Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “forecast error”

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 253 records · Page 14
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Moving lens effect: Simulations, forecasts, and foreground mitigation

The peculiar motion of massive objects across the line of sight imprints a dipolar temperature anisotropy pattern on the cosmic microwave background known as the moving lens effect. This effect provides a unique probe of the transverse components of the peculiar velocity field, but has not yet been detected due to its small size. We implement and validate a stacking estimator for the moving lens signal using a galaxy catalog as a tracer of massive haloes combined with reconstructed velocities from the galaxy number density field. Using simulations, we forecast detection prospects for the moving lens signal from current and upcoming microwave background and galaxy surveys. Here, we demonstrate a new foreground mitigation strategy likely sufficient for current datasets, and discuss various sources of systematic error and noise. Upcoming galaxy surveys will provide high-significance statistical detections of the moving lens effect.

Astrophysical & cosmological simulations↗
DOE OSTI2025

Adaptive Algebraic Derivative Estimation for Battery Electric Buses Energy Consumption Forecasting

The limited service life of onboard batteries for EVs is a challenge, underscoring the need for real-time battery usage prediction. This paper proposes an adaptive Algebraic Derivative Estimation (ADE) approach for forecasting the energy consumption of battery electric buses. By dynamically adjusting the sliding window length, the adaptive ADE retains the fixed-length ADE’s key advantage—namely, operating online without reliance on extensive historical datasets—while substantially bolstering forecast accuracy by actively trading estimation bias off estimation variance. Comparative experiments against both the conventional ADE with a fixed length and a representative machine learning algorithm, XGBoost, were conducted, with performance evaluated via root mean square error, mean absolute error, and the coefficient of determination. The results demonstrate that the proposed approach significantly outperforms baseline methods.

Cui, Tianyang [The University of Texas at Dallas]↗
DOE OSTI2023

A semi-Lagrangian method for detecting and tracking deep convective clouds in geostationary satellite observations

Automated methods for the detection and tracking of deep convective clouds in geostationary satellite imagery have a vital role in both the forecasting of severe storms and research into their behaviour. Studying the interactions and feedbacks between multiple deep convective clouds (DCC), however, poses a challenge for existing algorithms due to the necessary compromise between false detection and missed detection errors. We utilise an optical flow method to determine the motion of deep convective clouds in GOES-16 ABI imagery in order to construct a semi-Lagrangian framework for the motion of the cloud field, independently of the detection and tracking of cloud objects. The semi-Lagrangian framework allows severe storms to be simultaneously detected and tracked in both spatial and temporal dimensions. For the purpose of this framework we have developed a novel Lagrangian convolution method and a number of novel implementations of morphological image operations that account for the motion of observed objects. These novel methods allow the accurate extension of computer vision techniques to the temporal domain for moving objects such as DCCs. By combining this framework with existing methods for detecting DCCs (including detection of growing cores through cloud top cooling and detection of anvil clouds using brightness temperature), we show that the novel framework enables reductions in errors due to both false and missed detections compared to any of the individual methods, reducing the need to compromise when compared with existing frameworks. The novel framework enables the continuous tracking of anvil clouds associated with detected deep convection after convective activity has stopped, enabling the study of the entire life cycle of DCCs and their associated anvils. Furthermore, we expect this framework to be applicable to a wide range of cases including the detection and tracking of low-level clouds and other atmospheric phenomena. In addition, this framework may be used to combine observations from multiple sources, including satellite observations, weather radar and reanalysis model data.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Modelling and Simulation of a Thermoelectric Heat Pump with Micro-Channel Heat Transfer

Thermoelectric heat pumps (TEHPs) have advantages of modularity and simple design for heating and cooling in some applications. Models of thermoelectric (TE) heat pumps are widely studied. Nevertheless, most existing modeling work focuses on the TE material or air-sourced TE modules. TEHP modeling, especially the relationship between modules and heat exchangers, has not been comprehensively conducted. This work presents a water-to-water TEHP modeling framework that combines Goldsmid’s approach for TE material performance, Gnielinski’s correlation for convective heat transfer, and thermal balance theory for heat exchange networks. This combined framework provides an accurate theoretical analysis of the water-to-water TEHP system. Subsequently, the framework was used to empirically determine TE material properties (electric resistivity, thermal conductivity, and Seebeck coefficient) that minimize modeling errors versus experimentally observed values from the literature. Finally, an additional set of experimental TEHP data was used to validate the model, all with relative absolute deviations of approximately 10% when predicting heating capacity and 10%–20% when forecasting cooling capacity at a 30 K temperature lift. For future work, people can further develop models of TEHPs with an air-based heat sink on one side and water channels on the other side.

