Search NASA⌕ Search

SEARCH · Search NASA

Results for “Covariance operator estimation”

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.

49 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

HighDimMixedModels.jl: Robust high-dimensional mixed-effects models across omics data

High-dimensional mixed-effects models are an increasingly important form of regression in which the number of covariates rivals or exceeds the number of samples, which are collected in groups or clusters. The penalized likelihood approach to fitting these models relies on a coordinate descent algorithm that lacks guarantees of convergence to a global optimum. Here, we empirically study the behavior of this algorithm on simulated and real examples of three types of data that are common in modern biology: transcriptome, genome-wide association, and microbiome data. Our simulations provide new insights into the algorithm’s behavior in these settings, and, comparing the performance of two popular penalties, we demonstrate that the smoothly clipped absolute deviation (SCAD) penalty consistently outperforms the least absolute shrinkage and selection operator (LASSO) penalty in terms of both variable selection and estimation accuracy across omics data. To empower researchers in biology and other fields to fit models with the SCAD penalty, we implement the algorithm in a Julia package, HighDimMixedModels.jl .

Gorstein, Evan↗
DOE OSTI2023

Smoothing Lexis diagrams using kernel functions: A contemporary approach

Lexis diagrams are rectangular arrays of event rates indexed by age and period. Analysis of Lexis diagrams is a cornerstone of cancer surveillance research. Typically, population-based descriptive studies analyze multiple Lexis diagrams defined by sex, tumor characteristics, race/ethnicity, geographic region, etc. Inevitably the amount of information per Lexis diminishes with increasing stratification. Several methods have been proposed to smooth observed Lexis diagrams up front to clarify salient patterns and improve summary estimates of averages, gradients, and trends. In this article, we develop a novel bivariate kernel-based smoother that incorporates two key innovations. First, for any given kernel, we calculate its singular values decomposition, and select an optimal truncation point—the number of leading singular vectors to retain—based on the bias-corrected Akaike information criterion. Second, we model-average over a panel of candidate kernels with diverse shapes and bandwidths. The truncated model averaging approach is fast, automatic, has excellent performance, and provides a variance-covariance matrix that takes model selection into account. We present an in-depth case study (invasive estrogen receptor-negative breast cancer incidence among non-Hispanic white women in the United States) and simulate operating characteristics for 20 representative cancers. The truncated model averaging approach consistently outperforms any fixed kernel. Our results support the routine use of the truncated model averaging approach in descriptive studies of cancer.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

A Bayesian nonparametric analysis for zero-inflated multivariate count data with application to microbiome study

High-throughput sequencing technology has enabled researchers to profile microbial communities from a variety of environments, but analysis of multivariate taxon count data remains challenging. Here, we develop a Bayesian nonparametric (BNP) regression model with zero inflation to analyse multivariate count data from microbiome studies. A BNP approach flexibly models microbial associations with covariates, such as environmental factors and clinical characteristics. The model produces estimates for probability distributions which relate microbial diversity and differential abundance to covariates, and facilitates community comparisons beyond those provided by simple statistical tests. We compare the model to simpler models and popular alternatives in simulation studies, showing, in addition to these additional community-level insights, it yields superior parameter estimates and model fit in various settings. The model's utility is demonstrated by applying it to a chronic wound microbiome data set and a Human Microbiome Project data set, where it is used to compare microbial communities present in different environments.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Physics-Informed Gaussian Process Regression for States Estimation and Forecasting in Power Grids

Real-time state estimation and forecasting are critical for the efficient operation of power grids. In this paper, a physics-informed Gaussian process regression (PhI-GPR) method is presented and used for forecasting and estimating the phase angle, angular speed, and wind mechanical power of a three-generator power grid system using sparse measurements. In standard data-driven Gaussian process regression (GPR), parameterized models for the prior statistics are fit by maximizing the marginal likelihood of observed data. In the PhI-GPR method, we propose to compute the prior statistics offline by solving stochastic differential equations (SDEs) governing the power grid dynamics. The short-term forecast of a power grid system dominated by wind generation is complicated by the stochastic nature of the wind and the resulting uncertainty in wind mechanical power. Here, we assume that the power grid dynamics are governed by swing equations, with the wind mechanical power fluctuating randomly in time. We solve these equations for the mean and covariances of the power grid states using the Monte Carlo simulation method. We demonstrate that the proposed PhI-GPR method can accurately forecast and estimate observed and unobserved states. For the considered problem, PhI-GPR has computational advantages over the ensemble Kalman filter (EnKF) method: In PhI-GPR, ensembles are computed offline and independently of the data acquisition process, whereas for EnFK, ensembles are computed online with data acquisition, rendering real-time forecast more challenging. We also demonstrate that the PhI-GPR forecast is more accurate than the EnKF forecast when the random mechanical wind power is non-Markovian. In contrast, the two methods produce similar forecasts for the Markovian mechanical wind power. For observed states, we show that PhI-GPR provides a forecast comparable to the standard data-driven GPR; both forecasts are significantly more accurate than the autoregressive integrated moving average (ARIMA) forecast. We also show that the ARIMA forecast is more sensitive to observation frequency and measurement errors than the PhI-GPR forecast.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

