Search NASA⌕ Search

SEARCH · Search NASA

Results for “sampling statistics”

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 1,063 records · Page 59
All sourcesNASA NTRSDOE OSTI
NASA NTRSDate not supplied

Small-Sample Estimator Decisions – Certainly Uncertain in Human Spaceflight at NASA

Within statistics when estimating means, we rely heavily on the Central Limit Theorem (CLT) to aid in our inference, however in cases where the CLT does not apply we begin to be strongly limited in our choices. If a violation to the CLT comes in the form of small samples and unknown population variance, yet we maintain the assumption of normality, the use of t-distributions is perfectly valid. We explore cases where population normality is not assumed, where sample sizes are small, and where some actionable estimate of the central tendency of a distribution is needed such as for spaceflight operational decisions. Here we identify different scenarios based on measures taken from spaceflight, characterizing performance of estimators of central tendency within our set of simulations. We examine bias and variability of standard estimators of central tendency as they apply to varying small sample sizes and varying population distributions. Further work in this area is needed to develop a framework or set of guidelines for individuals set in these situations. Even with limited information, decisions need to be made.

Central Tendency↗
DOE OSTI2025

Investigation of oscillations above the acoustic cut-off frequency for a sample of Kepler stars. II. Identifying temporal frequency shifts

ABSTRACT Stellar oscillations with frequencies greater than the acoustic cut-off frequency are not trapped within the star’s interior. Geometric interference between these high-frequency waves produces a peak-like structure in the power spectrum of the star, known as the pseudo-modes. In this study, we aim to analyse 51 stars previously identified (from a sample of 91 Kepler stars) to contain statistically significant evidence for pseudo-modes, to determine if the pseudo-mode frequencies vary in time. We obtained these frequency variations and the uncertainty by the resampled periodogram approach, where periodogram realisations were created from successive, overlapping 90-d time segments, and a cross-correlation function was generated between them. The computed shifts were compared against temporal p-mode frequency shifts and a stellar magnetic activity proxy, $S_{\mathrm{ph}}$. We found that for 16 per cent of the stars pseudo-mode frequency shifts were significantly anticorrelated with p-mode shifts, as is the case for the Sun. However, we also found pseudo-mode and p-mode shifts to be significantly correlated in-phase in 8 per cent of our sample. The remainder showed no significant correlation. We also searched for trends between the maximum variation of pseudo-mode frequency shifts and stellar parameters. Whilst we found no correlation between the maximum variation and stellar photospheric and chromospheric proxies, we found frequency variations to have a greater amplitude for cooler and slower rotating stars, which is in opposition to p-mode frequency shift variation. Temporal changes to pseudo-mode frequencies may be attributable to variations in stellar magnetic activity, and therefore can contribute to our understanding of stellar activity cycles.

Millson, L. J. (ORCID:0009000342542676)↗
DOE OSTI2018

Estimating Highway Volumes Using Vehicle Probe Data - Proof of Concept: Preprint

This paper examines the feasibility of using sampled commercial probe data in combination with validated continuous counter data to accurately estimate vehicle volume across the entire roadway network, for any hour during the year. Currently either real time or archived volume data for roadways at specific times are extremely sparse. Most volume data are average annual daily traffic (AADT) measures derived from the Highway Performance Monitoring System (HPMS). Although methods to factor the AADT to hourly averages for typical day of week exist, actual volume data is limited to a sparse collection of locations in which volumes are continuously recorded. This paper explores the use of commercial probe data to generate accurate volume measures that span the highway network providing ubiquitous coverage in space, and specific point-in-time measures for a specific date and time. The paper examines the need for the data, fundamental accuracy limitations based on a basic statistical model that take into account the sampling nature of probe data, and early results from a proof of concept exercise revealing the potential of probe type data calibrated with public continuous count data to meet end user expectations in terms of accuracy of volume estimates.

