SEARCH · Engineering Papers

Results for “gradient methods”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,405 records · Page 78

All sources NASA NTRS DOE OSTI

DOE OSTI2022

Effect of Initial Water Saturation on The Performance of Fracturing Fluids With and Without Polyallylamine under Simulated EGS Conditions

Objectives/Scope: StimuFrac (US Patents 9,873,828 B2 and 9,447,315 B2), a CO2-reactive polymer aqueous solution [polyallylamine (PAA) 1wt% in water] combined with CO2, can be used as a less water-intensive fracturing fluid for enhanced geothermal systems (EGS). Our previous results show that in hot dry rock (HDR), PAA/CO2 fracturing fluids outperformed other fluids such as water, CO2, and CO2/water in generating large fractures with less fluid consumed. The objective of this work is to study the effect of initial water saturation of rock on the performance of StimuFrac fluid in ½ foot cubic rock samples and under representative EGS pressure/temperature conditions using cyclic and constant flow rate injection strategies. The fracturing results are compared with results using different fracturing fluids in terms of controlling fracture propagation rates, fracture hydraulic conductivity, breakdown pressures and volumes of fluids required. Methods/Procedures/Process: In all tests, water was initially injected into the rock to increase the water saturation before the fracturing processes to simulate actual geothermal reservoir conditions. For the cyclic injection, one complete cycle consisted of (1) a PAA slug (or water slug) injection followed by (2) CO2 injection to initiate the fracture. In the second step of the first cycle, the pressure of CO2 is increased until a maximum pressure is reached (fracture is initiated at this moment), and then the injection of CO2 is allowed for another 30 seconds to propagate the fracture. Then, the two-step cycle of PAA followed by CO2 injection (up to 2-4 mL/min) was continued. For the constant flow rate injection strategies, the initial water saturation was increased by injecting water at 1000 psi and 200°C for three days. After that, an initial slug of water, CO2, or PAA was injected and then fracturing was initiated and propagated by injecting CO2 at a constant flow rate. Applications/Significance/Novelty: The results of this study suggest that water saturation, especially near the wellbore region, will significantly affect the fracturing fluid transmission into the rock porous media by changing the relative permeability of CO2 or water, thus affecting the fracture initiation and propagation. In this study, fracturing with cyclic injection or constant flow rate injection methods were performed using three different kinds of fluids systems. These fluids are water, CO2, or CO2 with PAA. Splitting the rock samples in half after fracturing reveals that the fracture propagation is significantly limited under these high water saturated conditions compared to dry initial conditions: The fractures propagate less than 1/3 length of the distance from the wellbore to rock surface, and in some cases no fracture is generated. This may be caused by the fact that leak-off is dominating the fracturing process and the injected fluid flow rate is not high enough to overcome the leak-off even under high flow rate injection conditions. Additionally, CO2 could be leaking off into the wellbore annulus and this may be making it more difficult to generate pressure gradients away from the near-wellbore region.

