. As such, an understanding of surface charge distribution, charge accumulation, and decay is required. An in-house no-contact electrostatic probe designed with a 2 mm resolution was fabricated to measure and map surface charge distribution. Designed as a contactless instrument, the probe consisting of a metal rod and low leakage amplifier circuit probes the electric potential near the surface. A scan of the surface then provides the raw data, a potential distribution. The actual probe response - i.e., the surface charge to voltage transfer function, is accounted for in post-processing. This is accomplished via an Inverse Wien Filter - a technique often applied in image processing - to deconvolve the probe response from the measured data. A commercially available electrostatic probe, the Trek 341B meter with a 3455ET probe, capable of measuring +/- 20 kV was compared to the in- house probe that is designed to cover a wider range from +/- 35 kV. A resolution better than the simple probe resolution is achieved through the distribution’s scanning voltage method and deconvolution. Applying repeated scans, surface charge decay was tracked on various polymer materials to determine the material and environment dependence; materials included were PA6, PTFE, and others. As an example of material dependence, samples charged to 20 kV at 65% humidity experience full charge decay in approximately 45 seconds for PA5 (152 mm dia.), while it took about 100 times longer for PTFE (51 mm dia.) to observe the same decay/redistribution of charge.

36 MATERIALS SCIENCE↗
DOE OSTI2024

TNet: A Model-Constrained Tikhonov Network Approach for Inverse Problems

Deep learning (DL), in particular deep neural networks, by default is purely data-driven and in general does not require physics. This is the strength of DL but also one of its key limitations when applied to science and engineering problems in which underlying physical properties—such as stability, conservation, and positivity—and accuracy are required. DL methods in their original forms are often not capable of respecting the underlying mathematical models or achieving desired accuracy even in big-data regimes. On the other hand, many data-driven science and engineering problems, such as inverse problems, typically have limited experimental or observational data, and DL would overfit the data in this case. Leveraging information encoded in the underlying mathematical models, we argue, not only compensates for missing information in low data regimes but also provides opportunities to equip DL methods with the underlying physics, hence promoting better generalization. This paper develops a model-constrained DL approach and its variant TNet—a Tikhonov neural network—which are capable of learning not only information hidden in the training data but also in the underlying mathematical models to solve inverse problems governed by partial differential equations in low data regimes. We provide the constructions and some theoretical results for the proposed approaches for both linear and nonlinear inverse problems. Since TNet is designed to learn inverse solutions with Tikhonov regularization, it is interpretable: in fact it recovers Tikhonov solutions for linear cases while potentially approximating Tikhonov solutions for nonlinear inverse problems. We also prove that data randomization can enhance not only the smoothness of the networks but also their generalizations. Comprehensive numerical results confirm the theoretical findings and show that with even as little as 1 training data sample for one-dimensional (1D) deconvolution, 5 for an inverse 2D heat conductivity problem, 100 for inverse initial conditions for a time-dependent 2D Burgers’s equation, and 50 for inverse initial conditions for 2D Navier–Stokes equations, TNet solutions can be as accurate as Tikhonov solutions while being several orders of magnitude faster. Furthermore, this is possible owing to the model-constrained term, replications, and randomization.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Design Trade-Offs in Composite Fuel Cell Membranes: Effects of Reinforcement and Chemical Additives

Perfluorosulfonic acid (PFSA) membranes are critical components in proton exchange membrane fuel cells, where performance depends on balancing ionic conductivity, mechanical durability, and chemical stability. This study characterizes a composite membrane (NC700) featuring PFSA-impregnated expanded polytetrafluoroethylene (ePTFE) reinforcement and cerium-based radical scavengers, benchmarked against unreinforced NR211. Complementary techniques, including electron microscopy, X-ray scattering, infrared spectroscopy, thermogravimetric analysis, and dynamic mechanical analysis, identify the structural and compositional strategies employed in NC700. Water sorption isotherms reveal lower water uptake for NC700 across all conditions, attributed to reinforcement and cerium incorporation. Reinforcement reduces in-plane swelling from 11% to 2.1% at 90% RH, confirming strong swelling anisotropy, while maintaining mechanical properties at elevated temperatures. While the ionic conductivity of NC700 is approximately 10% lower than that of NR211, the reduced thickness yields a 40% decrease in calculated area-specific resistance, suggesting the composite architecture can favorably shift the conductivity-stability trade-off. The composite structure also reduces gas permeability, indicating potential for improved separator function alongside favorable transport properties. Systematic deconvolution of reinforcement and additive contributions shows that conductivity losses from cerium incorporation are largely offset by gains from the lower equivalent-weight polymer, providing quantitative relationships that may guide composite membrane design for fuel cells and other electrochemical applications.