25 ENERGY STORAGE↗
NASA NTRS2010

Mid-IR Properties of an Unbiased AGN Sample of the Local Universe: Emission-Line Diagnostics - 1

\Ve compare mid-IR emission-lines properties, from high-resolution Spitzer IRS spectra of a statistically-complete hard X-ray (14-195 keV) selected sample of nearby (z < 0.05) AGN detected by the Burst Alert Telescope (BAT) aboard Swift. The luminosity distribution for the mid-infrared emission-lines, [O IV] 25.89 microns, [Ne II] 12.81 microns, [Ne III] 15.56 microns and [Ne V] 14.32 microns, and hard X-ray continuum show no differences between Seyfert 1 and Seyfert 2 populations, although six newly discovered BAT AGNs are shown to be under-luminous in [O IV], most likely the result of dust extinction in the host galaxy. The overall tightness of the mid-infrared correlations and BAT luminosities suggests that the emission lines primarily arise in gas ionized by the AGN. We also compared the mid-IR emission-lines in the BAT AGNs with those from published studies of star-forming galaxies and LINERs. We found that the BAT AGN fall into a distinctive region when comparing the [Ne III]/[Ne II] and the [O IV]/[Ne III] quantities. From this we found that sources that have been previously classified in the mid-infrared/optical as AGN have smaller emission line ratios than those found for the BAT AGNs, suggesting that, in our X-ray selected sample, the AGN represents the main contribution to the observed line emission. Overall, we present a different set of emission line diagnostics to distinguish between AGN and star forming galaxies that can be used as a tool to find new AGN.

Weaver, K. A.↗
NASA NTRS1991

Far-infrared properties of Markarian galaxies with multiple nuclei - Warm dust emission in mergers

An investigation of coadded IRAS data is performed on 187 Markarian galaxies where distinguishing morphological characteristics or multiple optical nuclei are present. The far-IR properties of Markarian galaxies are compared to the IRAS Bright Galaxy Sample, and a much higher median dust temperature is found in the multiple nucleus galaxies, suggesting that more far-IR luminosity results from active star formation. Both optical/UV and far-IR selection techniques are necessary to extract complete samples of AGNs since the far-IR two-color plane can miss up to 50 percent of the galaxies. A systematic increase in the contribution of warm dust emission due to active star formation and AGNs is found in a statistical comparison of merger candidates and other galaxy samples. The assumed nature of precursor galaxies determines the assumed enhancement of far-IR luminosity caused by galaxy collisions. A model is presented which describes the properties of the Markarian galaxies in terms of enhanced OB star formation and different grain size distributions. The results of the investigation are shown to be consistent with a 'subdued' interpretation of merging galaxies with high luminosities.

Mazzarella, Joseph M.↗
NASA NTRS1995

Designing for Small Volume Assembly of Advanced Electronics Packages

We describe a general methodology to Design for Producibility and Reliability (DFPAR) for very small volume production runs. In cases where the entire volume for fabrication is less than five products, traditional Statistical Process Control (SPC) is inadequate due to reliance on statistics of much larger volumes and the Central Limit Theorem. Data acquisition for process parameter estimation from such a small sample size is difficult; however, it is critical to producing high reliability product.

Central Limit Theorem statistics high reliability ↗
DOE OSTI2018

Comparison of experimental methods for estimating matrix diffusion coefficients for contaminant transport modeling

Here, diffusion cell and diffusion wafer experiments were conducted to compare methods for estimating effective matrix diffusion coefficients in rock core samples from Pahute Mesa at the Nevada Nuclear Security Site (NNSS). A diffusion wafer method, in which a solute diffuses out of a rock matrix that is pre-saturated with water containing the solute, is presented as a simpler alternative to the traditional through-diffusion (diffusion cell) method. Both methods yielded estimates of effective matrix diffusion coefficients that were within the range of values previously reported for NNSS volcanic rocks. The difference between the estimates of the two methods ranged from 14 to 30%, and there was no systematic high or low bias of one method relative to the other. From a transport modeling perspective, these differences are relatively minor when one considers that other variables (e.g., fracture apertures, fracture spacings) influence matrix diffusion to a greater degree and tend to have greater uncertainty than effective matrix diffusion coefficients. For the same relative random errors in concentration measurements, the diffusion cell method yields effective matrix diffusion coefficient estimates that have less uncertainty than the wafer method. However, the wafer method is easier and less costly to implement and yields estimates more quickly, thus allowing a greater number of samples to be analyzed for the same cost and time. Given the relatively good agreement between the methods, and the lack of any apparent bias between the methods, the diffusion wafer method appears to offer advantages over the diffusion cell method if better statistical representation of a given set of rock samples is desired.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2017