Jian, Guoqing↗

DOE OSTI2024

Super Resolving Unrolled Neural Networks for Remote Sensing

In remote sensing systems, the capabilities of the system are constrained by the complex interactions between size, weight, and power (SWAP) of potential designs. In electro-optical (EO) systems, examples of these critical parameters include the system’s sensitivity and resolution. Those parameters can be increased by ever larger optical apertures and focal planes but at the cost of more SWAP. Multi-image super resolution (MISR) techniques allow resolution to be enhanced via computation rather than more sophisticated optical hardware. These algorithms combine multiple images together into a single, higher resolution image, trading temporal resolution and computation for spatial resolution. Fielded MISR techniques, such as Drizzle, can require several hundred images to create a single super resolved image, implying reduced temporal resolution, increased data acquisition load, and limiting mission applications. Iterative techniques, such as model-based image reconstruction and compressive sensing, have been shown to create super resolved images using fewer images than Drizzle. They do this by posing an optimization problem that balances accuracy between a highly accurate physical model and an image model. In the case of super resolution, the physical model is defined by the relation between low resolution input images and the desired high resolution output image. The image model encodes some assumptions about the super resolved image. These assumptions are meant to suppress reconstruction artifacts that arise due to deterministic physical model error, stochastic measurement noise, and potential undersampling. In practice, the performance of iterative methods are limited by imaging models compatible with optimization. Deep learning-based methods can effectively learn image models of arbitrary complexity, but lack the theoretical explainability and robustness of iterative techniques. Consensus equilibrium (CE) generalizes the iterative techniques beyond optimization, enabling blackbox algorithms such as traditional and neural image denoisers to be used as the image model. CE-based approaches retain much of the explainability and robustness of iterative techniques while allowing the expressiveness of machine learning image models to be used. Additionally, by unrolling iterations of CE with an embedded image denoiser, the image denoiser can be further trained and specialized to the specific application with potentially higher quality reconstructions. Under this project, we demonstrated the feasibility of training an unrolled neural network based upon CE. While we didn’t train one, we showed that the CE process is differentiable and its gradient can be tractably computed. We also explored the usage of a variants of CE akin to generative neural works. Most importantly, we applied the CE framework to a number of problems including non-blind deconvolution, upsampling, single-image super resolution, MISR, event-based sensing, and saturated deconvolution. Our MISR prototype creates high quality reconstructions with an order of magnitude fewer images than previous approaches and, critically, produces these reconstructions fast enough for practical usage.

47 OTHER INSTRUMENTATION↗

DOE OSTI2024

Development of an Optical Library for Coevaporated CdSe x Te 1– x

The conversion efficiency of CdTe solar cells may be improved by bandgap engineering, i.e., changing the bandgap value through the addition of Se in the absorber. The Se alloying enables a short-circuit current density improvement, as it leads to a bandgap energy value decrease. Furthermore, it has been associated with increased minority carrier lifetimes, assuring high open-circuit voltage values. An Se gradient profile control can further optimize the solar cell performance. Thus, an optical model baseline of the CdSe x Te 1–x (CST) compound was developed. Spectroscopic ellipsometry measurements were conducted to accurately extract the optical constants of ten CST layers deposited through coevaporation with x varying from 0 to 1. Using the measured dielectric function spectra from the discrete CST layers with varying x, and considering the composition-induced shift in the critical point energies, an energy-shift model was employed to develop the accurate optical library for the CST compound for any x value to provide data for future modeling and optimization. Furthermore, the library accuracy was validated through optical simulations of the quantum efficiency of a graded CST solar cell using the finite-difference time-domain method by replicating the Se profile in the absorber layer measured through secondary ion mass spectrometry.

14 SOLAR ENERGY↗

DOE OSTI2026

Superconducting Material Growth for Radio-Frequency (RF) Cavities

Superconducting radio-frequency (SRF) cavities, usually manufactured from Niobium (Nb), are vital components of modern particle accelerators because of their ability to achieve high acceleration gradients with little power dissipation. Naturally forming Nb surface oxides significantly alter cavity performance by changing surface resistance. A nondestructive characterization of oxide thickness is helpful for relating surface processing treatments to cavity performance. This work develops a protocol to measure Nb oxide thickness using Angle-Resolved X-ray Photoelectron Spectroscopy (ARXPS) while correcting instrumental errors. Using uniform bulk standard samples (Silver, Aluminum oxide, and Germanium), we determined a baseline correction factor to account for analyzer-related intensity evolution as the measurement angle increases. The correction factor was then applied to Nb 3d ARXPS data. Applying the Strohmeier equation to the corrected data yielded a Nb2O5 thickness of 6.04 nm, closely matching the 5.5 (±.05) nm value obtained from cross-sectional transmission electron microscopy. Our approach will bring a method to incorporate inherent errors in the thickness measurements using ARXPS and can be broadly applied to improve the accuracy of thickness measurements in a wide range of heterostructures.

