Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed inference”

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 217 records · Page 12

Physics of spinning gases and plasmas

Initially motivated by the problem of compression of spinning plasma in Z-pinch devices and related applications, the thesis explores a number of interesting smaller-scale problems related to physics of gas and plasma rotation. In particular, thermodynamics of ideal spinning gas is studied. It is found that rotation modifies the heat capacity of the gas and reduces the gas compressibility. It is also proposed that, by performing a series of measurement of external parameters of a spinning gas, one can infer the distribution of masses of gas constituents. It is also proposed how to use the rotation-dependent heat capacity for improving the thermodynamic efficiency of internal combustion engines. To that end, two possible engine embodiments are proposed and explored in detail. In addition, a transient piezothermal effect is discovered numerically and is given a theoretical explanation. The effect consists of the formation of a radial temperature gradient driven by gas heating or compression along the rotation axis. By elaborating on this idea, a theoretical explanation is proposed also for the operation of so-called vortex tubes, which so far have been lacking rigorous theory. Finally, adiabatic compression of spinning plasmas and ionized gases are considered, and the effect of the electrostatic interactions on the compressibility and heat capacity is predicted.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Dark Energy Survey Year 6 Results: Redshift Calibration of the MagLim++ Lens Sample

In this work, we derive and calibrate the redshift distribution of the MagLim++ lens galaxy sample used in the Dark Energy Survey Year 6 (DES Y6) 3x2pt cosmology analysis. The 3x2pt analysis combines galaxy clustering from the lens galaxy sample and weak gravitational lensing. The redshift distributions are inferred using the SOMPZ method - a Self-Organizing Map framework that combines deep-field multi-band photometry, wide-field data, and a synthetic source injection (Balrog) catalog. Key improvements over the DES Year 3 (Y3) calibration include a noise-weighted SOM metric, an expanded Balrog catalogue, and an improved scheme for propagating systematic uncertainties, which allows us to generate O($10^8$) redshift realizations that collectively span the dominant sources of uncertainty. These realizations are then combined with independent clustering-redshift measurements via importance sampling. The resulting calibration achieves typical uncertainties on the mean redshift of 1-2%, corresponding to a 20-30% average reduction relative to DES Y3. We compress the $n(z)$ uncertainties into a small number of orthogonal modes for use in cosmological inference. Marginalizing over these modes leads to only a minor degradation in cosmological constraints. This analysis establishes the MagLim++ sample as a robust lens sample for precision cosmology with DES Y6 and provides a scalable framework for future surveys.

Giannini, G. [Chicago U., Astron. Astrophys. Ctr.;↗

Data Center High-Temperature Liquid Cooling and Heat Reuse Techno-Economic Study: Preprint

Data centers are energy-intensive facilities with growing demands for efficiency and cost-effective operations. Smaller, more distributed edge inference data centers are expected to proliferate as AI applications require low latency closer to the user of AI tools, which presents a growing opportunity to explore the systems implications of liquid cooling on water and energy use. This study analyzes the implementation of high-temperature liquid cooling systems in a prototypical inference 1-MW data center and explores the potential for heat reuse across varying climates with a goal to optimize energy efficiency, reduce capital and operational costs, and identify opportunities for high-performance cooling and water use reduction infrastructure. This analysis evaluated configurations utilizing a peak day hourly sizing and systems performance spreadsheet to evaluate design and operational conditions from which component sizes, installed cost, operational cost, and performance metrics were determined for the Base case and the Elevated case. The techno-economic analysis included heat reuse applications across a range of heat recovery temperatures and heat rejection options. The analysis shows that high-temperature liquid cooling allows for improved energy efficiency, lower water consumption, and lower capital costs compared to traditional cooling approaches. Transitioning to elevated water inlet/outlet temperatures (50 degrees C/60 degrees C) eliminates the need for chillers, cooling towers, and heat recovery equipment in many scenarios across three distinct climate zones. This results in up to 75% capital cost savings for the cooling and heat recovery equipment, and with significantly reduced water consumption, especially in non-heat reuse applications. Heat generated from data centers can also be repurposed for space heating, domestic hot water, and other applications, and is most cost-effective when data center outlet temperatures exceed 55-60 degrees C.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Obscuration-dependent evolution of active galactic nuclei

