Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “prediction”

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 541 records · Page 30

Building thermal load prediction through shallow machine learning and deep learning

Building thermal load prediction informs the optimization of cooling plant and thermal energy storage. Physics-based prediction models of building thermal load are constrained by the model and input complexity. In this study, we developed 12 data-driven models (7 shallow learning, 2 deep learning, and 3 heuristic methods) to predict building thermal load and compared shallow machine learning and deep learning. The 12 prediction models were compared with the measured cooling demand. It was found XGBoost (Extreme Gradient Boost) and LSTM (Long Short Term Memory) provided the most accurate load prediction in the shallow and deep learning category, and both outperformed the best baseline model, which uses the previous day's data for prediction. Then, we discussed how the prediction horizon and input uncertainty would influence the load prediction accuracy. Major conclusions are twofold: first, LSTM performs well in short-term prediction (1 h ahead) but not in long term prediction (24 h ahead), because the sequential information becomes less relevant and accordingly not so useful when the prediction horizon is long. Second, the presence of weather forecast uncertainty deteriorates XGBoost's accuracy and favors LSTM, because the sequential information makes the model more robust to input uncertainty. Training the model with the uncertain rather than accurate weather data could enhance the model's robustness. Finally, our findings have two implications for practice. First, LSTM is recommended for short-term load prediction given that weather forecast uncertainty is unavoidable. Second, XGBoost is recommended for long term prediction, and the model should be trained with the presence of input uncertainty.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Application of a fuzzy neural network model in predicting polycyclic aromatic hydrocarbon-mediated perturbations of the Cyp1b1 transcriptional regulatory network in mouse skin

Polycyclic aromatic hydrocarbons (PAHs) are present in the environment as complex mixtures with components that have diverse carcinogenic potencies and mostly unknown interactive effects. Non-additive PAH interactions have been observed in regulation of cytochrome P450 (CYP) gene expression in the CYP1 family. To better understand and predict biological effects of complex mixtures, such as environmental PAHs, an 11 gene input-1 gene output fuzzy neural network (FNN) was developed for predicting PAH-mediated perturbations of dermal Cyp1b1 transcription in mice. Input values were generalized using fuzzy logic into low, medium, and high fuzzy subsets, and sorted using k-means clustering to create Mamdani logic functions for predicting Cyp1b1 mRNA expression. Model testing was performed with data from microarray analysis of skin samples from FVB/N mice treated with toluene (vehicle control), dibenzo[def,p]chrysene (DBC), benzo[a]pyrene (BaP), or 1 of 3 combinations of diesel particulate extract (DPE), coal tar extract (CTE) and cigarette smoke condensate (CSC) using leave-one-out cross-validation. Predictions were within 1 log{sub 2} fold change unit of microarray data, with the exception of the DBC treatment group, where the unexpected down-regulation of Cyp1b1 expression was predicted but did not reach statistical significance on the microarrays. Adding CTE to DPE was predicted to increase Cyp1b1 expression, whereas adding CSC to CTE and DPE was predicted to have no effect, in agreement with microarray results. The aryl hydrocarbon receptor repressor (Ahrr) was determined to be the most significant input variable for model predictions using back-propagation and normalization of FNN weights. - Highlights: ► Tested a model to predict PAH mixture-mediated changes in Cyp1b1 expression ► Quantitative predictions in agreement with microarrays for Cyp1b1 induction ► Unexpected difference in expression between DBC and other treatments predicted ► Model predictions for combining PAH mixtures in agreement with microarrays ► Predictions highly dependent on aryl hydrocarbon receptor repressor expression.

