Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “error vector”

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 379 records · Page 21
All sourcesNASA NTRSDOE OSTI
DOE OSTI2014

An initial study on the estimation of time-varying volumetric treatment images and 3D tumor localization from single MV cine EPID images

Purpose: In this work the authors develop and investigate the feasibility of a method to estimate time-varying volumetric images from individual MV cine electronic portal image device (EPID) images. Methods: The authors adopt a two-step approach to time-varying volumetric image estimation from a single cine EPID image. In the first step, a patient-specific motion model is constructed from 4DCT. In the second step, parameters in the motion model are tuned according to the information in the EPID image. The patient-specific motion model is based on a compact representation of lung motion represented in displacement vector fields (DVFs). DVFs are calculated through deformable image registration (DIR) of a reference 4DCT phase image (typically peak-exhale) to a set of 4DCT images corresponding to different phases of a breathing cycle. The salient characteristics in the DVFs are captured in a compact representation through principal component analysis (PCA). PCA decouples the spatial and temporal components of the DVFs. Spatial information is represented in eigenvectors and the temporal information is represented by eigen-coefficients. To generate a new volumetric image, the eigen-coefficients are updated via cost function optimization based on digitally reconstructed radiographs and projection images. The updated eigen-coefficients are then multiplied with the eigenvectors to obtain updated DVFs that, in turn, give the volumetric image corresponding to the cine EPID image. Results: The algorithm was tested on (1) Eight digital eXtended CArdiac-Torso phantom datasets based on different irregular patient breathing patterns and (2) patient cine EPID images acquired during SBRT treatments. The root-mean-squared tumor localization error is (0.73 ± 0.63 mm) for the XCAT data and (0.90 ± 0.65 mm) for the patient data. Conclusions: The authors introduced a novel method of estimating volumetric time-varying images from single cine EPID images and a PCA-based lung motion model. This is the first method to estimate volumetric time-varying images from single MV cine EPID images, and has the potential to provide volumetric information with no additional imaging dose to the patient.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Three-dimensional continuum point cloud method for large deformation and its verification

This study presents a strong form based meshfree collocation method, which is named Continuum Point Cloud Method, to solve nonlinear field equations derived from classical mechanics for deformed bodies in three-dimensional Euclidean space. The method and its implementation are benchmarked against a nonlinear vector field using manufactured solutions. The analysis of mechanical fields firstly focuses on the study of St. Venant Kirchhoff and compressible neo-Hookean materials. Results for various initial boundary value problems are presented, including benchmark cases involving unidirectional tension and simple shear. Subsequently, the study concludes with an analysis of a displacement-controlled simulation of a compressible neo-Hookean material, specifically a bar that is pulled to 50% of its original length and rotated 90°. The pure tension case yields a 1.5% error in displacement between computed and expected values and a combined tension and torsion loading case provides further insight into material behavior under complex loading conditions. The resulting normal axial and transverse stress-strain curves are also presented. Lastly, the consistency and robustness of the proposed nonlinear numerical schemes are successfully demonstrated through various numerical experiments.

Compressible neo-Hookean materials↗
DOE OSTI2011

Deformable Image Registration of Liver With Consideration of Lung Sliding Motion

Purpose: A feature based deformable registration model with sliding transformation was developed in the upper abdominal region for liver cancer. Methods: A two-step thin-plate spline (bi-TPS) algorithm was implemented to deformably register the liver organ. The first TPS registration was performed to exclusively quantify the sliding displacement component. A manual segmentation of the thoracic and abdominal cavity was performed as a priori knowledge. Tissue feature points were automatically identified inside the segmented contour on the images. The scale invariant feature transform method was utilized to match feature points that served as landmarks for the subsequent TPS registration to derive the sliding displacement vector field. To a good approximation, only motion along superior/inferior (SI) direction of voxels on each slice was averaged to obtain the sliding displacement for each slice. A second TPS transformation, as the last step, was carried out to obtain the local deformation field. Manual identification of bifurcation on liver, together with the manual segmentation of liver organ, was employed as a ''ground truth'' for assessing the algorithm's performance. Results: The proposed two-step TPS was assessed with six liver patients. The average error of liver bifurcation between manual identification and calculation for these patients was less than 1.8 mm. The residual errors between manual contour and propagated contour of liver organ using the algorithm fell in the range between 2.1 and 2.8 mm. An index of Dice similarity coefficient (DSC) between manual contour and calculated contour for liver tumor was 93.6% compared with 71.2% from the conventional TPS calculation. Conclusions: A high accuracy ({approx}2 mm) of the two-step feature based TPS registration algorithm was achievable for registering the liver organ. The discontinuous motion in the upper abdominal region was properly taken into consideration. Clinical implementation of the algorithm will find broad application in radiation therapy of liver cancer.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2014