We aim to constrain the evolution of active galactic nuclei (AGNs) as a function of obscuration using an X-ray-selected sample of ∼2000 AGNs from a multi-tiered survey including the CDFS, AEGIS-XD, COSMOS, and XMM-XXL fields. The spectra of individual X-ray sources are analyzed using a Bayesian methodology with a physically realistic model to infer the posterior distribution of the hydrogen column density and intrinsic X-ray luminosity. We develop a novel non-parametric method that allows us to robustly infer the distribution of the AGN population in X-ray luminosity, redshift, and obscuring column density, relying only on minimal smoothness assumptions. Our analysis properly incorporates uncertainties from low count spectra, photometric redshift measurements, association incompleteness, and the limited sample size. We find that obscured AGNs with N {sub H} > 10{sup 22} cm{sup –2} account for 77{sub −5}{sup +4}% of the number density and luminosity density of the accretion supermassive black hole population with L {sub X} > 10{sup 43} erg s{sup –1}, averaged over cosmic time. Compton-thick AGNs account for approximately half the number and luminosity density of the obscured population, and 38{sub −7}{sup +8}% of the total. We also find evidence that the evolution is obscuration dependent, with the strongest evolution around N {sub H} ≈ 10{sup 23} cm{sup –2}. We highlight this by measuring the obscured fraction in Compton-thin AGNs, which increases toward z ∼ 3, where it is 25% higher than the local value. In contrast, the fraction of Compton-thick AGNs is consistent with being constant at ≈35%, independent of redshift and accretion luminosity. We discuss our findings in the context of existing models and conclude that the observed evolution is, to first order, a side effect of anti-hierarchical growth.

79 ASTRONOMY AND ASTROPHYSICS↗

Relationships between the Raindrop Size Distribution and Properties of the Environment and Clouds Inferred from TRMM

Variability in the raindrop sized distribution (DSD) has long been recognized as a source of uncertainty in relationships between radar reflectivity Z and rain rate R. In this study, we analyze DSD retrievals from two years of data gathered by the Tropical Rainfall Measuring Mission (TRMM) satellite and processed with a combined radar-radiometer retrieval algorithm over the global oceans equatorward of 35◦. Numerous variables describing properties of each reflectivity profile, large-scale organization, and the background environment are examined for relationships to the reflectivity-normalized median drop diameter, epsilonDSD. In general, we find that higher freezing levels and relative humidities are associated with smaller epsilonDSD. Within a given environment, the mesoscale organization of precipitation and the vertical profile of reflectivity are associated with DSD characteristics. In the tropics, the smallest epsilonDSD values are found in large but shallow convective systems, where warm rain formation processes are thought to be predominant, whereas larger sizes are found in the stratiform regions of organized deep convection. In the extratropics, the largest epsilonDSD values are found in the scattered convection that occurs when cold, dry continental air moves over the much warmer ocean after the passage of a cold front. The geographical distribution of the retrieved DSDs is consistent with many of the observed regional Z−R relationships found in the literature as well as discrepancies between the TRMM radar-only and radiometer-only precipitation products. In particular, mid-latitude and tropical regions near land tend to have larger drops for a given reflectivity, whereas the smallest drops are found in the eastern Pacific Intertropical Convergence Zone.