60 APPLIED LIFE SCIENCES↗

Heavy Ion Collisions at the LHC - Last Call for Predictions

In August 2006, the CERN Theory Unit announced to restructure its visitor program and to create a 'CERN Theory Institute', where 1-3 month long specific programs can take place. The first such Institute was held from 14 May to 10 June 2007, focusing on 'Heavy Ion Collisions at the LHC - Last Call for Predictions'. It brought together close to 100 scientists working on the theory of ultra-relativistic heavy ion collisions. The aim of this workshop was to review and document the status of expectations and predictions for the heavy ion program at the Large Hadron Collider LHC before its start. LHC will explore heavy ion collisions at {approx} 30 times higher center of mass energy than explored previously at the Relativistic Heavy Ion Collider RHIC. So, on the one hand, the charge of this workshop provided a natural forum for the exchange of the most recent ideas, and allowed to monitor how the understanding of heavy ion collisions has evolved in recent years with the data from RHIC, and with the preparation of the LHC experimental program. On the other hand, the workshop aimed at a documentation which helps to distinguish pre- from post-dictions. An analogous documentation of the 'Last Call for Predictions' [1] was prepared prior to the start of the heavy-ion program at the Relativistic Heavy Ion Collider RHIC, and it proved useful in the subsequent discussion and interpretation of RHIC data. The present write-up is the documentation of predictions for the LHC heavy ion program, received or presented during the CERN TH Institute. The set-up of the CERN TH Institute allowed us to aim for the wide-most coverage of predictions. There were more than 100 presentations and discussions during the workshop. Moreover, those unable to attend could still participate by submitting predictions in written form during the workshop. This followed the spirit that everybody interested in making a prediction had the right to be heard. To arrive at a concise document, we required that each prediction should be summarized on at most two pages, and that predictions should be presented, whenever possible, in figures which display measurable quantities. Full model descriptions were not accepted--the authors were encouraged to indicate the relevant references for the interested reader. Participants had the possibility to submit multiple contributions on different topics, but it was part of the subsequent editing process to ensure that predictions on neighboring topics were merged wherever possible. The contributions summarized here are organized in several sections,--though some of them contain material related with more than one section--roughly by going from low transverse momentum to high transverse momentum and from abundant to rare measurements. In the low transverse momentum regime, we start with predictions on multiplicity distributions, azimuthal asymmetries in particle production and hadronic flavor observables, followed by correlation and fluctuation measurements. The contributions on hard probes at the LHC start with predictions for single inclusive high transverse momentum spectra, and jets, followed by heavy quark and quarkonium measurements, leptonic probes and photons. A final section 'Others' encompasses those predictions which do not fall naturally within one of the above-mentioned categories, or discuss the more speculative phenomena that may be explored at the LHC.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Suppressing simulation bias in multi-modal data using transfer learning

Abstract Many problems in science and engineering require making predictions based on few observations. To build a robust predictive model, these sparse data may need to be augmented with simulated data, especially when the design space is multi-dimensional. Simulations, however, often suffer from an inherent bias. Estimation of this bias may be poorly constrained not only because of data sparsity, but also because traditional predictive models fit only one type of observed outputs, such as scalars or images, instead of all available output data modalities, which might have been acquired and simulated at great cost. To break this limitation and open up the path for multi-modal calibration, we propose to combine a novel, transfer learning technique for suppressing the bias with recent developments in deep learning, which allow building predictive models with multi-modal outputs. First, we train an initial neural network model on simulated data to learn important correlations between different output modalities and between simulation inputs and outputs. Then, the model is partially retrained, or transfer learned, to fit the experiments; a method that has never been implemented in this type of architecture. Using fewer than 10 inertial confinement fusion experiments for training, transfer learning systematically improves the simulation predictions while a simple output calibration, which we design as a baseline, makes the predictions worse. We also offer extensive cross-validation with real and carefully designed synthetic data. The method described in this paper can be applied to a wide range of problems that require transferring knowledge from simulations to the domain of experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Evaluating the Trustworthiness of Explainable Artificial Intelligence (XAI) Methods Applied to Regression Predictions of Arctic Sea Ice Motion

Abstract Recent advances in explainable artificial intelligence (XAI) methods show promise for understanding predictions made by machine learning (ML) models. XAI explains how the input features are relevant or important for the model predictions. We train linear regression (LR) and convolutional neural network (CNN) models to make 1-day predictions of sea ice velocity in the Arctic from inputs of present-day wind velocity and previous-day ice velocity and concentration. We apply XAI methods to the CNN and compare explanations to variance explained by LR. We confirm the feasibility of using a novel XAI method [i.e., global layerwise relevance propagation (LRP)] to understand ML model predictions of sea ice motion by comparing it to established techniques. We investigate a suite of linear, perturbation-based, and propagation-based XAI methods in both local and global forms. Outputs from different explainability methods are generally consistent in showing that wind speed is the input feature with the highest contribution to ML predictions of ice motion, and we discuss inconsistencies in the spatial variability of the explanations. Additionally, we show that the CNN relies on both linear and nonlinear relationships between the inputs and uses nonlocal information to make predictions. LRP shows that wind speed over land is highly relevant for predicting ice motion offshore. This provides a framework to show how knowledge of environmental variables (i.e., wind) on land could be useful for predicting other properties (i.e., sea ice velocity) elsewhere. Significance Statement Explainable artificial intelligence (XAI) is useful for understanding predictions made by machine learning models. Our research establishes trustability in a novel implementation of an explainable AI method known as layerwise relevance propagation for Earth science applications. To do this, we provide a comparative evaluation of a suite of explainable AI methods applied to machine learning models that make 1-day predictions of Arctic sea ice velocity. We use explainable AI outputs to understand how the input features are used by the machine learning to predict ice motion. Additionally, we show that a convolutional neural network uses nonlinear and nonlocal information in making its predictions. We take advantage of the nonlocality to investigate the extent to which knowledge of wind on land is useful for predicting sea ice velocity elsewhere.