Lambert, Nathan [Fermilab]↗

DOE OSTI2026

Superconducting Material Growth for Radio-Frequency (RF) Cavities

Lambert, Nathan [Fermilab]↗

DOE OSTI2024

Residual Stress in Cold Spray SS304L Measured Via Neutron Diffraction and Comparison of Analytical Models to Predict the Residual Stress

Here, this study employs neutron diffraction to investigate the relationship between residual stress and coating thickness in cold sprayed 304L austenitic stainless steel. Results show that shot peening predominantly impacts the residual stress profile, leading to substantial in-plane compressive force. The impact of laser heating, a widely used method to alter cold spray's microstructural properties, on the coating's residual stress is also analyzed. The findings indicate that the maximum compressive residual stress in the in-plane component is mainly independent of coating thickness, which suggests that the material properties determine the maximum residual stress. The cold sprayed deposits possessed compressive, nearly biaxial strain and stresses. After laser heating, these stresses were replaced by tensile residual stresses. Two analytical models, the Tsui and Clyne and the Boruah models, for predicting residual stresses are also evaluated, and both models provide reasonable fits to the experimental data. At this point, the deviations between the experimental results and the models are principally caused by the inability of the current models to address plastic deformation and relaxation, and the residual stresses generated by thermal gradients.

36 MATERIALS SCIENCE↗

DOE OSTI2020

Computationally efficient CFD prediction of bubbly flow using physics-guided deep learning

To realize efficient computational fluid dynamics (CFD) prediction of two-phase flow, a multi-scale framework was proposed in this paper by applying a physics-guided data-driven approach. Instrumental to this framework, Feature Similarity Measurement (FSM) technique was developed for error estimation in two-phase flow simulation using coarse-mesh CFD, to achieve a comparable accuracy as fine-mesh simulations with fast-running feature. In this work, by defining physics-guided parameters and variable gradients as physical features, FSM has the capability to capture the underlying local patterns in the coarse-mesh CFD simulation. Massive low-fidelity data and respective high-fidelity data are used to explore the underlying information relevant to the main simulation errors and the effects of phenomenological scaling. By learning from previous simulation data, a surrogate model using deep feedforward neural network (DFNN) can be developed and trained to estimate the simulation error of coarse-mesh CFD. In a demonstration case of two-phase bubbly flow, the DFNN model well captured and corrected the unphysical “peaks” in the velocity and void fraction profiles near the wall in the coarse-mesh configuration, even for extrapolative predictions. The research documented supports the feasibility of the physics-guided deep learning methods for coarse mesh CFD simulations which has a potential for the efficient industrial design.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

DOE OSTI2020

DEMONSTRATION OF A DATA-DRIVEN PHYSICS-BASED APPROACH FOR COMPUTATIONALLY EFFICIENT CFD PREDICTION OF TWO-PHASE BUBBLY FLOW

To realize efficient computational fluid dynamics (CFD) prediction of two-phase flow, a multi-scale physics-guided data-driven approach, Feature Similarity Measurement (FSM) technique was developed for error estimation in two-phase flow simulation using coarse-mesh CFD, to achieve a comparable accuracy as fine-mesh simulations with fast-running feature. By defining physics-guided parameters and variable gradients as physical features, FSM has the capability to capture the underlying local patterns in the coarse-mesh CFD simulation. Massive low-fidelity data and respective high-fidelity data are used to explore the underlying information relevant to the main simulation errors and the effects of phenomenological scaling. By learning from previous simulation data, a surrogate model using deep feedforward neural network (DFNN) can be developed and trained to estimate the simulation error of coarse-mesh CFD. In a demonstration case of two-phase bubbly flow, the DFNN model well captured and corrected the unphysical “peaks” in the velocity and void fraction profiles near the wall in the coarse-mesh configuration, even for extrapolative predictions. The research documented supports the feasibility of the physics-guided deep learning methods for coarse mesh CFD simulations which has a potential for the efficient industrial design.