Wan, Hanlong↗
DOE OSTI2024

Cosmological parameter forecasts for a CMB-HD survey

We present forecasts on cosmological parameters for a CMB-HD survey. For a $\Lambda$CDM + $N_{eff}$ + $\sum m_\nu$ model, we find $\sigma(n_s) = 0.0013$ and $\sigma(N_{eff}) = 0.014$ using CMB and CMB lensing multipoles in the range of $\ell \in [30, 20000]$, after adding anticipated residual foregrounds, delensing the acoustic peaks, and adding DESI BAO data. This is about a factor of two improvement in ability to probe inflation via $n_s$ compared to precursor CMB surveys. The $N_{eff}$ constraint can rule out light thermal particles back to the end of inflation with 95% CL; for example, it can rule out the QCD axion in a model-independent way assuming the Universe's reheating temperature was high enough that the axion thermalized. We find that delensing the acoustic peaks and adding DESI BAO tightens parameter constraints. We also find that baryonic effects can bias parameters if not marginalized over, and that uncertainties in baryonic effects can increase parameter error bars; however, the latter can be mitigated by including information about baryonic effects from kinetic and thermal Sunyaev-Zel'dovich measurements by CMB-HD. The CMB-HD likelihood and Fisher estimation codes used here are publicly available; the likelihood is integrated with Cobaya to facilitate parameter forecasting.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2026

Evaluation of a high-resolution regional climate simulation for surface and hub-height wind climatology over North America

Assessing the availability of key wind resources requires augmenting observations to support the implementation of wind energy infrastructure. However, observations are limited, necessitating the development of high-resolution, long-term gridded datasets. This study presents a robust, dynamically downscaled climatological dataset, offering 20 years of hourly wind data at a 4 km spatial resolution across North America, and evaluates its performance against observations, including meteorological towers and automated surface-observing system (ASOS) stations, as well as coarse-resolution reanalysis data (the European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis version 5 (ERA5)). Results demonstrate that the downscaled high-resolution wind data outperform ERA5 in regions of complex terrain and coastal areas, with improved overlap coefficients for wind data distributions and reduced root mean square errors (RMSEs) for hub-height and near-surface diurnal wind patterns. The downscaled simulation also captures the synoptic drivers of seasonal wind direction patterns reasonably well, indicated by high wind rose similarity indices. This study also provides an analysis of interannual variability, utilizing the dataset's full 20-year period, and model uncertainty, generated by varying model initial conditions and physics parameterizations across 1-year ensemble members, which are key considerations for wind resource assessment in wind farm development.

17 WIND ENERGY↗
DOE OSTI2026

Autoregressive long-horizon prediction of plasma edge dynamics *

Accurate modeling of scrape-off layer (SOL) and divertor-edge dynamics is vital for designing plasma-facing components in fusion devices. High-fidelity edge fluid/neutral codes such as SOLPS-ITER capture SOL physics with high accuracy, but their computational cost limits broad parameter scans and long transient studies. We present transformer-based, autoregressive surrogates for efficient prediction of 2D, time-dependent plasma edge state fields. Trained on SOLPS-ITER spatiotemporal data for the KSTAR tokamak, the surrogates forecast electron temperature, electron density, and radiated power over extended horizons. We evaluate model variants trained with increasing autoregressive horizons (1–100 steps) on short- and long-horizon prediction tasks. Longer-horizon training systematically improves rollout stability and mitigates error accumulation, enabling stable predictions over hundreds to thousands of steps and reproducing key dynamical features such as the motion of high-radiation regions. Measured end-to-end wall-clock times show the surrogate is orders of magnitude faster than SOLPS-ITER, enabling rapid parameter exploration. Prediction accuracy degrades when the surrogate enters physical regimes not represented in the training dataset, motivating future work on data enrichment and physics-informed constraints. Overall, this approach provides a fast, accurate surrogate for computationally intensive plasma edge simulations, supporting rapid scenario exploration, control-oriented studies, and progress toward real-time applications in fusion devices.