Hoffman, Lauren [Scripps Institution of Oceanograp↗

Predicting September Arctic Sea Ice: A Multimodel Seasonal Skill Comparison

This study quantifies the state of the art in the rapidly growing field of seasonal Arctic sea ice prediction. A novel multimodel dataset of retrospective seasonal predictions of September Arctic sea ice is created and analyzed, consisting of community contributions from 17 statistical models and 17 dynamical models. Prediction skill is compared over the period 2001–20 for predictions of pan-Arctic sea ice extent (SIE), regional SIE, and local sea ice concentration (SIC) initialized on 1 June, 1 July, 1 August, and 1 September. This diverse set of statistical and dynamical models can individually predict linearly detrended pan-Arctic SIE anomalies with skill, and a multimodel median prediction has correlation coefficients of 0.79, 0.86, 0.92, and 0.99 at these respective initialization times. Regional SIE predictions have similar skill to pan-Arctic predictions in the Alaskan and Siberian regions, whereas regional skill is lower in the Canadian, Atlantic, and central Arctic sectors. The skill of dynamical and statistical models is generally comparable for pan-Arctic SIE, whereas dynamical models outperform their statistical counterparts for regional and local predictions. The prediction systems are found to provide the most value added relative to basic reference forecasts in the extreme SIE years of 1996, 2007, and 2012. SIE prediction errors do not show clear trends over time, suggesting that there has been minimal change in inherent sea ice predictability over the satellite era. Overall, this study demonstrates that there are bright prospects for skillful operational predictions of September sea ice at least 3 months in advance.

54 ENVIRONMENTAL SCIENCES↗

Subseasonal Earth System Prediction with CESM2

Abstract Prediction systems to enable Earth system predictability research on the subseasonal time scale have been developed with the Community Earth System Model, version 2 (CESM2) using two configurations that differ in their atmospheric components. One system uses the Community Atmosphere Model, version 6 (CAM6) with its top near 40 km, referred to as CESM2(CAM6). The other employs the Whole Atmosphere Community Climate Model, version 6 (WACCM6) whose top extends to ∼140 km, and it includes fully interactive tropospheric and stratospheric chemistry [CESM2(WACCM6)]. Both systems are utilized to carry out subseasonal reforecasts for the 1999–2020 period following the Subseasonal Experiment’s (SubX) protocol. Subseasonal prediction skill from both systems is compared to those of the National Oceanic and Atmospheric Administration CFSv2 and European Centre for Medium-Range Weather Forecasts (ECMWF) operational models. CESM2(CAM6) and CESM2(WACCM6) show very similar subseasonal prediction skill of 2-m temperature, precipitation, the Madden–Julian oscillation, and North Atlantic Oscillation to its previous version and to the NOAA CFSv2 model. Overall, skill of CESM2(CAM6) and CESM2(WACCM6) is a little lower than that of the ECMWF system. In addition to typical output provided by subseasonal prediction systems, CESM2 reforecasts provide comprehensive datasets for predictability research of multiple Earth system components, including three-dimensional output for many variables, and output specific to the mesosphere and lower-thermosphere (MLT) region from CESM2(WACCM6). It is shown that sudden stratosphere warming events, and the associated variability in the MLT, can be predicted ∼10 days in advance. Weekly real-time forecasts and reforecasts with CESM2(CAM6) and CESM2(WACCM6) are freely available. Significance Statement We describe here the design and prediction skill of two subseasonal prediction systems based on two configurations of the Community Earth System Model, version 2 (CESM2): CESM2 with the Community Atmosphere Model, version 6 [CESM2(CAM6)] and CESM 2 with Whole Atmosphere Community Climate Model, version 6 [CESM2(WACCM6)] as its atmospheric component. These two systems provide a foundation for community-model based subseasonal prediction research. The CESM2(WACCM6) system provides a novel capability to explore the predictability of the stratosphere, mesosphere, and lower thermosphere. Both CESM2(CAM6) and CESM2(WACCM6) demonstrate subseasonal surface prediction skill comparable to that of the NOAA CFSv2 model, and a little lower than that of the ECMWF forecasting system. CESM2 reforecasts provide a comprehensive dataset for predictability research of multiple aspects of the Earth system, including the whole atmosphere up to 140 km, land, and sea ice. Weekly real-time forecasts, reforecasts, and models are publicly available.