raindrop size distribution↗

THE BLACK HOLE MASS DISTRIBUTION IN THE GALAXY

We use dynamical mass measurements of 16 black holes in transient low-mass X-ray binaries to infer the stellar black hole mass distribution in the parent population. We find that the observations are best described by a narrow mass distribution at 7.8 {+-} 1.2 M{sub sun}. We identify a selection effect related to the choice of targets for optical follow-ups that results in a flux-limited sample. We demonstrate, however, that this selection effect does not introduce a bias in the observed distribution and cannot explain the absence of black holes in the 2-5 M{sub sun} mass range. On the high-mass end, we argue that the rapid decline in the inferred distribution may be the result of the particular evolutionary channel followed by low-mass X-ray binaries. This is consistent with the presence of high-mass black holes in the persistent, high-mass X-ray binary sources. If the paucity of low-mass black holes is caused by a sudden decrease of the supernova explosion energy with increasing progenitor mass, this would have observable implications for ongoing transient surveys that target core-collapse supernovae. Our results also have significant implications for the calculation of event rates from the coalescence of black hole binaries for gravitational wave detectors.

79 ASTRONOMY AND ASTROPHYSICS↗

Risk-Informed Condition Evaluation of Solar-centered Energy Generation and Distribution Networks through Bayesian Learning and Inference

We develop a methodology based on Bayesian inference over Probabilistic Graphical Models (PGMs) to understand and quantify risk in solar-centered grids using targeted measurements and learned system behavior. Being non-prescriptive but, rather, able to infer system behavior and, ultimately, address risk queries from data, our machine learning-type paradigm is tailored for diverse topologies and threat scenarios often associated with distributed energy generation and photovoltaic distributed energy resources (PV-DERs) in particular. We describe algorithmic processes for: (i) learning the structure of PGMs that result from attack-prone PV-DER-proliferated distribution systems, (ii) quantifying cause-effect relationships, and (iii) evaluating risk queries based on diverse evidence. The contributions are illustrated on a residential grid subject to output impairment attacks on its PV-DER infrastructure.

24 POWER TRANSMISSION AND DISTRIBUTION↗

An Efficient Bayesian Approach to Learning Droplet Collision Kernels: Proof of Concept Using “Cloudy,” a New n -Moment Bulk Microphysics Scheme

The small-scale microphysical processes governing the formation of precipitation particles cannot be resolved explicitly by cloud resolving and climate models. Instead, they are represented by microphysics schemes that are based on a combination of theoretical knowledge, statistical assumptions, and fitting to data (“tuning”). Historically, tuning was done in an ad hoc fashion, leading to parameter choices that are not explainable or repeatable. Recent work has treated it as an inverse problem that can be solved by Bayesian inference. The posterior distribution of the parameters given the data—the solution of Bayesian inference—is found through computationally expensive sampling methods, which require over $\mathcal{O}$(10 5 ) evaluations of the forward model; this is prohibitive for many models. We present a proof of concept of Bayesian learning applied to a new bulk microphysics scheme named “Cloudy,” using the recently developed Calibrate-Emulate-Sample (CES) algorithm. Cloudy models collision-coalescence and collisional breakup of cloud droplets with an adjustable number of prognostic moments and with easily modifiable assumptions for the cloud droplet mass distribution and the collision kernel. The CES algorithm uses machine learning tools to accelerate Bayesian inference by reducing the number of forward evaluations needed to $\mathcal{O}$(10 2 ). It also exhibits a smoothing effect when forward evaluations are polluted by noise. In a suite of perfect-model experiments, we show that CES enables computationally efficient Bayesian inference of parameters in Cloudy from noisy observations of moments of the droplet mass distribution. In an additional imperfect-model experiment, a collision kernel parameter is successfully learned from output generated by a Lagrangian particle-based microphysics model.