25 ENERGY STORAGE↗
NASA NTRS1999

Correction Factor for Gaussian Deconvolution of Optically Thick Linewidths in Homogeneous Sources

Profiles of optically thick, non-Gaussian emission line profiles convoluted with Gaussian instrumental profiles are constructed, and are deconvoluted on the usual Gaussian basis to examine the departure from accuracy thereby caused in "measured" linewidths. It is found that "measured" linewidths underestimate the true linewidths of optically thick lines, by a factor which depends on the resolution factor r congruent to Doppler width/instrumental width and on the optical thickness tau(sub 0). An approximating expression is obtained for this factor, applicable in the range of at least 0 <= tau(sub 0) <= 10, which can provide estimates of the true linewidth and optical thickness.

Kastner, S. O.↗
NASA NTRS2012

Wavespace-Based Coherent Deconvolution

Array deconvolution is commonly used in aeroacoustic analysis to remove the influence of a microphone array's point spread function from a conventional beamforming map. Unfortunately, the majority of deconvolution algorithms assume that the acoustic sources in a measurement are incoherent, which can be problematic for some aeroacoustic phenomena with coherent, spatially-distributed characteristics. While several algorithms have been proposed to handle coherent sources, some are computationally intractable for many problems while others require restrictive assumptions about the source field. Newer generalized inverse techniques hold promise, but are still under investigation for general use. An alternate coherent deconvolution method is proposed based on a wavespace transformation of the array data. Wavespace analysis offers advantages over curved-wave array processing, such as providing an explicit shift-invariance in the convolution of the array sampling function with the acoustic wave field. However, usage of the wavespace transformation assumes the acoustic wave field is accurately approximated as a superposition of plane wave fields, regardless of true wavefront curvature. The wavespace technique leverages Fourier transforms to quickly evaluate a shift-invariant convolution. The method is derived for and applied to ideal incoherent and coherent plane wave fields to demonstrate its ability to determine magnitude and relative phase of multiple coherent sources. Multi-scale processing is explored as a means of accelerating solution convergence. A case with a spherical wave front is evaluated. Finally, a trailing edge noise experiment case is considered. Results show the method successfully deconvolves incoherent, partially-coherent, and coherent plane wave fields to a degree necessary for quantitative evaluation. Curved wave front cases warrant further investigation. A potential extension to nearfield beamforming is proposed.