54 ENVIRONMENTAL SCIENCES↗

Lifetime predictions for the Solar Maximum Mission (SMM) and San Marco spacecraft

Lifetime prediction techniques developed by the Goddard Space Flight Center (GSFC) Flight Dynamics Division (FDD) are described. These techniques were developed to predict the Solar Maximum Mission (SMM) spacecraft orbit, which is decaying due to atmospheric drag, with reentry predicted to occur before the end of 1989. Lifetime predictions were also performed for the Long Duration Exposure Facility (LDEF), which was deployed on the 1984 SMM repair mission and is scheduled for retrieval on another Space Transportation System (STS) mission later this year. Concepts used in the lifetime predictions were tested on the San Marco spacecraft, which reentered the Earth's atmosphere on December 6, 1988. Ephemerides predicting the orbit evolution of the San Marco spacecraft until reentry were generated over the final 90 days of the mission when the altitude was less than 380 kilometers. The errors in the predicted ephemerides are due to errors in the prediction of atmospheric density variations over the lifetime of the satellite. To model the time dependence of the atmospheric densities, predictions of the solar flux at the 10.7-centimeter wavelength were used in conjunction with Harris-Priester (HP) atmospheric density tables. Orbital state vectors, together with the spacecraft mass and area, are used as input to the Goddard Trajectory Determination System (GTDS). Propagations proceed in monthly segments, with the nominal atmospheric drag model scaled for each month according to the predicted monthly average value of F10.7. Calibration propagations are performed over a period of known orbital decay to obtain the effective ballistic coefficient. Progagations using plus or minus 2 sigma solar flux predictions are also generated to estimate the despersion in expected reentry dates. Definitive orbits are compared with these predictions as time expases. As updated vectors are received, these are also propagated to reentryto continually update the lifetime predictions.

Smith, E. A.↗

DRA/NASA/ONERA Collaboration on Icing Research: Prediction of Airfoil Ice Accretion - Part 2

This report presents results from a joint study by DRA, NASA, and ONERA for the purpose of comparing, improving, and validating the aircraft icing computer codes developed by each agency. These codes are of three kinds: (1) water droplet trajectory prediction, (2) ice accretion modeling, and (3) transient electrothermal deicer analysis. In this joint study, the agencies compared their code predictions with each other and with experimental results. These comparison exercises were published in three technical reports, each with joint authorship. DRA published and had first authorship of Part 1 - Droplet Trajectory Calculations, NASA of Part 2 - Ice Accretion Prediction, and ONERA of Part 3 - Electrothermal Deicer Analysis. The results cover work done during the period from August 1986 to late 1991. As a result, all of the information in this report is dated. Where necessary, current information is provided to show the direction of current research. In this present report on ice accretion, each agency predicted ice shapes on two dimensional airfoils under icing conditions for which experimental ice shapes were available. In general, all three codes did a reasonable job of predicting the measured ice shapes. For any given experimental condition, one of the three codes predicted the general ice features (i.e., shape, impingement limits, mass of ice) somewhat better than did the other two. However, no single code consistently did better than the other two over the full range of conditions examined, which included rime, mixed, and glaze ice conditions. In several of the cases, DRA showed that the user's knowledge of icing can significantly improve the accuracy of the code prediction. Rime ice predictions were reasonably accurate and consistent among the codes, because droplets freeze on impact and the freezing model is simple. Glaze ice predictions were less accurate and less consistent among the codes, because the freezing model is more complex and is critically dependent upon unsubstantiated heat transfer and surface roughness models. Thus, heat transfer prediction methods used in the codes became the subject for a separate study in this report to compare predicted heat transfer coefficients with a limited experimental database of heat transfer coefficients for cylinders with simulated glaze and rime ice shapes. The codes did a good job of predicting heat transfer coefficients near the stagnation region of the ice shapes. But in the region of the ice horns, all three codes predicted heat transfer coefficients considerably higher than the measured values. An important conclusion of this study is that further research is needed to understand the finer detail of of the glaze ice accretion process and to develop improved glaze ice accretion models.