autoregressive deep learning↗
DOE OSTI2026

Hourly PM 2.5 Estimates across California from 2018 to 2023

This study presents a new data set of hourly PM 2.5 concentrations across California from 2018 to 2023 at a three-kilometer resolution. This data set was developed by assimilating observations from PurpleAir and the U.S. EPA Air Quality System monitors into wildfire smoke forecasts from the High-Resolution Rapid Refresh Smoke (HRRR-Smoke) model using the Gridpoint Statistical Interpolation (GSI) three-dimensional variational data assimilation framework. Archived forecasts of modeled wildfire smoke PM 2.5 from HRRR-Smoke create the background field for assimilation, which is then corrected using surface observations of total PM 2.5 . The resulting reanalysis from GSI provides an estimate of total PM 2.5 that minimizes error from both the observational and the model data. Validation results indicate strong performance, with monthly R 2 values ranging from 0.73 to 0.91 across the six-year data set, comparable to other PM 2.5 data sets. Case studies are presented for three major fire events, the 2018 Camp Fire, 2019 Kincade Fire, and 2020 Lightning Complex Fires to demonstrate the data set’s fidelity in resolving plume dynamics and local exposure patterns. Root-mean-squared error averaged over each month scales with average PM 2.5 concentrations, resulting in a low error under typical conditions but higher absolute errors during extreme smoke events. This is the first long-term, hourly PM 2.5 data set of its kind for California and enables the generation of subdaily exposure metrics, such as peak hourly concentrations, exceedance durations, and time-of-day exposure peaks. The novelty and strong validation of this data set make it a compelling resource for future studies on the impact and significance of subdaily PM 2.5 exposure.

PM2.5↗
DOE OSTI2026

Optimizing resource allocation in Miscanthus breeding via sparse testing designs for genomic prediction

Phenotyping high-biomass perennial crops is laborious and the rate of genetic gain in conventional perennial crop breeding programs is typically low. So, it is especially important to identify methods that produce efficiency gains in the breeding process. Miscanthus is a C4 perennial grass with favorable characteristics for producing biomass as a feedstock for biofuels and diverse bio-based products. Increasing biomass yield will increase profitability and environmental benefits, so it is a key target for Miscanthus breeding. In addition, the identification of well-adapted genotypes across a wide range of environmental conditions requires the establishment of multi-environment trials (METs). Sparse testing is a genomic prediction-based strategy that reduces the phenotyping costs in METs by selecting a subset of genotypes to evaluate in a subset of environments and then predicts the performance of the unobserved genotype-environment combinations. A Miscanthus sacchariflorus (MSA) population comprising 336 genotypes observed across three environments was analyzed implementing sparse testing designs. Three prediction models considering main effects (environments, genotypes, genomic) and interaction effects (genotype-by-environment; G×E interaction) were implemented for forecasting dry biomass yield (YDY), total culm (TCM), average internode length (AIL), and culm node number (CNN). Multiple calibration sets based on different compositions and sizes were considered to evaluate performance in terms of the predictive ability (PA) and the mean square error (MSE) for a fixed testing set size. The training set size ranged from 52 to 112 to predict a fixed set of 224 unobserved genotypes across all three environments. The results showed that the model accounting for G×E interaction consistently presented the highest PA and the lowest MSE: for CNN (PA: ~0.77, MSE: ~0.5) and YDY (PA: ~0.70, MSE: ~1.3) while for TCM and AIL these ranged from ~0.28 to 0.41 and ~1.3 to 4.3, respectively. Overall, varying training sets and allocation strategies did not affect PA and MSE, with 52 non-overlapping and 0 overlapping genotypes per environment as the optimal cost-effective allocation framework. This suggests that implementing sparse testing designs could significantly reduce phenotyping costs by fivefold, without compromising PA in breeding programs for perennial crops such as Miscanthus.

Miscanthus sacchariflorus (MSA)↗
DOE OSTI2022

Tracing and Forecasting Metabolic Indices of Cancer Patients Using Patient-Specific Deep Learning Models

We develop a patient-specific dynamical system model from the time series data of the cancer patient’s metabolic panel taken during the period of cancer treatment and recovery. The model consists of a pair of stacked long short-term memory (LSTM) recurrent neural networks and a fully connected neural network in each unit. It is intended to be used by physicians to trace back and look forward at the patient’s metabolic indices, to identify potential adverse events, and to make short-term predictions. When the model is used in making short-term predictions, the relative error in every index is less than 10% in the L ∞ norm and less than 6.3% in the L 1 norm in the validation process. Once a master model is built, the patient-specific model can be calibrated through transfer learning. As an example, we obtain patient-specific models for four more cancer patients through transfer learning, which all exhibit reduced training time and a comparable level of accuracy. This study demonstrates that this modeling approach is reliable and can deliver clinically acceptable physiological models for tracking and forecasting patients’ metabolic indices.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Accuracy of Vaisala RS41 and RS92 Upper Tropospheric Humidity Compared to Satellite Hyperspectral Infrared Measurements