RELAP5-3D Developer Guidelines and Programming Practices

Our ultimate goal is to create and maintain RELAP5-3D as the best software tool available to analyze nuclear power plants. This begins with writing excellent programming and requires thorough testing. This document covers development of RELAP5-3D software, the behavior of the RELAP5-3D program that must be maintained, and code testing. RELAP5-3D must perform in a manner consistent with previous code versions with backward compatibility for the sake of the users. Thus file operations, code termination, input and output must remain consistent in form and content while adding appropriate new files, input and output as new features are developed. As computer hardware, operating systems, and other software change, RELAP5-3D must adapt and maintain performance. The code must be thoroughly tested to ensure that it continues to perform robustly on the supported platforms. The coding must be written in a consistent manner that makes the program easy to read to reduce the time and cost of development, maintenance and error resolution. The programming guidelines presented her are intended to institutionalize a consistent way of writing FORTRAN code for the RELAP5-3D computer program that will minimize errors and rework. A common format and organization of program units creates a unifying look and feel to the code. This in turn increases readability and reduces time required for maintenance, development and debugging. It also aids new programmers in reading and understanding the program. Therefore, when undertaking development of the RELAP5-3D computer program, the programmer must write computer code that follows these guidelines. This set of programming guidelines creates a framework of good programming practices, such as initialization, structured programming, and vector-friendly coding. It sets out formatting rules for lines of code, such as indentation, capitalization, spacing, etc. It creates limits on program units, such as subprograms, functions, and modules. It establishes documentation guidance on internal comments. The guidelines apply to both existing and new subprograms. They are written for both FORTRAN 77 and FORTRAN 95. The guidelines are not so rigorous as to inhibit a programmer’s unique style, but do restrict the variations in acceptable coding to create sufficient commonality that new readers will find the coding in each new subroutine familiar. It is recognized that this is a “living” document and must be updated as languages, compilers, and computer hardware and software evolve.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2018

Assessing Alignment Between a Top-Down LCA Model and a Bottom-Up Supply Chain Model for Data Gap Identification and Prioritization

This work presents a method for evaluating the alignment between the results of bottom-up and top-down models and using the evaluation to enhance the bottom-up model and underlying data set. Disaggregation is used to enhance the embodied energy results of top-down models, which are then compared to bottom-up model results. Using the disaggregated top-down results makes it possible to identify specific data gaps and deficiencies within the bottom-up model and its data set. The top-down results will also provide information around which data gaps in the bottom-up model are contributing the most error to the results; in this way, data gaps can be identified as either critical or non-critical, and the effort needed to acquire and add data can be prioritized. Although the proposed method could be applied to any physical quantity or environmental impact that can be quantified with both bottom-up models (typically in physical units) and top-down models (typically in monetary units), embodied energy is the focus of this work. The disaggregated energy baseline is calculated from the throughput vector x, which gives the total economic activity required from each sector to provide a final demand. Using the direct requirements matrix, each element in x is disaggregated into a vector of inputs to the corresponding sector. Energy data from the interventions matrix B is then used to convert the monetary flows of the disaggregated x into physical energy flows. For each energy flow, the origin sector, the destination sector, and the energy carrier type will all be known. These energetic flows are analogous but not identical to direct energy inputs that would appear in a unit process used in life cycle analysis and other bottom-up methods, and thus can be used to evaluate the alignment between the top-down results and the results of a bottom-up model. To demonstrate and further develop the proposed method, embodied energy for chemicals assessed in the 2015 U.S. DOE bandwidth study on energy use within the chemical sector will be calculated using a top-down and a bottom-up model. The Materials Flows through Industry (MFI) modeling tool developed at the National Renewable Energy Laboratory is used as the bottom-up model, and the U.S. EEIO model, recently published by the U.S. Environmental Protection Agency, is used as the top-down model. U.S. EEIO and MFI results will be evaluated to determine how well the model results align. Where the two models are not in good alignment, the disaggregation method described above will be applied to the top-down model to generate a more detailed set of results that will aid in identifying data gaps in the MFI database. Using only MFI, data gaps can be identified in terms of which sectors have few or no unit processes in the database, but no information is available regarding the impact of the missing data on MFI results. The disaggregated top-down results will provide this information on which gaps in the MFI data are critical and have a significant impact on the results, and which are less important.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2024