Bahr, Christopher J.↗
NASA NTRS2003

The Filtered Abel Transform and Its Application in Combustion Diagnostics

Many non-intrusive combustion diagnosis methods generate line-of-sight projections of a flame field. To reconstruct the spatial field of the measured properties, these projections need to be deconvoluted. When the spatial field is axisymmetric, commonly used deconvolution method include the Abel transforms, the onion peeling method and the two-dimensional Fourier transform method and its derivatives such as the filtered back projection methods. This paper proposes a new approach for performing the Abel transform method is developed, which possesses the exactness of the Abel transform and the flexibility of incorporating various filters in the reconstruction process. The Abel transform is an exact method and the simplest among these commonly used methods. It is evinced in this paper that all the exact reconstruction methods for axisymmetric distributions must be equivalent to the Abel transform because of its uniqueness and exactness. Detailed proof is presented to show that the two dimensional Fourier methods when applied to axisymmetric cases is identical to the Abel transform. Discrepancies among various reconstruction method stem from the different approximations made to perform numerical calculations. An equation relating the spectrum of a set of projection date to that of the corresponding spatial distribution is obtained, which shows that the spectrum of the projection is equal to the Abel transform of the spectrum of the corresponding spatial distribution. From the equation, if either the projection or the distribution is bandwidth limited, the other is also bandwidth limited, and both have the same bandwidth. If the two are not bandwidth limited, the Abel transform has a bias against low wave number components in most practical cases. This explains why the Abel transform and all exact deconvolution methods are sensitive to high wave number noises. The filtered Abel transform is based on the fact that the Abel transform of filtered projection data is equal to an integral transform of the original projection data with the kernel function being the Abel transform of the filtering function. The kernel function is independent of the projection data and can be obtained separately when the filtering function is selected. Users can select the best filtering function for a particular set of experimental data. When the kernal function is obtained, it can be used repeatedly to a number of projection data sets (rovs) from the same experiment. When an entire flame image that contains a large number of projection lines needs to be processed, the new approach significantly reduces computational effort in comparison with the conventional approach in which each projection data set is deconvoluted separately. Computer codes have been developed to perform the filter Abel transform for an entire flame field. Measured soot volume fraction data of a jet diffusion flame are processed as an example.

Simons, Stephen N.↗
NASA NTRS1992

Far-infrared emission, gas, and ultraluminous HII regions in M101

We have made deconvolved maps of M101 in each of the 4 IRAS bands using a new maximum-entropy based model for reconstructing distributions with correlated structure on multiple scales. This new deconvolution procedure, developed by Weir (1992, J.Opt.Soc.Am., submitted), is superior to other maximum entropy-based techniques for several reasons. For our purposes, an important advantage is that it has less artifacts and greatly reduced systematic biases compared with the 'HiRes' maps of M101 produced by Aumann, Fowler, and Melnyk (1990, AJ, 99, 1674), thus it is accurate enough to perform aperture photometry. The deconvolved maps have a maximum resolution of approximately 30 sec, sufficient to resolve the brightest HII complexes and much of the spiral structure. The new maps and our comparison of the far-infrared, gas, and optical distributions are being written up in Kenney, Weir & Scoville (1992, in preparation). The quality of the reconstructed far-infrared maps is good enough for us to carry out most of the analyses outlined in the original proposal. From a ratio of the 60 micron and 100 micron maps, we have found that the ultraluminous HII complexes in the outer galaxy have the hottest dust temperatures with T approximately 50-60 K, which is twice as hot as most of the disk. Their extraordinary luminosity in the far-infrared is due to a relatively small amount of dust being heated to high temperatures, rather than a large concentration of dust and gas. A map of the dust opacity at 60 microns shows good overall agreement with a map of cold gas (HI+H2), indicating that throughout most of the galaxy only approximately 20 percent of the dust is warm enough to be detected by IRAS. The ultraluminous HII complexes have a high luminosity-to-gas mass ratio, independent of whether the far-infrared or the H-alpha emission line is used to measure the luminosity, which implies that gas is being converted into high mass stars more rapidly in these complexes compared to other locations in the disk.

Kenney, J. D. P.↗
NASA NTRS2007

Highly Collimated Jets and Wide-angle Outflows in HH 46/47: New Evidence from Spitzer Infrared Images

We present new details of the structure and morphology of the jets and outflows in HH 46/47 as seen in Spitzer infrared images from IRAC and MIPS, reprocessed using the 'HiRes' deconvolution technique. HiRes improves the visualization of spatial morphology by enhancing resolution (to subarcsecond levels in IRAC bands) and removing the contaminating side lobes from bright sources. In addition to sharper views of previously reported bow shocks, we have detected (1) the sharply delineated cavity walls of the wide-angle biconical outflow, seen in scattered light on both sides of the protostar, (2) several very narrow jet features at distances approximately 400 AU to approximately 0.1 pc from the star, and (3) compact emissions at MIPS 24 m with the jet heads, tracing the hottest atomic/ionic gas in the bow shocks. Together the IRAC and MIPS images provide a more complete picture of the bow shocks, tracing both the molecular and atomic/ionic gases, respectively. The narrow width and alignment of all jet-related features indicate a high degree of jet collimation and low divergence (width of approximately 400 AU increasing by only a factor of 2.3 over 0.2 pc). The morphology of this jet, bow shocks, wide-angle outflows, and the fact that the jet is nonprecessing and episodic, constrain the mechanisms for producing the jet's entrained molecular gas, and origins of the fast jet, and slower wide-angle outflow.