Wright, William B.↗

Assessment of Current Jet Noise Prediction Capabilities

An assessment was made of the capability of jet noise prediction codes over a broad range of jet flows, with the objective of quantifying current capabilities and identifying areas requiring future research investment. Three separate codes in NASA s possession, representative of two classes of jet noise prediction codes, were evaluated, one empirical and two statistical. The empirical code is the Stone Jet Noise Module (ST2JET) contained within the ANOPP aircraft noise prediction code. It is well documented, and represents the state of the art in semi-empirical acoustic prediction codes where virtual sources are attributed to various aspects of noise generation in each jet. These sources, in combination, predict the spectral directivity of a jet plume. A total of 258 jet noise cases were examined on the ST2JET code, each run requiring only fractions of a second to complete. Two statistical jet noise prediction codes were also evaluated, JeNo v1, and Jet3D. Fewer cases were run for the statistical prediction methods because they require substantially more resources, typically a Reynolds-Averaged Navier-Stokes solution of the jet, volume integration of the source statistical models over the entire plume, and a numerical solution of the governing propagation equation within the jet. In the evaluation process, substantial justification of experimental datasets used in the evaluations was made. In the end, none of the current codes can predict jet noise within experimental uncertainty. The empirical code came within 2dB on a 1/3 octave spectral basis for a wide range of flows. The statistical code Jet3D was within experimental uncertainty at broadside angles for hot supersonic jets, but errors in peak frequency and amplitude put it out of experimental uncertainty at cooler, lower speed conditions. Jet3D did not predict changes in directivity in the downstream angles. The statistical code JeNo,v1 was within experimental uncertainty predicting noise from cold subsonic jets at all angles, but did not predict changes with heating of the jet and did not account for directivity changes at supersonic conditions. Shortcomings addressed here give direction for future work relevant to the statistical-based prediction methods. A full report will be released as a chapter in a NASA publication assessing the state of the art in aircraft noise prediction.

Hunter, Craid A.↗

A multiscale recurrent neural network model for predicting energy production from geothermal reservoirs

Optimization of energy production from geothermal reservoirs requires reliable prediction of energy production performance under alternative operation and development scenarios. Traditionally, reservoir simulation models are used for the evaluation and screening of alternative production and development plans. However, simulation models require extensive data collection and modeling efforts and are time-consuming to build, run, and update. Data-driven predictive models, on the other hand, can serve as efficient prediction tools that can be used for decision support and management of daily operations and surveillance activities. Data-driven models become particularly attractive when a reservoir simulation model for a field does not exist and/or is difficult to build. Machine learning (ML)-based data-driven models that have recently become popular in several fields exploit statistical patterns and relations in training data to generate predictions. As such, they tend to perform better in interpolation problems (that is, prediction within the training data range) than when they are used to extrapolate beyond the training data. Production data from geothermal reservoirs tend to exhibit short-term variabilities as well as long-term trends, such as monotonically declining production temperatures. Capturing both short-term features and long-term trends with ML-based models is not trivial. We evaluate the use of recurrent neural networks (RNN) for the prediction of energy production from geothermal reservoirs. RNN is a class of ML architectures that are used to represent and predict sequential/dynamic data. Thus, it can be challenging to apply RNN to problems where long-term trends must be captured and extrapolation beyond the training data range is needed. We introduce the multiscale RNN architecture to extend the application of RNN to detect and predict both short-term variabilities and long-term trends in geothermal data. The developed architecture consists of a long-term component to only capture low-frequency data patterns, and a short-term component to detect features with higher frequency and more nonlinearity. The final prediction is obtained by combining the long-term and short-term predictions. Both synthetic and field data are used to evaluate the presented multiscale RNN model. The prediction performance of the multiscale RNN is compared against those obtained from the regular RNN and the autoregressive (AR) model. The results suggest that the multiscale architecture improves the long-term prediction performance of the regular RNN and enhances its robustness against noise.