Isovector Axial Charge and Form Factors of Nucleons from Lattice QCD

A survey of the calculations of the isovector axial vector form factor of the nucleon using lattice QCD is presented. Attention is paid to statistical and systematic uncertainties, in particular those due to excited state contributions. Based on a comparison of results from various collaborations, a case is made that lattice results are consistent within 10%. A similar level of uncertainty is in the axial charge gAu−d, the mean squared axial charge radius ⟨rA2⟩, the induced pseudoscalar charge gP∗, and the pion–nucleon coupling gπNN. Even with the current methodology, a significant reduction in errors is expected over the next few years with higher statistics data on more ensembles closer to the physical point. Lattice QCD results for the form factor GA(Q2) are compatible with those obtained from the recent MINERνA experiment but lie 2–3σ higher than the phenomenological extraction from the old ν–deuterium bubble chamber scattering data for Q2>0.3 GeV2. Current data show that the dipole ansatz does not have enough parameters to fit the form factor over the range 0≤Q2≤1 GeV2, whereas even a z2 truncation of the z expansion or a low order Padé are sufficient. Looking ahead, lattice QCD calculations will provide increasingly precise results over the range 0≤Q2≤1 GeV2, and MINERνA-like experiments will extend the range to Q2∼2 GeV2 or higher. Nevertheless, improvements in lattice methods to (i) further control excited state contributions and (ii) extend the range of Q2 are needed.

Astronomy & Astrophysics↗
DOE OSTI2012

THE M31 VELOCITY VECTOR. I. HUBBLE SPACE TELESCOPE PROPER-MOTION MEASUREMENTS

We present the first proper-motion (PM) measurements for the galaxy M31. We obtained new V-band imaging data with the Hubble Space Telescope ACS/WFC and the WFC3/UVIS instruments of three fields: a spheroid field near the minor axis, an outer disk field along the major axis, and a field on the Giant Southern Stream. The data provide five to seven year time baselines with respect to pre-existing deep first-epoch observations of the same fields. We measure the positions of thousands of M31 stars and hundreds of compact background galaxies in each field. High accuracy and robustness is achieved by building and fitting a unique template for each individual object. The average PM for each field is obtained from the average motion of the M31 stars between the epochs with respect to the background galaxies. For the three fields, the observed PMs ({mu}{sub W}, {mu}{sub N}) are, in units of mas yr{sup -1}, (- 0.0458, -0.0376) {+-} (0.0165, 0.0154), (- 0.0533, -0.0104) {+-} (0.0246, 0.0244), and (- 0.0179, -0.0357) {+-} (0.0278, 0.0272), respectively. The ability to average over large numbers of objects and over the three fields yields a final displacement accuracy of a few thousandths of a pixel, corresponding to only 12 {mu}as yr{sup -1}. This is comparable to what has been achieved for other Local Group galaxies using Very Long Baseline Array observations of water masers. Potential systematic errors are controlled by an analysis strategy that corrects for detector charge transfer inefficiency, spatially and time-dependent geometric distortion, and point-spread function variations. The robustness of the PM measurements and uncertainties are supported by the fact that data from different instruments, taken at different times and with different telescope orientations, as well as measurements of different fields, all yield statistically consistent results. Papers II and III of this series explore the implications of the new measurements for our understanding of the history, future, and mass of the Local Group.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Situational awareness-enhancing community-level load mapping with opportunistic machine learning

Motivated by present and forthcoming challenges in the adoption and integration of distributed renewable energy, we develop a machine learning (ML) approach that builds short-fuse mappings connecting the occasionally-unobservable true load in one target community with information-rich signals collected from relatively more instrumented reference communities. Our setting is inspired by and tailored to target communities with significant unobservable behind-the-meter solar generation, where true load (a relatively well-behaved quantity of interest to grid operators) is hard to discern during daytime due to insufficient instrumentation and/or privacy reasons, but that can be related to reference communities with low unobservable distributed variable generation or with sufficient instrumentation. The developed mapping, herein realized with Support Vector Machine regression, is built using nighttime data from all communities, when their distributed generation is low or zero. Our ML algorithm opportunistically learns to correlate signals of interest and then is operationally used the next day to shed light into target community load evolution. The mapping is subsequently rebuilt, rolling its short-fuse scope perpetually forward in time. Here, we demonstrate the efficacy of our approach on nine synthetically generated topologies and associated timeseries stemming from real-world data, on which we observe cumulative error performance that yields lower than 10% and 15% daily-averaged mean absolute percentage errors in target community load estimation on more than about 75% and 90% of days, respectively, in multiple yearly evaluations that shed light on long-term performance also under seasonal and one-off effects. The proposed ML-powered methodology can offer grid operators much-improved visibility into a previously obscure space and can also serve as an additional source of information in broader, multi-modal solar disaggregation solutions.