Herbig-Haro objects↗
NASA NTRS2015

Characterization of Flap Edge Noise Radiation from a High-Fidelity Airframe Model

The results of an experimental study of the noise generated by a baseline high-fidelity airframe model are presented. The test campaign was conducted in the open-jet test section of the NASA Langley 14- by 22-foot Subsonic Tunnel on an 18%-scale, semi-span Gulfstream airframe model incorporating a trailing edge flap and main landing gear. Unsteady surface pressure measurements were obtained from a series of sensors positioned along the two flap edges, and far field acoustic measurements were obtained using a 97-microphone phased array that viewed the pressure side of the airframe. The DAMAS array deconvolution method was employed to determine the locations and strengths of relevant noise sources in the vicinity of the flap edges and the landing gear. A Coherent Output Power (COP) spectral method was used to couple the unsteady surface pressures measured along the flap edges with the phased array output. The results indicate that outboard flap edge noise is dominated by the flap bulb seal cavity with very strong COP coherence over an approximate model-scale frequency range of 1 to 5 kHz observed between the array output and those unsteady pressure sensors nearest the aft end of the cavity. An examination of experimental COP spectra for the inboard flap proved inconclusive, most likely due to a combination of coherence loss caused by decorrelation of acoustic waves propagating through the thick wind tunnel shear layer and contamination of the spectra by tunnel background noise at lower frequencies. Directivity measurements obtained from integration of DAMAS pressure-squared values over defined geometric zones around the model show that the baseline flap and landing gear are only moderately directional as a function of polar emission angle.

Humphreys, William M., Jr.↗
NASA NTRS1994

Dimensions and Fragmentation of the Nuclei of Comet Shoemaker-Levi 9

Central regions on the digital maps of 13 nuclear condensations of Comet Shoemaker-Levi 9, obtained with the Planetary Camera of the Hubble Space Telescope on January 27, March 30, and July 4, 1994, have been analyzed with the aim to identify the presence of distinct, major fragments in each condensation, to deconvolve their contributions to the signal that also includes the contribution from a surrounding cloud of dust (modeled as an extended source, using two different laws), to estimate the dimensions of the fragments and to study their temporal variations, and to determine the spatial distributions of the fragments as projected on to the plane of the sky. The deconvolution method applied is described and the results of the analysis are summarized, including the finding that sizable fragments did survive until the time of atmospheric entry. This result does not contradict evidence of the comet's continuing, apparently spontaneous fragmentation, which still went on long after the extremely close approach to Jupiter in July 1992 and which, because of the Jovian tidal effects, may have intensified in the final days before the crash on Jupiter. Since the developed approach is based on certain premises and involves approximations, the results should be viewed as preliminary and the problem should be the subject of further investigation.

Comet Shoemaker-Levi 9 Hubble Space Telescope come↗
NASA NTRSDate not supplied

Phased Array Characterization of Slat Noise Radiation from a High-Lift Common Research Model