15 GEOTHERMAL ENERGY↗

Assessment of Arctic and Antarctic Sea Ice Predictability in CMIP5 Decadal Hindcasts

This paper examines the ability of coupled global climate models to predict decadal variability of Arctic and Antarctic sea ice. We analyze decadal hindcasts/predictions of 11 Coupled Model Intercomparison Project Phase 5 (CMIP5) models. Decadal hindcasts exhibit a large multimodel spread in the simulated sea ice extent, with some models deviating significantly from the observations as the predicted ice extent quickly drifts away from the initial constraint. The anomaly correlation analysis between the decadal hindcast and observed sea ice suggests that in the Arctic, for most models, the areas showing significant predictive skill become broader associated with increasing lead times. This area expansion is largely because nearly all the models are capable of predicting the observed decreasing Arctic sea ice cover. Sea ice extent in the North Pacific has better predictive skill than that in the North Atlantic (particularly at a lead time of 3-7 years), but there is a reemerging predictive skill in the North Atlantic at a lead time of 6-8 years. In contrast to the Arctic, Antarctic sea ice decadal hindcasts do not show broad predictive skill at any timescales, and there is no obvious improvement linking the areal extent of significant predictive skill to lead time increase. This might be because nearly all the models predict a retreating Antarctic sea ice cover, opposite to the observations. For the Arctic, the predictive skill of the multi-model ensemble mean outperforms most models and the persistence prediction at longer timescales, which is not the case for the Antarctic. Overall, for the Arctic, initialized decadal hindcasts show improved predictive skill compared to uninitialized simulations, although this improvement is not present in the Antarctic.

sea ice↗

EPIsembleVis: A geo-visual analysis and comparison of the prediction ensembles of multiple COVID-19 models

In this work, we present EPIsembleVis, a web-based comparative visual analysis tool for evaluating the consistency of multiple COVID-19 prediction models. Our approach analyzes a collection of COVID-19 predictions from different epidemiological models as an ensemble and utilizes two metrics to quantify model performance. These metrics include (a) prediction uncertainty (represented as the dispersion of predictions in each ensemble) and (b) prediction error (calculated by comparing individual model predictions with the recorded data). Through an interactive visual interface, our approach provides a data-driven workflow for (a) selecting and constructing the COVID-19 model prediction ensemble based on the spatiotemporal overlap of available predictions of multiple epidemiological models, (b) quantifying the model performance using both the uncertainty of each model prediction ensemble, and the error of each ensemble member that represents individual model predictions, and (c) visualizing the spatiotemporal variability in the projection performance of individual models using a suite of novel ensemble visualization techniques, such as the data availability map, a spatiotemporal textured-tile calendar, multivariate rose chart, and time-series leaflet glyph. We demonstrate the capability of our ensemble visual interface through a case study that investigates the performance of weekly COVID-19 predictions, which are provided through the COVID-19 Forecast Hub UMass-Amherst Influenza Forecasting Center of Excellence [47] for the United States and United States Territories. The EPIsembleVis tool is implemented using open-source web technologies and adaptive system design, rendering it interoperable with Elasticsearch and Kibana for automatically ingesting COVID-19 predictions from online repositories, and it is generalizable for analyzing worldwide projections from more epidemiological models.

60 APPLIED LIFE SCIENCES↗

Denoising Autoencoder for Reconstructing Sensor Observation Data and Predicting Evapotranspiration: Noisy and Missing Values Repair and Uncertainty Quantification

Abstract Machine learning (ML) methods applied in scientific research often deal with interrelated features in high‐dimensional data. Reducing data noise and redundancy is needed to increase prediction accuracy and efficiency especially when dealing with data from field sensors. We explored an unsupervised learning method, the denoising autoencoder (DAE), to extract the underlying data structure from noisy raw data in the context of predicting hydrologic quantities from multiple field sensors. These sensors have intrinsic instrumental noise and occasional malfunctions that cause missing values. Our DAE neural network reconstructed meteorological sensor data containing noise and missing values to predict evapotranspiration in a mountainous watershed. The DAE reconstructed the sensor variables with a mean coefficient of determination value of 0.77 across 15 dimensions representing individual sensors. It reduced variance and bias uncertainties compared to a classical autoencoder model. The reconstruction quality varied across dimensions depending on their cross‐correlation and alignment with the underlying data structure. Uncertainties arising from the model structure were overall higher than those resulting from data corruption. We attached the DAE structure to a downstream ET‐prediction neural network in three formats and achieved reasonably accurate ET predictions . The use of the DAE notably reduced variance uncertainty in ET prediction. However, excessive variance reduction may be accompanied by an increase in bias due to the intrinsic bias‐variance tradeoff. Our method of evaluating and reducing uncertainties in aggregated data from different sources can be used to improve predictive models, process understanding, and uncertainty quantification for better water resource management. Plain Language Summary We present a machine learning method, namely the denoising autoencoder, which reduces the effects of data noise and missing values typically present in scientific data sets collected through sensor measurements. This method selects the most relevant information from noisy raw data collected by the instruments and fills in missing values. To demonstrate the effectiveness of our method, we applied it to predict evapotranspiration, a hydrologic variable that represents the water moved from the land surface to the atmosphere through a combination of evaporation and plant water use (transpiration). We also used a random sampling technique (the Monte Carlo method) to compare the uncertainty in the predictions when using the raw and noisy data versus the reconstructed data. The denoising process produced more accurate predictions of evapotranspiration with less uncertainty. Improved predictions of evapotranspiration can lead to a better understanding and accounting of water budgets. This ML approach is broadly suitable for a wide variety of applications that involve noisy sensor data with missing values. Key Points We used a denoising autoencoder (DAE) neural network to reduce noise in meteorological and soil sensor observations by on average We used Monte Carlo sampling to estimate the bias and variance of all model outputs, including uncertainty sources from data and the model We attached the DAE component to a downstream neural network to predict ET with the variance reduced by , compared to that without the DAE