Comparison of Experimental Methods for Estimating Matrix Diffusion Coefficients for Contaminant Transport Modeling

Diffusion cell and diffusion wafer experiments were conducted to compare methods for estimating matrix diffusion coefficients in rock core samples from Pahute Mesa at the Nevada Nuclear Security Site (NNSS). A diffusion wafer method, in which a solute diffuses out of a rock matrix that is pre-saturated with water containing the solute, is presented as a simpler alternative to the traditional through-diffusion (diffusion cell) method. Both methods yielded estimates of matrix diffusion coefficients that were within the range of values previously reported for NNSS volcanic rocks. The difference between the estimates of the two methods ranged from 14 to 30%, and there was no systematic high or low bias of one method relative to the other. From a transport modeling perspective, these differences are relatively minor when one considers that other variables (e.g., fracture apertures, fracture spacings) influence matrix diffusion to a greater degree and tend to have greater uncertainty than diffusion coefficients. For the same relative random errors in concentration measurements, the diffusion cell method yields diffusion coefficient estimates that have less uncertainty than the wafer method. However, the wafer method is easier and less costly to implement and yields estimates more quickly, thus allowing a greater number of samples to be analyzed for the same cost and time. Given the relatively good agreement between the methods, and the lack of any apparent bias between the methods, the diffusion wafer method appears to offer advantages over the diffusion cell method if better statistical representation of a given set of rock samples is desired.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
DOE OSTI2018

Identifying structural changes with unsupervised machine learning methods

In this work, unsupervised machine learning methods are used to identify structural changes using the melting point transition in classical molecular dynamics simulations as an example application of the approach. Dimensionality reduction and clustering methods are applied to instantaneous radial distributions of atomic configurations from classical molecular dynamics simulations of metallic systems over a large temperature range. Principal component analysis is used to dramatically reduce the dimensionality of the feature space across the samples using an orthogonal linear transformation that preserves the statistical variance of the data under the condition that the new feature space is linearly independent. From there, k-means clustering is used to partition the samples into solid and liquid phases through a criterion motivated by the geometry of the reduced feature space of the samples, allowing for an estimation of the melting point transition. This pattern criterion is conceptually similar to how humans interpret the data but with far greater throughput, as the shapes of the radial distributions are different for each phase and easily distinguishable by humans. The transition temperature estimates derived from this machine learning approach produce comparable results to other methods on similarly small system sizes. These results show that machine learning approaches can be applied to structural changes in physical systems.

42 ENGINEERING↗
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↗
NASA NTRS2011

Shuttle Hypervelocity Impact Database

With three missions outstanding, the Shuttle Hypervelocity Impact Database has nearly 3000 entries. The data is divided into tables for crew module windows, payload bay door radiators and thermal protection system regions, with window impacts compromising just over half the records. In general, the database provides dimensions of hypervelocity impact damage, a component level location (i.e., window number or radiator panel number) and the orbiter mission when the impact occurred. Additional detail on the type of particle that produced the damage site is provided when sampling data and definitive analysis results are available. Details and insights on the contents of the database including examples of descriptive statistics will be provided. Post flight impact damage inspection and sampling techniques that were employed during the different observation campaigns will also be discussed. Potential enhancements to the database structure and availability of the data for other researchers will be addressed in the Future Work section. A related database of returned surfaces from the International Space Station will also be introduced.