A test campaign was conducted in the open-jet test section of the NASA Langley 14- by 22-Foot Subsonic Tunnel on a 10%-scale, semispan version of a High-Lift Common Research Model (CRM-HL) incorporating a leading-edge slat, trailing-edge flap and removable high-fidelity main landing gear. Far field acoustic measurements were obtained on baseline and acoustically treated model configurations using a traversing 97-microphone phased array that viewed the pressure side of the airframe. A modified version of the DAMAS deconvolution method incorporating corrections for shear layer acoustic wave decorrelation was employed to determine the locations and strengths of relevant noise sources along the span of the slat and in the vicinity of the flap edges and landing gear. The main goal of the test campaign was to use the array data to evaluate the effectiveness of various slat noise reduction concepts. It was found that slat-gap fillers produced significant broadband noise reduction for all Mach numbers and airfoil angles of attack that were investigated. Mach number scaling revealed an approximate velocity to the sixth power relationship for the slat noise at higher frequencies. Directivity measurements obtained from integration of DAMAS pressure-squared values over defined geometric zones around the model showed that baseline slat noise radiation approximated a dipole as a function of polar emission angle, becoming approximately omnidirectional with the introduction of the slat-gap filler. The test campaign clearly demonstrated the noise reduction benefits that can be obtained by applying appropriate treatments to leading edge slats on commercial transport-class aircraft.

William M Humphreys, Jr.↗
NASA NTRS2011

Aeroacoustic Characterization of the NASA Ames Experimental Aero-Physics Branch 32- by 48-Inch Subsonic Wind Tunnel with a 24-Element Phased Microphone Array

The Aero-Physics Branch at NASA Ames Research Center utilizes a 32- by 48-inch subsonic wind tunnel for aerodynamics research. The feasibility of acquiring acoustic measurements with a phased microphone array was recently explored. Acoustic characterization of the wind tunnel was carried out with a floor-mounted 24-element array and two ceiling-mounted speakers. The minimum speaker level for accurate level measurement was evaluated for various tunnel speeds up to a Mach number of 0.15 and streamwise speaker locations. A variety of post-processing procedures, including conventional beamforming and deconvolutional processing such as TIDY, were used. The speaker measurements, with and without flow, were used to compare actual versus simulated in-flow speaker calibrations. Data for wind-off speaker sound and wind-on tunnel background noise were found valuable for predicting sound levels for which the speakers were detectable when the wind was on. Speaker sources were detectable 2 - 10 dB below the peak background noise level with conventional data processing. The effectiveness of background noise cross-spectral matrix subtraction was assessed and found to improve the detectability of test sound sources by approximately 10 dB over a wide frequency range.

Costanza, Bryan T.↗
DOE OSTI2022

Raman signatures of detonation soot

Soot generated from the detonation of nine explosives was investigated with Raman microspectroscopy and high-resolution transmission electron microscopy (HRTEM). The first order bands, located at approximately 1,580 (G), 1,350 (D1), 1,620 cm –1 (D2), 1,500 (D3), and 1,200 (D4), are used to provide information on the carbon allotropes and defects. The intensity ratios and the full-width at half-maximum provide an indication of the disorder in the material. The Raman spectra of soot acquired from the detonation of nine explosives showed differences which can be used to identify the precursor explosive. Overall, most differences in the spectra were correlated with the presence of amorphous carbon in the soot. In additional to identification of distinct bands in the spectra via deconvolution, principal component analysis was also used to differentiate the different types of soot. HRTEM performed on soot samples with different explosive precursors identified several distinct allotropes of carbon such as nanodiamonds, carbon onions, graphite fibers, and amorphous carbon.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗
NASA NTRS2015

Fe and O EELS Studies of Ion Irradiated Murchison CM2 Carbonaceous Chondrite Matrix