14 SOLAR ENERGY↗
DOE OSTI2023

Exploiting Kronecker structure in exponential integrators: Fast approximation of the action of φ $-$functions of matrices via quadrature

Here, in this article, we propose an algorithm for approximating the action of φ $-$ functions of matrices against vectors, which is a key operation in exponential time integrators. In particular, we consider matrices with Kronecker sum structure, which arise from problems admitting a tensor product representation. The method is based on quadrature approximations of the integral form of the φ $-$ functions combined with a scaling and modified squaring method. Owing to the Kronecker sum representation, only actions of 1D matrix exponentials are needed at each quadrature node and assembly of the full matrix can be avoided. Additionally, we derive a priori bounds for the quadrature error, which show that, as expected by classical theory, the rate of convergence of our method is supergeometric. Guided by our analysis, we construct a fast and robust method for estimating the optimal scaling factor and number of quadrature nodes that minimizes the total cost for a prescribed error tolerance. We investigate the performance of our algorithm by solving several linear and semilinear time-dependent problems in 2D and 3D. The results show that our method is accurate and orders of magnitude faster than the current state-of-the-art.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Alignment of a vector magnetometer to an optical prism

A method for alignment of a vector magnetometer to a rigidly attached prism is presented. This enables optical comparison of the magnetometer axes to physical surfaces in the apparatus, and thus an absolute determination of the magnetic field direction in space. This is in contrast with more common techniques, which focus on precise determination of the relative angles between magnetometer axes, and so are more suited to measuring differences in the direction of magnetic fields. Here we demonstrate precision better than 500 μrad on a fluxgate magnetometer, which also gives the coil orthogonality errors to a similar precision. As a result, the relative sensitivity of the three axes is also determined, with a precision of about 5 ×10 –4 .

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2017

Image phase shift invariance based cloud motion displacement vector calculation method for ultra-short-term solar PV power forecasting

Irradiance received on the earth's surface is the main factor that affects the output power of solar PV plants, and is chiefly determined by the cloud distribution seen in a ground-based sky image at the corresponding moment in time. It is the foundation for those linear extrapolation-based ultra-short-term solar PV power forecasting approaches to obtain the cloud distribution in future sky images from the accurate calculation of cloud motion displacement vectors (CMDVs) by using historical sky images. Theoretically, the CMDV can be obtained from the coordinate of the peak pulse calculated from a Fourier phase correlation theory (FPCT) method through the frequency domain information of sky images. The peak pulse is significant and unique only when the cloud deformation between two consecutive sky images is slight enough, which is likely possible for a very short time interval (such as 1?min or shorter) with common changes in the speed of cloud. Sometimes, there will be more than one pulse with similar values when the deformation of the clouds between two consecutive sky images is comparatively obvious under fast changing cloud speeds. This would probably lead to significant errors if the CMDVs were still only obtained from the single coordinate of the peak value pulse. However, the deformation estimation of clouds between two images and its influence on FPCT-based CMDV calculations are terrifically complex and difficult because the motion of clouds is complicated to describe and model. Therefore, to improve the accuracy and reliability under these circumstances in a simple manner, an image-phase-shift-invariance (IPSI) based CMDV calculation method using FPCT is proposed for minute time scale solar power forecasting. First, multiple different CMDVs are calculated from the corresponding consecutive images pairs obtained through different synchronous rotation angles compared to the original images by using the FPCT method. Second, the final CMDV is generated from all of the calculated CMDVs through a centroid iteration strategy based on its density and distance distribution. Third, the influence of different rotation angle resolution on the final CMDV is analyzed as a means of parameter estimation. Simulations under various scenarios including both thick and thin clouds conditions indicated that the proposed IPSI-based CMDV calculation method using FPCT is more accurate and reliable than the original FPCT method, optimal flow (OF) method, and particle image velocimetry (PIV) method.