Hyde, James L.↗
DOE OSTI2023

Using elliptical galaxy kinematics to compare the strength of gravity in cosmological regions of differing gravitational potential – a first look

ABSTRACT Various models of modified gravity invoke ‘screening’ mechanisms that are sensitive to the value of the local gravitational potential. This could have observable consequences for galaxies. These consequences might be seen by comparing two proxies for galaxy mass – their luminosity and their internal kinematics – as a function of local galaxy density. Motivated by this prospect, we have compared the observed properties of luminous red galaxies (LRGs) inside and outside of voids in the cosmic large scale structure. We used archival measurements of line widths, luminosities, redshifts, colours, and positions of galaxies in conjunction with recent void catalogues to construct comparison LRG samples inside and outside of voids. We fitted these two samples to the well-established fundamental plane of elliptical galaxies to constrain any differences between the inferred value of the Newtonian gravitational constant G for the two samples. We obtained a null result, with an upper limit on any fractional difference in G within and outside of cosmological voids to be α = δG/G ∼ 40 per cent. This upper bound is dominated by the small-number statistics of our N ∼ 100 within-void LRG sample. With the caveat that environmental effects could influence various parameters such as star formation, we estimate that a 1 per cent statistical limit on α could be attained with data from 105 elliptical galaxies within voids. This is within the reach of future photometric and spectroscopic surveys, both of which are required to pursue this method.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Assessment of Bulk Oxygen Capacity and Transient Redox Behavior of Foamed Lanthanum Strontium Manganese Perovskites

Synthesis of solar thermochemical hydrogen (STCH) production redox materials with engineered structures, for example, replica foams, can enable efficient heat and mass transport and are critical for scaled-up systems. Prior work has motivated the use of lanthanum strontium manganese (LSM)-type perovskites as foamed STCH materials, but the effect of their morphology on bulk and kinetic behavior has not been reported. In this work, replica and direct foamed samples of La 0.65 Sr 0.35 MnO 3-δ (LSM35) were fabricated and compared to synthesized powders and dense monoliths, and similarly synthesized CeO 2-δ (ceria) foams, regarding their specific reaction rates and bulk oxygen capacity/H 2 yields. Changes in oxygen capacity (Δδ) and reaction rates were measured between 1200 °C and 1400 °C by using fixed ratios of steam and hydrogen during both reduction and oxidation steps, allowing for analysis under practical high conversion conditions. Results suggest bulk behavior and reaction rates of the foamed LSM materials are comparable to their powder analogues. Differences in reaction rates were observed only when replica foamed samples were subjected to rapid laser heating (emulating conditions expected in solar furnaces), which is expected but has not been demonstrated at such a small scale. Here, foamed samples were further subjected to 50 redox cycles at 1400 °C to evaluate their stability. Results show no statistically significant decrease in hydrogen production for any of the foamed samples, but the direct foamed samples became brittle with time. Together, these results demonstrate the viability of replica foamed LSM perovskites for integration in scaled-up STCH systems.

08 HYDROGEN↗
DOE OSTI2023

The Spatial Clustering of ROSAT All-Sky Survey Active Galactic Nuclei. V. The Evolution of Broad-line AGN Clustering Properties in the Last 6 Gyr

This is the fifth paper in a series of investigations of the clustering properties of luminous, broad-emission-line active galactic nuclei (AGNs) identified in the ROSAT All-Sky Survey (RASS) and Sloan Digital Sky Survey (SDSS). In this work we measure the cross-correlation function (CCF) between RASS/SDSS Data Release 14 AGNs with the SDSS CMASS galaxy sample at 0.44 < z < 0.64. We apply halo occupation distribution (HOD) modeling to the CCF along with the autocorrelation function of the CMASS galaxies. We find that X-ray-selected and optically selected AGNs at 0.44 < z < 0.64 reside in statistically identical halos with a typical dark matter halo (DMH) mass of $M$$^{typ,AGN}_{DMH}$ ~ 10 12.7 h –1 M ⊙ . The acceptable HOD parameter space for these two broad-line AGN samples have only statistically marginal differences caused by small deviations of the CCFs in the one-halo-dominated regime on small scales. In contrast to optically selected AGNs, the X-ray AGN sample may contain a larger population of satellites at MDMH ~ 1013h–1M⊙. We compare our measurements in this work with our earlier studies at lower independent redshift ranges, spanning a lookback time of 6 Gyr. The comparison over this wider redshift range of 0.07 < z < 0.64 reveals (i) no significant difference between the typical DMH masses of X-ray-selected and optically selected AGNs, (ii) weak positive clustering dependencies of $M$$^{typ,AGN}_{DMH}$ with L X and M BH , (iii) no significant dependence of $M$$^{typ,AGN}_{DMH}$ on Eddington ratio, and (iv) the same DMH masses host more-massive accreting black holes at high redshift than at low redshifts.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