JGI-Trichoderma v1.0

There is a series of Python and bash scripts to parse genomics datasets used to evaluate the coevolution of gene families and the feature importance of gene families using an SVM classifier. - Cover analysis: takes a list of single-copy genes in a set of genomes, aligns and builds the gene trees to determine if two gene families have a signature of covariation with one another. It parses the files to run phykit cover script described here: https://jlsteenwyk.com/PhyKIT/usage/index.html - SVM-classifier: This Python script is an SVM-based genomic classifier designed for biological data analysis. It combines machine learning with feature selection to identify important genomic markers and classify biological samples. Core Functionality: The script uses Support Vector Machines from scikit-learn to classify genomic data, incorporating SelectKBest for automated feature selection and leave-one-out cross-validation for performance assessment. It operates in multiple modes: feature ranking, optimal combination discovery, and sample prediction. Primary Applications: Genomic sample classification and biomarker discovery Feature importance analysis in high-dimensional biological datasets Prediction of sample categories based on genomic profiles Research applications requiring robust classification of biological data Key Advantages: High-dimensional handling: SVMs excel with genomic data's typical high feature-to-sample ratios Integrated feature selection: Reduces noise and computational overhead while identifying key markers Probability estimation: Provides confidence scores essential for biological interpretation Validation robustness: Leave-one-out cross-validation ensures reliable performance metrics Operational flexibility: Multiple analysis modes support different research phases from exploration to prediction

Stecca Steindorff, Andrei [Lawrence Berkeley Natio↗
DOE OSTI2022

A Variational Method for Analyzing Vortex Flows in Radar-Scanned Tornadic Mesocyclones. Part III: Sensitivities to Vortex Center Location Errors

Abstract When the vortex center location is estimated from a radar-scanned tornadic mesocyclone, the estimated location is not error-free. This raises an important issue concerning the sensitivities of analyzed vortex flow (VF) fields by the VF-Var (formulated in Part I of this paper series and tested in Part II) to vortex center location errors, denoted by Δ x c . Numerical experiments are performed to address this issue with the following findings: The increase of |Δ x c | from zero to a half of vortex core radius causes large analysis error increases in the vortex core but the increased analysis errors decrease rapidly away from the vortex core especially for dual-Doppler analyses. The increased horizontal-velocity errors in the vortex core are mainly in the Δ x c -normal component, because this component varies much more rapidly than the other component along the Δ x c direction in the vortex core. The vertical variations of Δ x c distort the vertical correlation structure of Δ x c -dislocated VF-dependent background error covariance, which can increase the analysis errors in the vortex core. The dual-Doppler analyses have adequate accuracies outside the vortex core even when |Δ x c | increases to a half of vortex core radius, while single-Doppler analyses can also have adequate accuracies outside the vortex core mainly for the single-Doppler-observed velocity component. The sensitivities to Δ x c are largely unaffected by the vortex slanting. The above findings are important and useful for assessing the accuracies of analyzed VFs for real radar-observed tornadic mesocyclones. Significance Statement When the vortex center location is estimated from a radar-scanned tornadic mesocyclone, the estimated location is not error-free. This raises an issue concerning the sensitivity of analyzed vortex flow (VF) by the VF-Var (formulated in Part I of this paper series and tested with simulated radar observations in Part II) to vortex center location error. This issue and its required investigations are very important for the VF-Var to be applied to real radar-observed tornadic mesocyclones, especially in an operational setting with the WSR-88Ds. Numerical experiments are performed to address this issue. The findings from these experiments are important and useful for assessing the accuracies of VF-Var analyzed VF fields for real radar-observed tornadic mesocyclones.