54 ENVIRONMENTAL SCIENCES↗

The distribution of [α/Fe] in the Milky Way disc

ABSTRACT Using a sample of red giant stars from the Apache Point Observatory Galactic Evolution Experiment (APOGEE) Data Release 16, we infer the conditional distribution $p([\alpha /{\rm Fe}]\, |\, [{\rm Fe}/{\rm H}])$ in the Milky Way disk for the α-elements Mg, O, Si, S, and Ca. In each bin of [Fe/H] and Galactocentric radius R, we model p([α/Fe]) as a sum of two Gaussians, representing ‘low-α’ and ‘high-α’ populations with scale heights $z_1=0.45\, {\rm kpc}$ and $z_2=0.95\, {\rm kpc}$, respectively. By accounting for age-dependent and z-dependent selection effects in APOGEE, we infer the [α/Fe] distributions that would be found for a fair sample of long-lived stars covering all z. Near the Solar circle, this distribution is bimodal at sub-solar [Fe/H], with the low-α and high-α peaks clearly separated by a minimum at intermediate [α/Fe]. In agreement with previous results, we find that the high-α population is more prominent at smaller R, lower [Fe/H], and larger |z|, and that the sequence separation is smaller for Si and Ca than for Mg, O, and S. We find significant intrinsic scatter in [α/Fe] at fixed [Fe/H] for both the low-α and high-α populations, typically ∼0.04-dex. The means, dispersions, and relative amplitudes of this two-Gaussian description, and the dependence of these parameters on R, [Fe/H], and α-element, provide a quantitative target for chemical evolution models and a test for hydrodynamic simulations of disk galaxy formation. We argue that explaining the observed bimodality will probably require one or more sharp transitions in the disk’s gas accretion, star formation, or outflow history in addition to radial mixing of stellar populations.

79 ASTRONOMY AND ASTROPHYSICS↗

MULTITHERMAL REPRESENTATION OF THE KAPPA-DISTRIBUTION OF SOLAR FLARE ELECTRONS AND APPLICATION TO SIMULTANEOUS X-RAY AND EUV OBSERVATIONS

Acceleration of particles and plasma heating is one of the fundamental problems in solar flare physics. An accurate determination of the spectrum of flare-energized electrons over a broad energy range is crucial for our understanding of aspects such as the acceleration mechanism and the total flare energy. Recent years have seen a growing interest in the kappa-distribution as a representation of the total spectrum of flare-accelerated electrons. In this work we present the kappa-distribution as a differential emission measure. This allows for inferring the electron distribution from X-ray observations and EUV observations by simultaneously fitting the proposed function to RHESSI and SDO/AIA data. This yields the spatially integrated electron spectra of a coronal source from less than 0.1 keV up to several tens of keV. The method is applied to a single-loop GOES C4.1 flare. The results show that the total energy can only be determined accurately by combining RHESSI and AIA observations. Simultaneously fitting the proposed representation of the kappa-distribution reduces the electron number density in the analyzed flare by a factor of ∼30 and the total flare energy by a factor of ∼5 compared with the commonly used fitting of RHESSI spectra. The spatially integrated electron spectrum of the investigated flare between 0.043 and 24 keV is consistent with the combination of a low-temperature (∼2 MK) component and a hot (∼11 MK) kappa-like component with spectral index 4, reminiscent of solar wind distributions.

79 ASTRONOMY AND ASTROPHYSICS↗

Macroinvertebrate Abundance and Biomass: 2007 Data, BPA-51; Preliminary Report, February 10, 2009..