Measuring infrared surface brightness fluctuation distances with HST WFC3: Calibration and advice

We present new calibrations of the near-infrared (near-IR) surface brightness fluctuation (SBF) distance method for the F110W (J{sub 110}) and F160W (H{sub 160}) bandpasses of the Wide Field Camera 3 Infrared Channel (WFC3/IR) on the Hubble Space Telescope. The calibrations are based on data for 16 early-type galaxies in the Virgo and Fornax clusters observed with WFC3/IR and are provided as functions of both the optical (g{sub 475}−z{sub 850}) and near-infrared (J{sub 110}−H{sub 160}) colors. The scatter about the linear calibration relations for the luminous red galaxies in the sample is approximately 0.10 mag, corresponding to a statistical error of 5% in distance. Our results imply that the distance to any suitably bright elliptical galaxy can be measured with this precision out to about 80 Mpc in a single-orbit observation with WFC3/IR, making this a remarkably powerful instrument for extragalactic distances. The calibration sample also includes much bluer and lower-luminosity galaxies than previously used for IR SBF studies, revealing interesting population differences that cause the calibration scatter to increase for dwarf galaxies. Comparisons with single-burst population models show that as expected, the redder early-type galaxies contain old, metal-rich populations, while the bluer dwarf ellipticals contain a wider range of ages and lower metallicities than their more massive counterparts. Radial SBF gradients reveal that IR color gradients are largely an age effect; the bluer dwarfs typically have their youngest populations near their centers, while the redder giant ellipticals show only weak trends and in the opposite sense. Because of the population variations among bluer galaxies, distance measurements in the near-IR are best limited to red early-type galaxies. We conclude with some practical guidelines for using WFC3/IR to measure reliable SBF distances.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2018

Alpha Air Sample Counting Efficiency Versus Dust Loading: Evaluation of a Large Data Set

Dust loading on air sample filters is known to cause a loss of efficiency for direct counting of alpha activity on the filters, but the amount of dust loading and the correction factor needed to account for attenuated alpha particles is difficult to assess. In this paper, correction factors are developed by statistical analysis of a large database of air sample results for a uranium and plutonium processing facility at the Savannah River Site. As is typically the case, dust-loading data is not directly available, but sample volume is found to be a reasonable proxy measure; the amount of dust loading is inferred by a combination of the derived correction factors and a Monte Carlo model. The technique compares the distribution of activity ratios [beta/(beta + alpha)] by volume and applies a range of correction factors on the raw alpha count rate. The best-fit results with this method are compared with MCNP modeling of activity uniformly deposited in the dust and analytical laboratory results of digested filters. Finally, a linear fit is proposed to evenly-deposited alpha activity collected on filters with dust loading over a range of about 2 mg cm -2 to 1,000 mg cm -2 .