Xu, Qin↗
DOE OSTI2024

Proceedings for the Workshop on Applied Nuclear Data Activities 2024

The Workshop for Applied Nuclear Data Activities (WANDA) is designed to increase communication among nuclear data (ND) users in multidisciplinary federal programs, ND producers, ND funders, and other ND experts. It also presents an opportunity to cross-pollinate ideas as well as introduce ND gaps identified by federal programs to ND experts and ND capabilities to the various federal ND users. WANDA 2024 included five technical sessions, three of which focused on Fusion Energy Sciences (FES)—FES Fusion Neutronics, FES Tritium Production, and FES Material Damage—and two stand-alone sessions—Isotopes and Targetry for Nuclear Data and Uncertainty Quantification. The FES sessions successfully brought new voices to the WANDA discussions, expanding the application space in which nuclear data are critical. FES programs need accurate nuclear data with realistic uncertainty quantification to properly estimate, for example, shielding, activation, tritium production, helium production, structural material integrity, and superconducting magnet operation. This includes a variety of projectile (neutrons, photons, charged particles) and target atoms. One of the action items common to all the FES sessions was a need to perform sensitivity studies to identify the prioritization of nuclear data needs. The Isotopes and Targetry session highlighted the many capabilities available to produce high-quality targets for nuclear data measurements, including 3D printing with spherical powders, combustion synthesis coupled with spin coating & electrospraying, inkjet printing, and isotopic doping. These new methods open doors for more accurate measurement, but it was also stressed that sample characterization following any method of fabrication is of the highest importance to accurately interpret nuclear data measurement results that used that sample. The Uncertainty Quantification (UQ) session was broken into two categories: nuclear data uncertainty quantification and the use of that uncertainty quantification. Thematic to the UQ session was the loss of information when going from nuclear data measurement, to evaluation, to evaluated file, and finally to neutron transport calculations. Current evaluated ND libraries typically only contain covariances, which assume that the probability distributions are Gaussian. Beyond being a simplified assumption for many evaluations, this can lead to negative values on many observables when attempting to sample the covariance. The covariance format, however, is very efficient in that a simple set of linear equations can transform uncertainty from parameters or cross sections to the application of interest. Focused collaboration is needed between nuclear data evaluators and nuclear data users to ensure that needs are being met.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

The NuMI Flux Prediction at ICARUS

The Deep Underground Neutrino Experiment (DUNE) is a next-generation long-baseline neutrino oscillation experiment seeking to probe fundamental symmetries within the structure of the Pontecorvo-Maki-Nakagawa-Sakata (PMNS) mixing matrix, and perform precision measurements its parameters including the neutrino mass ordering via the sign of $\Delta m^2_{31}$, and the charge-parity violating phase, $\delta_{CP}$. To make these measurements with high precision, DUNE will require external $\nu$-Ar scattering cross section data as a crucial input to the oscillation fit. Imaging Cosmic And Rare Underground Signals (ICARUS) is a 476 t liquid argon neutrino detector located at Fermi National Accelerator Laboratory (FNAL) where it is serving as the far detector for the Short-Baseline Neutrino (SBN) program along the Booster Neutrino Beam (BNB) axis. ICARUS additionally lies 795 m downstream and 100.1 mrad off-axis of the Neutrinos at the Main Injector (NuMI) neutrino beam. From this position, ICARUS is exposed to a large flux of NuMI (anti-)electron and (anti-)muon neutrinos, and poses a unique opportunity to provide high-statistics measurements of quasi-elastic and single pion-production cross sections for four neutrino flavors ($\nu_{\mu}$, $\nu_{e}$, $\bar{\nu}_{\mu}$, $\bar{\nu}_{e}$). This dissertation is centered around accurately characterizing the models and estimating their precision for use in making these measurements. This includes identifying major sources of uncertainty in the models such that they can be properly propagated to the cross section measurements. Specifically, this work focused on the model of the NuMI beamline and its impact on the neutrino fluxes, but also delved into the detector response model and its impact on reconstructed observables in the detector. Significant efforts were made to improve the characterization to enhance precision and thus reduce the level of propagated uncertainty. In particular, the NuMI flux was determined to be composed of 57% $\nu_{\mu}$, 38% $\bar{\nu}_{\mu}$, 3% $\nu_{e}$, and 2% $\bar{\nu}_{e}$ while the horns are operating in the positive-particle focusing configuration. The total uncertainty on the $\nu_{\mu} + \bar{\nu}_{\mu}$ ($\nu_{e} + \bar{\nu}_{e}$) flux while operating in the forward horn operating mode was determined to be 10.84% (9.04%). Compared to the on-axis flux, mesons that eventually decay to neutrinos more frequently reinteract within the NuMI structure, resulting in elevated uncertainty as these processes are not well-constrained by existing hadron interaction cross section measurements. Covariance matrices were calculated to propagate the flux uncertainty characterization to NuMI analyses.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