Introduction: The physical and chemical response of hydrated carbonaceous chondrite materials to space weathering processes is poorly understood. Improving this understanding is a key part of establishing how regoliths on primitive carbonaceous asteroids respond to space weathering processes, knowledge that supports future sample return missions (Hayabusa 2 and OSIRISREx) that are targeting objects of this type. We previously reported on He+ irradiation of Murchison matrix and showed that the irradiation resulted in amorphization of the matrix phyllosilicates, loss of OH, and surface vesiculation. Here, we report electron energy-loss spectroscopy (EELS) measurements of the irradiated material with emphasis on the Fe and O speciation. Sample and Methods: A polished thin section of the Murchison CM2 carbonaceous chondrite was irradiated with 4 kilovolts He(+) (normal incidence) to a total dose of 1 x 10(exp 18) He(+) per square centimeter. We extracted thin sections from both irradiated and unirradiated regions in matrix using focused ion beam (FIB) techniques with electron beam deposition for the protective carbon strap to minimize surface damage artifacts from the FIB milling. The FIB sections were analyzed using a JEOL 2500SE scanning and transmission electron microscope (STEM) equipped with a Gatan Tridiem imaging filter. EELS spectra were collected from 50 nanometer diameter regions with an energy resolution of 0.7 electronvolts FWHM at the zero loss. EELS spectra were collected at low electron doses to minimize possible artifacts from electron-beam irradiation damage. Results and Discussion: Fe L (sub 2,3) EELS spectra from matrix phyllosilicates in CM chondrites show mixed Fe(2+)/Fe(3+) oxidation states with Fe(3+)/Sigma Fe approximately 0.5. Fe L(sub 2,3) spectra from the irradiated/ amorphized matrix phyllosilicates show higher Fe(2+)/Fe(3+) ratios compared to spectra obtained from pristine material at depths beyond the implantation/amorphization layer. We also obtained O Ka spectra from phyllosilicates in both regions of the sample. The O Ka spectra show a pre-edge feature at approximately 530.5 electronvolts that is related to O 2p states hybridized with Fe 3d states. The intensity ratio of the O Ka pre-edge peak relative to the main part of the O Ka edge (that results from transitions of O 1s to 2p states) is lower in the irradiated layer compared to the pristine material and may reflect the loss of O (as OH) as was observed by IR spectroscopy. Conclusions: In addition to amorphization and OH loss, EELS spectra of He(+) irradiated matrix phyllosilicates in Murchison show that some of the Fe(3+) is reduced to Fe(2+). Spectral deconvolution is underway to extract quantitative ratios from the EELS spectra.

Keller, L. P.↗
NASA NTRS1994

Energetic electrons in the magnetophere of Saturn

The energy spectra and angular distributions of electrons observed by Pioneer 11 as a function of radial distance in the inner magnetosphere of Saturn are reanalyzed and phase space densities are then calculated. The radial dependence of phase space density requires a distributed loss process. The loss is greatest in the region of the E ring (5.5 less than L less than 8.5) and is attributed to collisions with the ring particles in agreement with earlier work by Van Allen et al. (1980b). Quantitative analysis yields the following properties of the E ring: the particle radii are in the range of 4 x 10(exp -5) to 3.2 x 10(exp -4) cm and the thickness of the ring is approximately 3 R(sub s). Between the inner edge of the E ring (5.5 R(sub s)) and the outer edge of the A ring (2.3 R(sub s)) there are more energetic electrons than can be supplied by radial diffusion from an external source. Detailed calculations show that a cosmic ray albedo neutron decay (CRAND) source in the A and B rings is a plausible source for this excess. The radial diffusion coefficient required to explain the E ring absorption and CRAND source for electrons is 1 x 10(exp -12) greater than D(sub 0) greater than 3 x 10(exp -12) R(sub s)(exp 2)/s, assuming that D(sub LL) = D(sub 0)L(exp 3). As part of the reanalysis program, a method for the deconvolution of pitch angle distributions observed by simple detectors on a rotating spacecraft is developed. This process removes the instrumental response and rotational smear due to finite sampling periods and yields true angular distributions.

Randall, B. A.↗
NASA NTRS1994

Energetic electrons in the magnetosphere of Saturn

The energy spectra and angular distributions of electrons observed by Pioneer 11 as a function of radial distance in the inner magnetosphere of Saturn are reanalyzed and phase space densities are then calculated. The radial dependence of phase space density requires a distributed loss process. The loss is greatest in the region of the E ring (5.5 less than L less than 8.5) and is attributed to collisions with the ring particles in agreement with earlier work by Van Allen et al. (1980). Quantitative analysis yields the following properties of the E ring: the particle radii are in the range of 4 x 10(exp -5) to 3.2 x 10(exp -4) cm and the thickness of the ring is approximately 3 R(sub s). Between the inner edge of the E ring (5.5 R(sub s)) and the outer edge of the A ring (2.3 R(sub s)) there are more energetic electrons than can be supplied by radial diffusion from an external source. Detailed calculations show that a cosmic ray albedo neutron decay (CRAND) source in the A and B rings is a plausible source for this excess. The radial diffusion coefficient required to explain the E ring absorption and CRAND source for electrons is 1 x 10(exp -12) greater than D(sub 0) greater than 3 x 10(exp -12) R(exp 2, sub s)/s, assuming that D(sub LL) = D(sub 0)L(exp 3). As part of the reanalysis program, a method for the deconvolution of pitch angle distributions observed by simple detectors on a rotating spacecraft is developed. This process removes the instrumental response and rotational smear due to finite sampling periods and yields true angular distributions.