Radiosondes are important for calibrating satellite sensors and assessing sounding retrievals. Vaisala RS41 radiosondes have mostly replaced RS92 in the Global Climate Observing System (GCOS) Reference Upper Air Network (GRUAN) and the conventional network. This study assesses RS41 and RS92 upper tropospheric humidity (UTH) accuracy by comparing with Infrared Atmospheric Sounding Interferometer (IASI) upper tropospheric water vapor absorption spectrum measurements. Using single RS41 and RS92 soundings at three GRUAN and DOE Atmospheric Radiation Measurement (ARM) sites and dual RS92/RS41 launches at three additional GRUAN sites, collocated with cloud-free IASI radiances (OBS), we compute Line-by-Line Radiative Transfer Model radiances for radiosonde profiles (CAL). We analyze OBS-CAL differences from 2015 to 2020, for daytime, nighttime, and dusk/dawn separately if data is available, for standard (STD) RS92 and RS41 processing, and RS92 GRUAN Data Processing (GDP; RS41 GDP is in development). We find that daytime RS41 (even without GDP) has ~1% smaller UTH errors than GDP RS92. RS41 may still have a dry bias of 1–1.5% for both daytime and nighttime, and a similar error for nighttime RS92 GDP, while standard RS92 may have a dry bias of 3–4%. These sonde humidity biases are probably upper limits since “cloud-free” scenes could still be cloud contaminated. Radiances computed from European Centre for Medium-Range Weather Forecasts (ECMWF) analyses match better than radiosondes with IASI measurements, perhaps because ECMWF assimilates IASI measurements. Relative differences between RS41 STD and RS92 GDP, or between radiosondes and ECMWF humidity profiles obtained from the radiance analysis, are consistent with their differences obtained directly from the RH measurements.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Decreasing wind speed extrapolation error via domain-specific feature extraction and selection

Abstract. Model uncertainty is a significant challenge in the wind energy industry and can lead to mischaracterization of millions of dollars' worth of wind resources. Machine learning methods, notably deep artificial neural networks (ANNs), are capable of modeling turbulent and chaotic systems and offer a promising tool to produce high-accuracy wind speed forecasts and extrapolations. This paper uses data collected by profiling Doppler lidars over three field campaigns to investigate the efficacy of using ANNs for wind speed vertical extrapolation in a variety of terrains, and it quantifies the role of domain knowledge in ANN extrapolation accuracy. A series of 11 meteorological parameters (features) are used as ANN inputs, and the resulting output accuracy is compared with that of both standard log-law and power-law extrapolations. It is found that extracted nondimensional inputs, namely turbulence intensity, current wind speed, and previous wind speed, are the features that most reliably improve the ANN's accuracy, providing up to a 65 % and 52 % increase in extrapolation accuracy over log-law and power-law predictions, respectively. The volume of input data is also deemed important for achieving robust results. One test case is analyzed in depth using dimensional and nondimensional features, showing that the feature nondimensionalization drastically improves network accuracy and robustness for sparsely sampled atmospheric cases.

17 WIND ENERGY↗
DOE OSTI2025

Evaluating the potential of short-term instrument deployment to improve distributed wind resource assessment