Stereo Camera Deployment in Support of TRACER (Field Campaign Report)

An improved understanding of the salient environmental controls on cloud formation, evolution, and eventual dissipation is critical to address ongoing challenges with cloud process and parameterization representations in global climate and Earth system models. One goal for the recent U.S. Department of Energy Atmospheric Radiation Measurement (ARM) user facility’s Tracking Aerosol Convection Interactions Experiment (TRACER) campaign was to collect a comprehensive data set that enabled such convective cloud process studies and key demonstrations for those controls that influenced cloud life cycle (i.e., aerosols, thermodynamics) in the Houston, Texas region. To help accomplish this, ARM instrumentation during TRACER was tasked with tracking a large number of individual, isolated convective cells – and to follow the evolution of these cells at high spatiotemporal resolution to characterize changes in cloud dynamic and microphysical properties. Since Houston experiences a range of convective clouds, it was known that the standard ARM Mobile Facility (AMF) instruments may not be sufficient to completely document initiating, transient, or dissipating low- or shallow-cloud behaviors that were also expected during this campaign (in terms of sensitivity, resolution, and/or operational availability). As one partial solution, a supplemental stereo camera deployment (this sub-campaign) was requested to augment the ARM AMF instrumentation to better address shallow and shallow-to-deep transitional types of cloud process drivers during TRACER (ARM stereo cameras for clouds [STEREOCAM]; Romps and Öktem 2018). The primary scientific focus was the relationships between cloud properties and the ambient conditions, which points to several key TRACER science questions including: ‘What is the relationship between cloud size or updraft intensity to the environmental wind shear and/or humidity?’ Overall, the ARM TRACER stereo camera deployment demonstrated unique effectiveness in observing a wide range of critical shallow, congestus, and transitioning or time-evolving cloud characteristics. The data sets from these cameras include information on the clouds' horizontal dimensions, elevations, and depths, while also enabling potential products for cloud initiation and dissipation rates, and vertical velocities. Stereo cameras simultaneously inform on cloud life cycle stage and spatial properties such as cloud fractional coverage, which should provide complementary information for ARM users when combined with TRACER cloud radars, lidar, and/or other profiling sensors. Moreover, camera products offer large-eddy simulation (LES),-scale-appropriate cloud coverage, depth, and spatial variability estimates, while opening additional avenues to challenge difficult process questions on cloud updrafts/entrainment and their covariability with environmental controls such as wind shear and humidity.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

King City Asbestos Corporation (KCAC) Mine Carbon Mineralization Field Test (Final Combined Report)