14 SOLAR ENERGY↗
DOE OSTI2020

Unified model of nucleon elastic form factors and implications for neutrino-oscillation experiments

Precise knowledge of the nucleon’s axial-current form factors is crucial for modeling GeV-scale neutrino-nucleus interactions. Unfortunately, the axial form factor remains insufficiently constrained to meet the precision requirements of upcoming long-baseline neutrino-oscillation experiments. This work studies the nucleon’s axial and vector form factors using the light-front approach to build a quark-diquark model of the nucleon with an explicit pion cloud. The light-front wave functions in both the quark and pion-baryon Fock spaces are first calibrated to existing experimental information on the nucleon’s electromagnetic form factors and then used to predict the axial form factor. The resulting squared charge radius of the axial pseudovector form factor is predicted to be $r$$^{2}_{A}$ = 0.29 ± 0.03 fm 2 , where the small error accounts for the model’s parametric uncertainty. We use our form factor results to explore the (quasi)elastic scattering of neutrinos by (nuclei)nucleons, with the result that the widely implemented dipole ansatz is an inadequate approximation of the full form factor for modeling both processes. Here, the approximation leads to a 5%–10% overestimation of the total cross section, depending on the (anti)neutrino energy. We project overestimations of similar size in the flux-averaged cross sections for the upcoming DUNE long-baseline neutrino-oscillation experiment.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2014

Forecasting longitudinal changes in oropharyngeal tumor morphology throughout the course of head and neck radiation therapy

Purpose: To create models that forecast longitudinal trends in changing tumor morphology and to evaluate and compare their predictive potential throughout the course of radiation therapy. Methods: Two morphology feature vectors were used to describe 35 gross tumor volumes (GTVs) throughout the course of intensity-modulated radiation therapy for oropharyngeal tumors. The feature vectors comprised the coordinates of the GTV centroids and a description of GTV shape using either interlandmark distances or a spherical harmonic decomposition of these distances. The change in the morphology feature vector observed at 33 time points throughout the course of treatment was described using static, linear, and mean models. Models were adjusted at 0, 1, 2, 3, or 5 different time points (adjustment points) to improve prediction accuracy. The potential of these models to forecast GTV morphology was evaluated using leave-one-out cross-validation, and the accuracy of the models was compared using Wilcoxon signed-rank tests. Results: Adding a single adjustment point to the static model without any adjustment points decreased the median error in forecasting the position of GTV surface landmarks by the largest amount (1.2 mm). Additional adjustment points further decreased the forecast error by about 0.4 mm each. Selection of the linear model decreased the forecast error for both the distance-based and spherical harmonic morphology descriptors (0.2 mm), while the mean model decreased the forecast error for the distance-based descriptor only (0.2 mm). The magnitude and statistical significance of these improvements decreased with each additional adjustment point, and the effect from model selection was not as large as that from adding the initial points. Conclusions: The authors present models that anticipate longitudinal changes in tumor morphology using various models and model adjustment schemes. The accuracy of these models depended on their form, and the utility of these models includes the characterization of patient-specific response with implications for treatment management and research study design.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

Strong coupling from hadronic τ -decay data including τ → π − π 0 ν τ from Belle

In previous work we have combined the π − π 0 , 2 π − π + π 0 , and π − 3 π 0 spectral data obtained from hadronic τ decays measured by the ALEPH and OPAL experiments, together with electroproduction data for several of the subleading hadronic modes and data for the K K ¯ mode to construct an inclusive nonstrange vector spectral function entirely based on experimental data, with no Monte-Carlo generated input. In this paper, we include, for the first time, the Belle τ → π − π 0 ν τ high-statistics decay data to construct a new inclusive nonstrange vector spectral function that combines more of the world’s available data. As no Belle data are at present available for the two 4 π modes, this requires a revised data analysis in comparison with our previous work. From the resulting new spectral function, we obtain a new determination of the strong coupling, α s , using our previously developed strategy based on finite-energy sum rules. We find, at the Z mass scale, α s ( m Z 2 ) = 0.1159 ( 14 ) . We discuss the smaller central value and larger error of our new result compared to our previous result, showing the shifts to be due mainly to significant changes in updated HFLAV results for the π − 3 π 0 decay mode. Published by the American Physical Society 2025