Randall, B. A.↗
NASA NTRS2023

Relationship Between the Sub-Micron Fraction (SMF) and Fine-Mode Fraction (FMF) in the Context of AERONET Retrievals

The sub-micron (SM) aerosol optical depth (AOD) is an optical separation based on the fraction of particles below a specified cutoff radius of the particle size distribution (PSD) at a given particle radius. It is fundamentally different from spectrally separated FM (fine-mode) AOD. We present a simple (AOD-normalized) SM fraction versus FM fraction (SMF vs. FMF) linear equation that explains the well-recognized empirical result of SMF generally being greater than the FMF. The AERONET inversion (AERinv) products (combined inputs of spectral AOD and sky radiance) and the spectral deconvolution algorithm (SDA) products (input of AOD spectra) enable, respectively, an empirical SMF vs. FMF comparison at similar (columnar) remote sensing scales across a variety of aerosol types. SMF (AERinv-derived) vs. FMF (SDA-derived) behavior is primarily dependent on the relative truncated portion (ε c ) of the coarse-mode (CM) AOD associated with the cutoff portion of the CM PSD and, to a second order, the cutoff FM PSD and FM AOD (ε f ). The SMF vs. FMF equation largely explains the SMF vs. FMF behavior of the AERinv vs. SDA products as a function of PSD cutoff radius (“inflection point”) across an ensemble of AERONET sites and aerosol types (urban-industrial, biomass burning, dust, maritime and a mixed class of Arctic aerosols). The overarching dynamic was that the linear SMF vs. FMF relation pivots clockwise about the approximate (SMF, FMF) singularity of (1, 1) in a “linearly inverse” fashion (slope and intercept of approximately 1−ε c and ε c ) with increasing cutoff radius. SMF vs. FMF slopes and intercepts derived from AERinv and SDA retrievals confirmed the general domination of ε c over ε f in controlling that dynamic. A more general conclusion is the apparent confirmation that the optical impact of truncating modal (whole) PSD features can be detected by an SMF vs. FMF analysis.

Norman T. O'Neill↗
DOE OSTI2021

Adversarial sampling of unknown and high-dimensional conditional distributions

Many engineering problems require the prediction of realization-to-realization variability or a refined description of modeled quantities. In that case, it is necessary to sample elements from unknown high-dimensional spaces with possibly millions of degrees of freedom. While there exist methods able to sample elements from probability density functions (PDF) with known shapes, several approximations need to be made when the distribution is unknown. In this paper the sampling method, as well as the inference of the underlying distribution, are both handled with a data-driven method known as generative adversarial networks (GAN), which trains two competing neural networks to produce a network that can effectively generate samples from the training set distribution. In practice, it is often necessary to draw samples from conditional distributions. When the conditional variables are continuous, only one (if any) data point corresponding to a particular value of a conditioning variable may be available, which is not sufficient to estimate the conditional distribution. This work handles this problem using an a priori estimation of the conditional moments of a PDF. Herein, two approaches, stochastic estimation, and an external neural network are compared for computing these moments; however, any preferred method can be used. The algorithm is demonstrated in the case of the deconvolution of a filtered turbulent flow field. It is shown that all the versions of the proposed algorithm effectively sample the target conditional distribution with minimal impact on the quality of the samples compared to state-of-the-art methods. Additionally, the procedure can be used as a metric for the diversity of samples generated by a conditional GAN (cGAN) conditioned with continuous variables.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