The overall objective of the KCAC carbon mineralization project is to determine the most successful approaches to CO 2 mineralization of asbestos-bearing serpentinite rocks and mine waste at the closed KCAC asbestos mine. The objectives of Budget Period 1 (BP1) are to characterize the site including, mineralogical and geochemical variability in site materials, baseline conditions including CO 2 flux and local climate, and to establish the designs for the second, experimental phase of the project. As described in Section 3 of this report, all tasks outlined in the Statement of Project Objective (SOPO) have been completed or are on schedule to be completed by the end of BP1. As described in Section 4, all success criteria and milestones have been met or will be met by the end of BP1. We conclude that the findings of BP1 strongly support moving to the second phase of the project to investigate methods to increase the rate of CO 2 mineralization. CO 2 flux monitoring during BP1 established an estimated baseline uptake of CO 2 over the 142,193 m 2 KCAC mine site of 147 and 179 metric tons CO 2 /yr using dynamic closed chamber and eddy covariance monitoring methods, respectively. Mineralogical analysis shows that the loose surficial waste material contains pyroaurite and minor hydromagnesite, indicating past carbon mineralization. Sediment samples at depths below 60 cm contained ~3 wt% brucite and localized high brucite contents of ~24 wt% were measured at depth, indicating significant unreacted material at relatively shallow depths. Together, these results demonstrate that despite prolonged exposure to the atmosphere, reactive material is still present and accessible and is naturally taking up atmospheric CO 2 . However, these natural rates are still well below rates that can be achieved by applying practices that can accelerate the CO 2 mineralization reaction. Designs for the Field-Testing Phase (BP2) of the project have been established, incorporating the findings of BP1. Based on results to date we have selected two methods to accelerate the rate of CO 2 mineralization, which we refer to as the Tilling Method and the Greenhouse Method. We hypothesize that these methods can increase the CO 2 mineralization rate by up to 5 times greater than the natural background rate measured in BP1. An Initial Lifecyle Assessment (LCA) and techno-economic analysis (TEA) were conducted for these two methods. The LCA concluded that the baseline version of the Greenhouse and Tilling methods at commercial scale are expected to achieve a CDR efficiency of 97% and 71%, respectively (cumulative, over 30 years of operation). Incorporating the net CO 2 removal rates calculated in the LCA, the TEA found the levelized, 30-year costs of the baseline Greenhouse and Tilling methods to be $\$$95/t-CO 2 and $\$$119/t-CO 2 , respectively, and that costs could be brought to well below $\$$100/t-CO 2 for either method by scaling up the volume of rock utilized. An initial EH&S Risk Assessment, evaluating the potential EH&S risks of the proposed carbon mineralization methods at commercial scale, was completed by LLNL EH&S professionals for these two methods.

58 GEOSCIENCES↗
DOE OSTI2024

A New Coupled Biogeochemical Modeling Approach Provides Accurate Predictions of Methane and Carbon Dioxide Fluxes Across Diverse Tidal Wetlands

Abstract Tidal wetlands provide valuable ecosystem services, including storing large amounts of carbon. However, the net exchanges of carbon dioxide (CO 2 ) and methane (CH 4 ) in tidal wetlands are highly uncertain. While several biogeochemical models can operate in tidal wetlands, they have yet to be parameterized and validated against high‐frequency, ecosystem‐scale CO 2 and CH 4 flux measurements across diverse sites. We paired the Cohort Marsh Equilibrium Model (CMEM) with a version of the PEPRMT model called PEPRMT‐Tidal, which considers the effects of water table height, sulfate, and nitrate availability on CO 2 and CH 4 emissions. Using a model‐data fusion approach, we parameterized the model with three sites and validated it with two independent sites, with representation from the three marine coasts of North America. Gross primary productivity (GPP) and ecosystem respiration (R eco ) modules explained, on average, 73% of the variation in CO 2 exchange with low model error (normalized root mean square error (nRMSE) <1). The CH 4 module also explained the majority of variance in CH 4 emissions in validation sites ( R 2 = 0.54; nRMSE = 1.15). The PEPRMT‐Tidal‐CMEM model coupling is a key advance toward constraining estimates of greenhouse gas emissions across diverse North American tidal wetlands. Further analyses of model error and case studies during changing salinity conditions guide future modeling efforts regarding four main processes: (a) the influence of salinity and nitrate on GPP, (b) the influence of laterally transported dissolved inorganic C on R eco , (c) heterogeneous sulfate availability and methylotrophic methanogenesis impacts on surface CH 4 emissions, and (d) CH 4 responses to non‐periodic changes in salinity.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

(U) A General-Purpose Code for Correlated Sampling Using Batch Statistics with MCNP6 for Fixed-Source Problems