Boito, Diogo (ORCID:0000000244267984)↗
DOE OSTI2022

Motion tomography via occupation kernels

The goal of motion tomography is to recover a description of a vector flow field using measurements along the trajectory of a sensing unit. In this paper, we develop a predictor corrector algorithm designed to recover vector flow fields from trajectory data with the use of occupation kernels developed by Rosenfeld et al. [9,10]. Specifically, we use the occupation kernels as an adaptive basis; that is, the trajectories defining our occupation kernels are iteratively updated to improve the estimation in the next stage. Initial estimates are established, then under mild assumptions, such as relatively straight trajectories, convergence is proven using the Contraction Mapping Theorem. We then compare the developed method with the established method by Chang et al. [5] by defining a set of error metrics. Here, we found that for simulated data, where a ground truth is available, our method offers a marked improvement over [5]. For a real-world example, where ground truth is not available, our results are similar results to the established method.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Enabling probabilistic learning on manifolds through double diffusion maps

Here, we present a generative learning framework for probabilistic sampling that extends Probabilistic Learning on Manifolds (PLoM), which is designed to generate statistically consistent realizations of a random vector in a finite-dimensional Euclidean space, informed by a (representative) set of observations. In its original form, PLoM constructs a reduced-order probabilistic model by combining three main components: (a) kernel density estimation to approximate the underlying probability measure, (b) Diffusion Maps to characterize the manifold of the data, and (c) a reduced-order Itô Stochastic Differential Equation (ISDE) to sample from the learned distribution. However, its sampling dynamics are posed in the ambient space and the retained number of reduced coordinates is chosen by projection-reconstruction error. In practice, this often (i) requires more coordinates than the data’s intrinsic dimension to achieve stable sampling and (ii) lacks a smooth, basis-independent lifting back to the data domain; moreover, standard Diffusion Maps emphasize harmonic eigenfunctions and can miss non-harmonic latent structure. We address these limitations by decoupling geometry learning from sampling: a first Diffusion Maps pass identifies non-harmonic coordinates on which we formulate a full-order ISDE directly in the latent space, while Double Diffusion Maps captures multiscale geometric features and Geometric Harmonics (GH) learns a smooth lifting map to the ambient variables that is independent of the particular diffusion basis. This hybrid design preserves the system’s dynamical richness with a compact geometric representation and enables principled out-of-sample inference. The effectiveness and robustness of the proposed method are illustrated through two numerical studies: one based on data generated from two-dimensional Hermite polynomial functions and another based on high-fidelity simulations of a detonation wave in a reactive flow.

Double diffusion maps↗
DOE OSTI2022

Stream Temperature Predictions for River Basin Management in the Pacific Northwest and Mid-Atlantic Regions Using Machine Learning

Stream temperature (Ts) is an important water quality parameter that affects ecosystem health and human water use for beneficial purposes. Accurate Ts predictions at different spatial and temporal scales can inform water management decisions that account for the effects of changing climate and extreme events. In particular, widespread predictions of Ts in unmonitored stream reaches can enable decision makers to be responsive to changes caused by unforeseen disturbances. In this study, we demonstrate the use of classical machine learning (ML) models, support vector regression and gradient boosted trees (XGBoost), for monthly Ts predictions in 78 pristine and human-impacted catchments of the Mid-Atlantic and Pacific Northwest hydrologic regions spanning different geologies, climate, and land use. The ML models were trained using long-term monitoring data from 1980–2020 for three scenarios: (1) temporal predictions at a single site, (2) temporal predictions for multiple sites within a region, and (3) spatiotemporal predictions in unmonitored basins (PUB). In the first two scenarios, the ML models predicted Ts with median root mean squared errors (RMSE) of 0.69–0.84 °C and 0.92–1.02 °C across different model types for the temporal predictions at single and multiple sites respectively. For the PUB scenario, we used a bootstrap aggregation approach using models trained with different subsets of data, for which an ensemble XGBoost implementation outperformed all other modeling configurations (median RMSE 0.62 °C).The ML models improved median monthly Ts estimates compared to baseline statistical multi-linear regression models by 15–48% depending on the site and scenario. Air temperature was found to be the primary driver of monthly Ts for all sites, with secondary influence of month of the year (seasonality) and solar radiation, while discharge was a significant predictor at only 10 sites. The predictive performance of the ML models was robust to configuration changes in model setup and inputs, but was influenced by the distance to the nearest dam with RMSE <1 °C at sites situated greater than 16 and 44 km from a dam for the temporal single site and regional scenarios, and over 1.4 km from a dam for the PUB scenario. Our results show that classical ML models with solely meteorological inputs can be used for spatial and temporal predictions of monthly Ts in pristine and managed basins with reasonable (<1 °C) accuracy for most locations.

54 ENVIRONMENTAL SCIENCES↗