denoising autoencoder↗

Subseasonal Forecasting and MJO Teleconnections in Machine Learning Weather Prediction Models

Abstract In recent years, machine‐learning (ML) models trained on reanalysis data have rivaled physics‐based forecast models in terms of performance skill for global weather forecasting. With increased rollout stability, the question of how these models perform for subseasonal to seasonal (S2S, week 3–8) forecasting has emerged. In this study we run a large set of subseasonal hindcasts over 2004–2023 to evaluate two ML weather forecast models at the S2S time scale, SFNO‐HENS (Nvidia, fully ML) and NeuralGCM (Google Research, hybrid). Corresponding hindcasts from the European Centre for Medium‐Range Weather Forecasts (ECMWF) are used as a baseline for comparison to a physics‐based model. Because our focus is on predicting moisture transport over the Western United States between October and March, we evaluate the models' prediction skill for the Madden‐Julian Oscillation (MJO) and its associated teleconnections in the North Pacific. We find that both ML models are competitive with the ECWMF model, with comparable skill in predicting the North Pacific large‐scale circulation and the MJO at week 3 and beyond. Even though overall the mid‐latitude subseasonal prediction skill remains low, the ML models exhibit interesting behavior such as a realistic propagation of the MJO across the Maritime Continent and realistic teleconnections. A SFNO‐HENS sensitivity experiment with altered initial conditions in the tropics demonstrates the stability of the model, and it illustrates the capability of ML models to represent important physical processes of the atmosphere at the S2S time scale. Plain Language Summary Predicting weather patterns and precipitation a few weeks in advance (subseasonal time scale) is of great interest for stakeholders such as water managers in the Southwest United States (US), where arid conditions prevail. Subseasonal forecasts from traditional weather forecast models exhibit low skill in the region, limiting their applicability. Here we examine whether the recent breakthrough in weather forecasting made with machine learning/artificial intelligence models can translate to improved subseasonal forecasts. Recently‐developed machine learning models exhibit comparable skill to a state‐of‐the‐art physics‐based model for predicting weather patterns in the North Pacific/North America region, and associated moisture transport. The same applies to their skill in predicting the tropical pattern, the Madden‐Julian Oscillation, and its important remote perturbations over the midlatitude East Pacific and Southwest US. Additionally, a perturbation experiment carried out with one of the machine learning models illustrates their ability to not only predict the evolution of atmospheric fields, but also to learn and represent physical processes such as tropics‐extratropics Rossby wave propagation. Key Points Two machine learning weather forecast models exhibit state‐of‐the‐art prediction skill at the subseasonal time scale in the Pacific sector The models equal ECWMF in terms of Madden‐Julian oscillation (MJO) prediction skill, and they accurately predict the MJO propagation and associated teleconnections The two machine‐learning models represent key physical processes for subseasonal prediction, despite being trained for weather forecasting

Peings, Yannick↗

Benchmarking core turbulence and transport predictions for an inductive compact tokamak reactor plasma