Correlated sampling can be used to reduce the uncertainty of a difference of tallies by taking advantage of the negative covariance term in the sandwich formula. Booth first showed how correlated sampling can be applied with batch statistics using MCNP’s tally fluctuation chart (TFC) to reduce the uncertainty of a difference of tallies in fixed-source problems. Booth presented a problem in which a 1273% uncertainty in a difference was reduced to 8% by accounting for correlations. Researchers He and Su recently studied correlated sampling using the TFC in MCNP version 5. They determined that the code did not print enough digits in the TFC tally means for accurate batch statistics in some cases. After modifying the source code, they concluded that “correlated sampling can yield a standard deviation of about one magnitude smaller than that predicted by the direct, un-correlated simulation when the changes in system response are small (say about 1%), which is equivalent to saving in CPU time by a factor of 100. Such saving [sic] becomes less significant as the change in system response becomes larger.” He and Su provided the formulas needed to apply batch statistics to compute the correlated uncertainty of a difference of tallies. In this report, we follow up on their work by providing the formulas needed to apply batch statistics to compute the correlated uncertainty of a ratio of tallies and of a difference of two tallies divided by a third tally. We extend these formulas to differences and ratios of ratios. These formulas are applied to reduce the uncertainty associated with calculating a relative sensitivity. He and Su did not investigate the accuracy of their correlated sampling uncertainty estimates. We use their test problems and evaluate the accuracy of the uncertainty estimates by comparing with results obtained from random sampling, and, in simple cases, with theoretical values of the “exact” uncertainties. We find that the uncertainties obtained from batch statistics are accurate as long as at least 100 batches are used. We present a new computer code, COSUBS (COrrelated Sampling Using Batch Statistics), that reads MCNP6 TFCs and applies correlated sampling using batch statistics for the tally combinations that the user specifies. COSUBS is a very general tool that compares all TFCs for a base case and one or two perturbed cases. It computes uncertainties for ratios if given only a base case. This report is organized as follows. The equations to apply batch statistics to the difference of random tallies are reviewed in Sec. II. Section III presents the equations for applying batch statistics to a ratio of random tallies; this is useful for computing relative sensitivities using a one-sided finite difference and the relative sensitivity using the differential operator method. Section IV presents the equations for applying batch statistics to a difference of two random tallies divided by a third; this is useful for computing a relative sensitivities using a central difference. Section V presents the equations for applying batch statistics to a difference of two ratios with four random tallies. Section VI presents the equations for applying batch statistics to a one-sided finite difference estimate of the relative sensitivity of a ratio (this uses four random tallies). Section VII presents the equations for applying batch statistics to a central difference estimate of the relative sensitivity of a ratio (this uses six random tallies). Section VIII presents the equations for applying batch statistics to a sum of random tallies. Section IX discusses how to apply batch statistics using MCNP6. Section X presents COSUBS, describing its command-line options and logic. Sections XI through XVI present numerical results for various test problems. Section XVII is a summary and conclusions. Appendix A derives the theoretical Monte Carlo tally variance given certain assumptions; these variances are used to verify the batch statistics for some of the problems. Appendix B lists the MCNP6 input for the unperturbed example problem. Appendix C presents modifications made to MCNP6.3 to support this work.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Assessing Photovoltaic Capacity Factor Variability Using Long-Term Satellite Derived Solar Resource Data Under Brazilian Climate

Accurate estimation of photovoltaic (PV) energy yield and its variability is essential for reducing financial risk and supporting reliable system planning for rapidly expanding PV markets. In Brazil, high solar adoption and increasing levels of distributed energy resources are beginning to introduce operational challenges such as curtailment and evolving grid requirements. Understanding how natural variability in solar resource propagates into PV system performance is therefore increasingly important for both project design and grid integration. Modern PV yield assessments commonly rely on multi-year meteorological datasets and probabilistic exceedance metrics (e.g., P50/P90) to quantify energy yield uncertainty for project financing. However, the implications of long-term solar resource variability for PV system design choices and high-adoption grid conditions remain less well characterized for rapidly expanding markets such as Brazil. In particular, understanding how weather-driven variability propagates into PV production distributions and capacity factor expectations is important for evaluating curtailment exposure, deployment strategies, and storage requirements in regions experiencing rapid growth of distributed and utility-scale PV. Seasonal and interannual variability in atmospheric conditions can produce substantial fluctuations in monthly PV energy production, which propagate into uncertainty in annual energy yield and capacity factor expectations. Characterizing this variability using long-term meteorological datasets allows probabilistic estimation of PV system performance and provides improved insight into the range of expected PV energy outcomes. This study explores the use of long-term satellite-derived meteorological data from the National Solar Radiation Database (NSRDB) to evaluate the variability of photovoltaic system performance across multiple locations in Brazil. Using a 27-year dataset (1998-2024), PV system simulations are performed to characterize the distribution of annual and seasonal capacity factors and energy yield outcomes, while propagating key sources of meteorological variability and model uncertainty through the PV modeling chain. The analysis also investigates the sensitivity of PV performance outcomes to key system design assumptions within the PV modeling chain, including tracking configuration and system sizing parameters. The resulting probabilistic performance characterization provides insight into how weather-driven variability influences PV production expectations and capacity factor distributions. These results provide a foundation for evaluating how weather-driven variability interacts with high PV adoption and potential storage or curtailment mitigation strategies.

14 SOLAR ENERGY↗