Four Excel files containing information on the 2007 macroinvertebrate data were initially provided to Statistical Consulting Services (SCS) by EcoAnalysts on 1/27/2009. These data files contained information on abundance and biomass data at the level of taxonomic groups. The data were subsequently reformatted and compiled, and aggregated for analysis by SCS. All descriptions and analyses below relate to this compiled data. Computations were carried out separately for each site over all sample periods. Basic summary information for both the abundance and biomass data is presented in Print Out No.2. The 14 sites varied widely in their minimum, mean, maximum and variance values. The number of observations ranged from 10 to 18. Some large abundance values (abundance > 40,000) were noted for sites KR6 and KR13. A more detailed summary of each site is given in Print Out No.3. Site KR3, for example, had a mean abundance of 6914 with a sample size of 17. The variance was 4591991 and the standard error of the mean was 1643. The skewness value, a measure of symmetry for the frequency distribution, was moderately large at 1.29 indicating an asymmetric distribution. Biomass for KR3 had a mean value of 0.87 g/m{sup 2} with 17 observations. The variance was 0.8872 and the standard error was 0.228 g/m{sup 2}. Skewness for biomass was also high at 1.29. Further examination of the quantiles and frequency plots for abundance and biomass also indicate considerable skewness. The stem and leaf diagram (frequency plot) for abundance in KR3 shows most of the data centered on smaller values with a few very large counts. The distribution for biomass has a similar pattern. Statistical tests for normality are significant for both response variables in KR3, thus, the hypothesis that the data originates from a symmetric normal distribution is rejected. Because sample size estimation and statistical inference assume normally distributed data, a transformation of the data is required prior to further analysis. As was the case for previous years, the natural logarithm was chosen as a transformation to mitigate distributional skewness. Abundance and biomass for the remaining sites were also notably skewed, therefore, these data were also log transformed prior to analysis. Summary information for the transformed data (referred to as L-abun and L-bio for abundance and biomass, respectively) are given in Print Out No.4. For site KR3, the logarithmic transformation reduced skewness value for biomass to -0.66. The distributions of abundance and biomass in the other sites also generally showed improvement as well. Hence, all subsequent statistical analyses reported here will be based on the log transformed data.

13 HYDRO ENERGY↗

The Intrinsic Shapes of Low Surface Brightness Galaxies (LSBGs): A Discriminant of LSBG Galaxy Formation Mechanisms

We use the low surface brightness galaxy (LSBG) samples created from the Hyper Suprime-Cam Subaru Strategic Program (781 galaxies), the Dark Energy Survey (20977 galaxies), and the Legacy Survey (selected via H ι detection in the Arecibo Legacy Fast ALFA Survey, 188 galaxies) to infer the intrinsic shape distribution of the LSBG population. To take into account the effect of the surface brightness cuts employed when constructing LSBG samples, we simultaneously model both the projected ellipticity and the apparent surface brightness in our shape inference. We find that the LSBG samples are well characterized by oblate spheroids, with no significant difference between red and blue LSBGs. This inferred shape distribution is in good agreement with similar inferences made for ultra-diffuse cluster galaxy samples, indicating that environment does not play a key role in determining the intrinsic shape of LSBGs. In this study, we also find some evidence that LSBGs are more thickened than similarly massive high surface brightness dwarfs. We compare our results to intrinsic shape measures from contemporary cosmological simulations, and find that the observed LSBG intrinsic shapes place considerable constraints on the formation path of such galaxies. In particular, LSBG production via the migration of star formation to large radii produces intrinsic shapes in good agreement with our observational findings.

79 ASTRONOMY AND ASTROPHYSICS↗

Trapped Melt in IIIAB Irons: Solid/Liquid Elemental Partitioning During the Fractionation of the IIIAB Magma

Group IIIAB, the largest iron-meteorite group, shows compositional trends (including a three-order-of-magnitude It concentration range) indicating that it formed by fractional crystallization of a metallic magma. Because about 200 irons are available, and all degrees of crystallization are well represented, IIIAB offers an excellent set of samples for the study of crystallization at all depths of the asteroidal core. On log-log Ir-Au, and Ir-As diagrams IIIAB forms a broad band; the breadth represents real meteorite-to-meteorite variations, far outside experimental or sampling uncertainties. A successful model must explain the width of this band; I suggest that it mainly resulted from the trapping of parental magma within the crystallizing solid. Because S is essentially insoluble in metal, the abundance of FeS is a measure of the fraction of trapped liquid. The trapped-melt model is supported by the observation that irons having higher S contents plot closer to the inferred composition of the magmatic parental liquid. The lowest S values are found in the irons occupying the left envelope of the IIIAB Ir-Au or Ir-As compositional fields, thus it is this set of irons that should be interpreted as the solid products of a fractionating magma. This simplifies the modeling of the crystallization process and allows inferences regarding the distribution ratios for other elements in the evolved IIIAB system. The large (multiton) Cape York irons show wide variations in their trapped-melt fractions; their compositions seem best understood in terms of a low initial S content of the IIIAB magma, about 20 mg/g. The inferred initial IIIAB distribution coefficient for Ir, 4.6, is much higher than published values based on laboratory studies of low-S systems; I suggest that low-S (and low-P) partition-ratio measurements tend to err in the direction of unity. In IIIAB distribution coefficients for Au, As, and Ni were still < 1 when the most evolved IIIAB irons formed, another indication of a low initial S content.