Distributed wind projects, which are connected at the distribution level of an electricity system or in off-grid applications to serve specific or local energy needs, often rely solely on wind resource models to establish wind speed and energy generation expectations. Historically, anemometer loan programs have provided an affordable avenue for more accurate onsite wind resource assessment, and the lowering cost of lidar systems has shown similar advantages for more recent assessments. While a full 12 months of onsite wind measurement is the standard for correcting model-based long-term wind speed estimates for utility-scale wind farms, the time and capital investment involved in gathering onsite measurements must be reconciled with the energy needs and funding opportunities that drive expedient deployment of distributed wind projects. Much literature exists to quantify the performance of correcting long-term wind speed estimates with 1 or more years of observational data, but few studies explore the impacts of correcting with months-long observational periods. This study aims to answer the question of how short you can go in terms of the observational time period needed to make impactful improvements to model-based long-term wind speed estimates. Three algorithms, multivariable linear regression, adaptive regression splines, and regression trees, are evaluated for their skill at correcting long-term wind resource estimates from the European Centre for Medium-Range Weather Forecasts Reanalysis version 5 (ERA5) using months-long periods of observational data from 66 locations across the US. On average, correction with even 1 month of observations provides significant improvement over the baseline ERA5 wind speed estimates and produces median bias magnitudes and relative errors within 0.22 m s −1 and 4 percentage points of the median bias magnitudes and relative errors achieved using the standard 12 months of data for correction. However, in cases when the shortest observational periods (1 to 2 months) used for correction are not well correlated with the overlapping ERA5 reference, the resultant long-term wind speed errors are worse than those produced using ERA5 without correction. Summer months, which are characterized by weaker relative wind speeds and standard deviations for most of the evaluation sites, tend to produce the worst results for long-term correction using months-long observations. The three tested algorithms perform similarly for long-term wind speed bias; however, regression trees perform notably worse than multivariable linear regression and adaptive regression splines in terms of correlation when using 6 months or less of observational data for correction. Translating the analysis to wind energy, median relative errors in the capacity factor are on average within 10 % using 1 month of training. If the observation period used for correction is not well correlated with the reference data, however, misrepresentation of the observed capacity factor can be substantial. The risk associated with poor correlation between the observed and reference datasets decreases with increasing training period length. In the worst-correlation scenarios, the median capacity factor relative errors from using 1, 3, and 6 months are within 47 %, 26 %, and 16 %, respectively.

17 WIND ENERGY↗
DOE OSTI2022

Sparsifying priors for Bayesian uncertainty quantification in model discovery

We propose a probabilistic model discovery method for identifying ordinary differential equations governing the dynamics of observed multivariate data. Our method is based on the sparse identification of nonlinear dynamics (SINDy) framework, where models are expressed as sparse linear combinations of pre-specified candidate functions. Promoting parsimony through sparsity leads to interpretable models that generalize to unknown data. Instead of targeting point estimates of the SINDy coefficients, we estimate these coefficients via sparse Bayesian inference. The resulting method, uncertainty quantification SINDy (UQ-SINDy), quantifies not only the uncertainty in the values of the SINDy coefficients due to observation errors and limited data, but also the probability of inclusion of each candidate function in the linear combination. UQ-SINDy promotes robustness against observation noise and limited data, interpretability (in terms of model selection and inclusion probabilities) and generalization capacity for out-of-sample forecast. Sparse inference for UQ-SINDy employs Markov chain Monte Carlo, and we explore two sparsifying priors: the spike and slab prior, and the regularized horseshoe prior. UQ-SINDy is shown to discover accurate models in the presence of noise and with orders-of-magnitude less data than current model discovery methods, thus providing a transformative method for real-world applications which have limited data.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Examining the Impacts of Great Lakes Temperature Perturbations on Simulated Precipitation in the Northeastern United States

Most inland water bodies are not resolved by general circulation models, requiring that lake surface temperatures be estimated. Given the large spatial and temporal variability of the surface temperatures of the North American Great Lakes, such estimations can introduce errors when used as lower boundary conditions for dynamical downscaling. Lake surface temperatures (LSTs) influence moisture and heat fluxes, thus impacting precipitation within the immediate region and potentially in regions downwind of the lakes. For this study, the Advanced Research version of the Weather Research and Forecasting Model (WRF-ARW) was used to simulate precipitation over the six New England states during a 5-yr historical period. The model simulation was repeated with perturbed LSTs, ranging from 10°C below to 10°C above baseline values obtained from reanalysis data, to determine whether the inclusion of erroneous LST values has an impact on simulated precipitation and synoptic-scale features. Results show that simulated precipitation in New England is statistically correlated with LST perturbations, but this region falls on a wet–dry line of a larger bimodal distribution. Wetter conditions occur to the north and drier conditions occur to the south with increasing LSTs, particularly during the warm season. Additionally, the precipitation differences coincide with large-scale anomalous temperature, pressure, and moisture patterns. Care must therefore be taken to ensure reasonably accurate Great Lakes surface temperatures when simulating precipitation, especially in southeastern Canada, Maine, and the mid-Atlantic region.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Predicting oxidation damage of ultra high-temperature carbide ceramics in extreme environments using machine learning