Motivated by the need for accurate, timely, and efficient calculations of plasma transport, predictions of plasma turbulence properties made using different TGLF saturation rules are benchmarked against corresponding predictions from linear and nonlinear gyrokinetic CGYRO simulations. This benchmarking is carried out using parameters taken from an inductive burning plasma scenario in a hypothetical compact high-field (R maj = 4 m, B T = 8 T) tokamak, lying in a much different regime of parameter space than either the TGLF calibration regime or current-day experiments. The core turbulent transport in this scenario is predicted to be dominated by ion temperature gradient (ITG) turbulence. In general, the ITG critical gradients predicted by various TGLF saturation rules are quite close to the CGYRO predictions. Both codes predict similar linear ITG growth rates and frequency spectra, as well as their scaling with R/L T i = −Rd ln(T i )/dr. However, TGLF systematically predicts unstable trapped-electron modes (TEMs) above k y ρ s ≃ 0.5 not seen by CGYRO for the same parameters, due to TGLF predicting a lower threshold in R/L T e than CGYRO for TEM onset. It is shown that for this scenario, nonlinear CGYRO simulations predict stiffer ITG turbulence than the TGLF SAT0 and SAT1 saturation rules, with energy fluxes close in magnitude and scaling with R/L T i to what is predicted by the SAT2 saturation rule. Self-consistent core profiles calculated using nonlinear CGYRO flux predictions and the PORTALS transport solver are shown to agree fairly well with corresponding predictions made using the TGLF SAT2 model, including a similar level of density peaking.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Mechanical loading prediction through accelerometry data during walking and running

ABSTRACT Currently, there is no way to assess mechanical loading variables such as peak ground reaction forces (pGRF) and peak loading rate (pLR) in clinical settings. The purpose of this study was to develop accelerometry‐based equations to predict both pGRF and pLR during walking and running. One hundred and thirty one subjects (79 females; 76.9 ± 19.6 kg) walked and ran at different speeds (2–14 km·h −1 ) on a force plate–instrumented treadmill while wearing accelerometers at their ankle, lower back and hip. Regression equations were developed to predict pGRF and pLR from accelerometry data. Leave‐one‐out cross‐validation was used to calculate prediction accuracy and Bland–Altman plots. Our pGRF prediction equation was compared with a reference equation previously published. Body mass and peak acceleration were included for pGRF prediction and body mass and peak acceleration rate for pLR prediction. All pGRF equation coefficients of determination were above 0.96, and a good agreement between actual and predicted pGRF was observed, with a mean absolute percent error (MAPE) below 7.3%. Accuracy indices from our equations were better than previously developed equations. All pLR prediction equations presented a lower accuracy compared to those developed to predict pGRF. Walking and running pGRF can be predicted with high accuracy by accelerometry‐based equations, representing an easy way to determine mechanical loading in free‐living conditions. The pLR prediction equations yielded a somewhat lower prediction accuracy compared with the pGRF equations.

Veras, Lucas↗

SU-F-R-04: Radiomics for Survival Prediction in Glioblastoma (GBM)

Purpose: To develop a quantitative radiomics approach for survival prediction of glioblastoma (GBM) patients treated with chemoradiotherapy (CRT). Methods: 28 GBM patients who received CRT at our institution were retrospectively studied. 255 radiomic features were extracted from 3 gadolinium-enhanced T1 weighted MRIs for 2 regions of interest (ROIs) (the surgical cavity and its surrounding enhancement rim). The 3 MRIs were at pre-treatment, 1-month and 3-month post-CRT. The imaging features comprehensively quantified the intensity, spatial variation (texture), geometric property and their spatial-temporal changes for the 2 ROIs. 3 demographics features (age, race, gender) and 12 clinical parameters (KPS, extent of resection, whether concurrent temozolomide was adjusted/stopped and radiotherapy related information) were also included. 4 Machine learning models (logistic regression (LR), support vector machine (SVM), decision tree (DT), neural network (NN)) were applied to predict overall survival (OS) and progression-free survival (PFS). The number of cases and percentage of cases predicted correctly were collected and AUC (area under the receiver operating characteristic (ROC) curve) were determined after leave-one-out cross-validation. Results: From univariate analysis, 27 features (1 demographic, 1 clinical and 25 imaging) were statistically significant (p<0.05) for both OS and PFS. Two sets of features (each contained 24 features) were algorithmically selected from all features to predict OS and PFS. High prediction accuracy of OS was achieved by using NN (96%, 27 of 28 cases were correctly predicted, AUC = 0.99), LR (93%, 26 of 28 cases were correctly predicted, AUC = 0.95) and SVM (93%, 26 of 28 cases were correctly predicted, AUC = 0.90). When predicting PFS, NN obtained the highest prediction accuracy (89%, 25 of 28 cases were correctly predicted, AUC = 0.92). Conclusion: Radiomics approach combined with patients’ demographics and clinical parameters can accurately predict survival in GBM patients treated with CRT.

60 APPLIED LIFE SCIENCES↗