Wasson, John T.↗

Estimating redshift distributions using hierarchical logistic Gaussian processes

ABSTRACT This work uses hierarchical logistic Gaussian processes to infer true redshift distributions of samples of galaxies, through their cross-correlations with spatially overlapping spectroscopic samples. We demonstrate that this method can accurately estimate these redshift distributions in a fully Bayesian manner jointly with galaxy-dark matter bias models. We forecast how systematic biases in the redshift-dependent galaxy-dark matter bias model affect redshift inference. Using published galaxy-dark matter bias measurements from the Illustris simulation, we compare these systematic biases with the statistical error budget from a forecasted weak gravitational lensing measurement. If the redshift-dependent galaxy-dark matter bias model is mis-specified, redshift inference can be biased. This can propagate into relative biases in the weak lensing convergence power spectrum on the 10–30 per cent level. We, therefore, showcase a methodology to detect these sources of error using Bayesian model selection techniques. Furthermore, we discuss the improvements that can be gained from incorporating prior information from Bayesian template fitting into the model, both in redshift prediction accuracy and in the detection of systematic modelling biases.

79 ASTRONOMY AND ASTROPHYSICS↗

Assessing the pollution risk of a groundwater source field at western Laizhou Bay under seawater intrusion

Coastal areas have great significance for human living, economy and society development in the world. With the rapid increase of pressures from human activities and climate change, the safety of groundwater resource is under the threat of seawater intrusion in coastal areas. The area of Laizhou Bay is one of the most serious seawater intruded areas in China, since seawater intrusion phenomenon was firstly recognized in the middle of 1970s. This study assessed the pollution risk of a groundwater source filed of western Laizhou Bay area by inferring the probability distribution of groundwater Cl{sup −} concentration. The numerical model of seawater intrusion process is built by using SEAWAT4. The parameter uncertainty of this model is evaluated by Markov Chain Monte Carlo (MCMC) simulation, and DREAM{sub (ZS)} is used as sampling algorithm. Then, the predictive distribution of Cl{sup -} concentration at groundwater source field is inferred by using the samples of model parameters obtained from MCMC. After that, the pollution risk of groundwater source filed is assessed by the predictive quantiles of Cl{sup -} concentration. The results of model calibration and verification demonstrate that the DREAM{sub (ZS)} based MCMC is efficient and reliable to estimate model parameters under current observation. Under the condition of 95% confidence level, the groundwater source point will not be polluted by seawater intrusion in future five years (2015–2019). In addition, the 2.5% and 97.5% predictive quantiles show that the Cl{sup −} concentration of groundwater source field always vary between 175 mg/l and 200 mg/l. - Highlights: • The parameter uncertainty of seawater intrusion model is evaluated by MCMC. • Groundwater source field won’t be polluted by seawater intrusion in future 5 years. • The pollution risk is assessed by the predictive quantiles of Cl{sup −} concentration.