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS2004

Prediction of Active-Region CME Productivity from Magnetograms

We report results of an expanded evaluation of whole-active-region magnetic measures as predictors of active-region coronal mass ejection (CME) productivity. Previously, in a sample of 17 vector magnetograms of 12 bipolar active regions observed by the Marshall Space Flight Center (MSFC) vector magnetograph, from each magnetogram we extracted a measure of the size of the active region (the active region s total magnetic flux a) and four measures of the nonpotentiality of the active region: the strong-shear length L(sub SS), the strong-gradient length L(sub SG), the net vertical electric current I(sub N), and the net-current magnetic twist parameter alpha (sub IN). This sample size allowed us to show that each of the four nonpotentiality measures was statistically significantly correlated with active-region CME productivity in time windows of a few days centered on the day of the magnetogram. We have now added a fifth measure of active-region nonpotentiality (the best-constant-alpha magnetic twist parameter (alpha sub BC)), and have expanded the sample to 36 MSFC vector magnetograms of 31 bipolar active regions. This larger sample allows us to demonstrate statistically significant correlations of each of the five nonpotentiality measures with future CME productivity, in time windows of a few days starting from the day of the magnetogram. The two magnetic twist parameters (alpha (sub 1N) and alpha (sub BC)) are normalized measures of an active region s nonpotentially in that they do not depend directly on the size of the active region, while the other three nonpotentiality measures (L(sub SS), L(sub SG), and I(sub N)) are non-normalized measures in that they do depend directly on active-region size. We find (1) Each of the five nonpotentiality measures is statistically significantly correlated (correlation confidence level greater than 95%) with future CME productivity and has a CME prediction success rate of approximately 80%. (2) None of the nonpotentiality measures is a significantly better CME predictor than the others. (3) The active-region phi shows some correlation with CME productivity, but well below a statistically significant level (correlation confidence level less than approximately 80%; CME prediction success rate less than approximately 65%). (4) In addition to depending on magnetic twist, CME productivity appears to have some direct dependence on active-region size (rather than only an indirect dependence through a correlation of magnetic twist with active-region size), but it will take a still larger sample of active regions (50 or more) to certify this. (5) Of the five nonpotentiality measures, L(sub SG) appears to be the best for operational CME forecasting because it is as good or better a CME predictor than the others and it alone does not require a vector magnetogram; L(sub SG) can be measured from a line-of-sight magnetogram such as from the Michelson Doppler Imager (MDI) on the Solar and Heliospheric Observatory (SOHO).

Falconer, D. A.↗
DOE OSTI2021

Evaluating the Benefits of Bayesian Hierarchical Methods for Analyzing Heterogeneous Environmental Datasets: A Case Study of Marine Organic Carbon Fluxes

Large compilations of heterogeneous environmental observations are increasingly available as public databases, allowing researchers to test hypotheses across datasets. Statistical complexities arise when analyzing compiled data due to unbalanced spatial sampling, variable environmental context, mixed measurement techniques, and other reasons. Hierarchical Bayesian modeling is increasingly used in environmental science to describe these complexities, however few studies explicitly compare the utility of hierarchical Bayesian models to simpler and more commonly applied methods. Here we demonstrate the utility of the hierarchical Bayesian approach with application to a large compiled environmental dataset consisting of 5,741 marine vertical organic carbon flux observations from 407 sampling locations spanning eight biomes across the global ocean. We fit a global scale Bayesian hierarchical model that describes the vertical profile of organic carbon flux with depth. Profile parameters within a particular biome are assumed to share a common deviation from the global mean profile. Individual station-level parameters are then modeled as deviations from the common biome-level profile. The hierarchical approach is shown to have several benefits over simpler and more common data aggregation methods. First, the hierarchical approach avoids statistical complexities introduced due to unbalanced sampling and allows for flexible incorporation of spatial heterogeneitites in model parameters. Second, the hierarchical approach uses the whole dataset simultaneously to fit the model parameters which shares information across datasets and reduces the uncertainty up to 95% in individual profiles. Third, the Bayesian approach incorporates prior scientific information about model parameters; for example, the non-negativity of chemical concentrations or mass-balance, which we apply here. We explicitly quantify each of these properties in turn. We emphasize the generality of the hierarchical Bayesian approach for diverse environmental applications and its increasing feasibility for large datasets due to recent developments in Markov Chain Monte Carlo algorithms and easy-to-use high-level software implementations.

54 ENVIRONMENTAL SCIENCES↗