Determining the oxidation resistance of UHTC carbides in extreme environments is challenging theoretically and experimentally due to the high dimensional complexity of influencing variables and intricate testing setups. Herein we demonstrate the use of machine learning (ML) models trained with experimental literature data to predict the oxide thickness of UHTC carbides exposed to air based on composition, mean grain size, relative densification, holding time, and temperature. A multi-dimensional database with 76 occurrences is created containing experimental results of Hf, Zr, and Ta carbides plus additives. In this study, the preprocessed database is then used to train ML models to predict their oxidation behavior. The trained model predicts the oxidation damage in the form of an average oxide thickness in UHTC carbides with a Mean Absolute Error (MAE) of ±65.45 μm for samples in the testing set that developed thicknesses up to 1000 μm. The model successfully predicted oxidation damage for a recession rate lower than 60 μm/min. It is noticed that the ensemble method MAE is increased to ±134.34 μm while forecasting the oxidation of samples with a recession rate higher than the threshold. The unprecedented approach is a novel way to predict the damage through the oxidation of carbide compounds before processing for a smarter design with room for improvement.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Streamlining Ocean Dynamics Modeling with Fourier Neural Operators: A Multiobjective Hyperparameter and Architecture Optimization Approach

Training an effective deep learning model to learn ocean processes involves careful choices of various hyperparameters. We leverage DeepHyper’s advanced search algorithms for multiobjective optimization, streamlining the development of neural networks tailored for ocean modeling. The focus is on optimizing Fourier neural operators (FNOs), a data-driven model capable of simulating complex ocean behaviors. Selecting the correct model and tuning the hyperparameters are challenging tasks, requiring much effort to ensure model accuracy. DeepHyper allows efficient exploration of hyperparameters associated with data preprocessing, FNO architecture-related hyperparameters, and various model training strategies. We aim to obtain an optimal set of hyperparameters leading to the most performant model. Moreover, on top of the commonly used mean squared error for model training, we propose adopting the negative anomaly correlation coefficient as the additional loss term to improve model performance and investigate the potential trade-off between the two terms. The numerical experiments show that the optimal set of hyperparameters enhanced model performance in single timestepping forecasting and greatly exceeded the baseline configuration in the autoregressive rollout for long-horizon forecasting up to 30 days. Utilizing DeepHyper, we demonstrate an approach to enhance the use of FNO in ocean dynamics forecasting, offering a scalable solution with improved precision.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Resiliency in numerical algorithm design for extreme scale simulations

Here this work is based on the seminar titled ‘Resiliency in Numerical Algorithm Design for Extreme Scale Simulations’ held March 1–6, 2020, at Schloss Dagstuhl, that was attended by all the authors. Advanced supercomputing is characterized by very high computation speeds at the cost of involving an enormous amount of resources and costs. A typical large-scale computation running for 48 h on a system consuming 20 MW, as predicted for exascale systems, would consume a million kWh, corresponding to about 100k Euro in energy cost for executing 10 23 floating-point operations. It is clearly unacceptable to lose the whole computation if any of the several million parallel processes fails during the execution. Moreover, if a single operation suffers from a bit-flip error, should the whole computation be declared invalid? What about the notion of reproducibility itself: should this core paradigm of science be revised and refined for results that are obtained by large-scale simulation? Naive versions of conventional resilience techniques will not scale to the exascale regime: with a main memory footprint of tens of Petabytes, synchronously writing checkpoint data all the way to background storage at frequent intervals will create intolerable overheads in runtime and energy consumption. Forecasts show that the mean time between failures could be lower than the time to recover from such a checkpoint, so that large calculations at scale might not make any progress if robust alternatives are not investigated. More advanced resilience techniques must be devised. The key may lie in exploiting both advanced system features as well as specific application knowledge. Research will face two essential questions: (1) what are the reliability requirements for a particular computation and (2) how do we best design the algorithms and software to meet these requirements? While the analysis of use cases can help understand the particular reliability requirements, the construction of remedies is currently wide open. One avenue would be to refine and improve on system- or application-level checkpointing and rollback strategies in the case an error is detected. Developers might use fault notification interfaces and flexible runtime systems to respond to node failures in an application-dependent fashion. Novel numerical algorithms or more stochastic computational approaches may be required to meet accuracy requirements in the face of undetectable soft errors. These ideas constituted an essential topic of the seminar. The goal of this Dagstuhl Seminar was to bring together a diverse group of scientists with expertise in exascale computing to discuss novel ways to make applications resilient against detected and undetected faults. In particular, participants explored the role that algorithms and applications play in the holistic approach needed to tackle this challenge. This article gathers a broad range of perspectives on the role of algorithms, applications and systems in achieving resilience for extreme scale simulations. The ultimate goal is to spark novel ideas and encourage the development of concrete solutions for achieving such resilience holistically.

79 ASTRONOMY AND ASTROPHYSICS↗