54 ENVIRONMENTAL SCIENCES↗

Evaluation of a Class of Simple and Effective Uncertainty Methods for Sparse Samples of Random Variables and Functions

When very few samples of a random quantity are available from a source distribution of unknown shape, it is usually not possible to accurately infer the exact distribution from which the data samples come. Under-estimation of important quantities such as response variance and failure probabilities can result. For many engineering purposes, including design and risk analysis, we attempt to avoid under-estimation with a strategy to conservatively estimate (bound) these types of quantities -- without being overly conservative -- when only a few samples of a random quantity are available from model predictions or replicate experiments. This report examines a class of related sparse-data uncertainty representation and inference approaches that are relatively simple, inexpensive, and effective. Tradeoffs between the methods' conservatism, reliability, and risk versus number of data samples (cost) are quantified with multi-attribute metrics use d to assess method performance for conservative estimation of two representative quantities: central 95% of response; and 10 -4 probability of exceeding a response threshold in a tail of the distribution. Each method's performance is characterized with 10,000 random trials on a large number of diverse and challenging distributions. The best method and number of samples to use in a given circumstance depends on the uncertainty quantity to be estimated, the PDF character, and the desired reliability of bounding the true value. On the basis of this large data base and study, a strategy is proposed for selecting the method and number of samples for attaining reasonable credibility levels in bounding these types of quantities when sparse samples of random variables or functions are available from experiments or simulations.

97 MATHEMATICS AND COMPUTING↗

The Occurrence of Particle Size Distribution Bimodality in Midlatitude Cirrus as Inferred from Ground-Based Remote Sensing Data

To better understand the role of small particles in the microphysical processes and the radiative properties of cirrus, the reliability of historical in-situ data must be understood. Recent studies call into question the validity of that data because of shattering of large crystals on probe and aircraft surfaces thereby artificially amplifying the concentration of crystals smaller than approximately 50 μm. We contend that the general character of the in-situ measurements must be consistent, in a broad sense, with statistics derived from long-term remote sensing data. To examine this consistency, an algorithm using Doppler radar moments and Raman lidar extinction is developed to retrieve a bimodal particle size distribution and its uncertainty. Using case studies and statistics compiled over one year we show that the existence of high concentrations (> 1 cm -3 ) of small (sub 50 μm) particles in cirrus are not consistent with any reasonable interpretation of the remote sensing data. We conclude that the high concentrations of small particles found in many aircraft data sets are therefore likely an artifact of the in situ measurement process.

54 ENVIRONMENTAL SCIENCES↗

Minimally Informative Prior Distributions for PSA

A salient feature of Bayesian inference is its ability to incorporate information from a variety of sources into the inference model, via the prior distribution (hereafter simply “the prior”). However, over-reliance on old information can lead to priors that dominate new data. Some analysts seek to avoid this by trying to work with a minimally informative prior distribution. Another reason for choosing a minimally informative prior is to avoid the often-voiced criticism of subjectivity in the choice of prior. Minimally informative priors fall into two broad classes: 1) so-called noninformative priors, which attempt to be completely objective, in that the posterior distribution is determined as completely as possible by the observed data, the most well known example in this class being the Jeffreys prior, and 2) priors that are diffuse over the region where the likelihood function is nonnegligible, but that incorporate some information about the parameters being estimated, such as a mean value. In this paper, we compare four approaches in the second class, with respect to their practical implications for Bayesian inference in Probabilistic Safety Assessment (PSA). The most commonly used such prior, the so-called constrained noninformative prior, is a special case of the maximum entropy prior. This is formulated as a conjugate distribution for the most commonly encountered aleatory models in PSA, and is correspondingly mathematically convenient; however, it has a relatively light tail and this can cause the posterior mean to be overly influenced by the prior in updates with sparse data. A more informative prior that is capable, in principle, of dealing more effectively with sparse data is a mixture of conjugate priors. A particular diffuse nonconjugate prior, the logistic-normal, is shown to behave similarly for some purposes. Finally, we review the so-called robust prior. Rather than relying on the mathematical abstraction of entropy, as does the constrained noninformative prior, the robust prior places a heavy-tailed Cauchy prior on the canonical parameter of the aleatory model.

99 GENERAL AND MISCELLANEOUS↗