42 ENGINEERING↗

DOE OSTI2020

First-principles study of intrinsic point defects and Xe impurities in uranium monocarbide

Based on density functional theory (DFT) calculations, we perform an extensive investigation of intrinsic point defects and Xe impurities in uranium monocarbide (UC). The DFT calculations involve both the conventional generalized gradient approximation (GGA) and the GGA+U approach with the Hubbard parametric term (U), using up to 5×5×5 supercells. GGA calculations for the formation energy of intrinsic defects demonstrate the significant effect of using larger supercells than in previous studies. Results confirm that the ⟨111⟩ and ⟨100⟩ dumbbell interstitials are the most stable interstitial configurations for U and C, respectively. The interstitial mechanisms are favored for self-diffusion of both uranium and carbon and diffusion of Xe under equilibrium conditions. Calculations also reveal that the Xe substitutional defect at the C lattice site tends to adopt an off-site configuration, which can be interpreted as a Xe interstitial–C vacancy complex. We also utilize GGA+U to assess the impact of effective U parameter (Ueff) on the results. Moreover, we introduce a method to estimate the carbon chemical potential by fitting the phase diagram composition data and propose a selection Ueff=1.25 eV based on the experimental Xe diffusion activation energy. With this approach, GGA+U calculations reproduce the available experimental data for the formation energy of the carbon Frenkel pair and can explain the stepwise recovery of intrinsic properties and burst Xe release behavior in UC observed in annealing experiments.

Huang, Gui-Yang (ORCID:0000000301447376)↗

DOE OSTI2019

Application and Evaluation of Surrogate Models for Radiation Source Search

Surrogate models are increasingly required for applications in which first-principles simulation models are prohibitively expensive to employ for uncertainty analysis, design, or control. They can also be used to approximate models whose discontinuous derivatives preclude the use of gradient-based optimization or data assimilation algorithms. We consider the problem of inferring the 2D location and intensity of a radiation source in an urban environment using a ray-tracing model based on Boltzmann transport theory. Whereas the code implementing this model is relatively efficient, extension to 3D Monte Carlo transport simulations precludes subsequent Bayesian inference to infer source locations, which typically requires thousands to millions of simulations. Additionally, the resulting likelihood exhibits discontinuous derivatives due to the presence of buildings. To address these issues, we discuss the construction of surrogate models for optimization, Bayesian inference, and uncertainty propagation. Specifically, we consider surrogate models based on Legendre polynomials, multivariate adaptive regression splines, radial basis functions, Gaussian processes, and neural networks. We detail strategies for computing training points and discuss the merits and deficits of each method.

Cook, Jared A.↗

DOE OSTI2024

Training quantum neural networks using the quantum information bottleneck method

Abstract We provide in this paper a concrete method for training a quantum neural network to maximize the relevant information about a property that is transmitted through the network. This is significant because it gives an operationally well founded quantity to optimize when training autoencoders for problems where the inputs and outputs are fully quantum. We provide a rigorous algorithm for computing the value of the quantum information bottleneck quantity within error ε that requires O ( log 2 ⁡ ( 1 / ϵ ) + 1 / δ 2 ) queries to a purification of the input density operator if its spectrum is supported on { 0 } ⋃ [ δ , 1 − δ ] for δ > 0 and the kernels of the relevant density matrices are disjoint. We further provide algorithms for estimating the derivatives of the QIB function, showing that quantum neural networks can be trained efficiently using the QIB quantity given that the number of gradient steps required is polynomial.

Çatlı, Ahmet Burak (ORCID:0000000152294141)↗

DOE OSTI2021

A machine learning aided interpretable model for rupture strength prediction in Fe-based martensitic and austenitic alloys

Abstract The class of 9–12% Cr ferritic-martensitic alloys (FMA) and austenitic stainless steels have received considerable attention due to their numerous applications in high temperature power generation industries. To design high strength steels with prolonged service life requires a thorough understanding of the long-term properties, e.g., creep rupture strength, rupture life, etc., as a function of the chemical composition and processing parameters that govern the microstructural characteristics. In this article, the creep rupture strength of both 9–12% Cr FMA and austenitic stainless steel has been parameterized using curated experimental datasets with a gradient boosting machine. The trained model has been cross validated against unseen test data and achieved high predictive performance in terms of correlation coefficient ( $$R^{2} > 0.98 $$ R 2 > 0.98 for 9–12% Cr FMA and $$R^{2} > 0.95 $$ R 2 > 0.95 for austenitic stainless steel) thus bypassing the need for additional comprehensive tensile test campaigns or physical theoretical calculations. Furthermore, the feature importance has been computed using the Shapley value analysis to understand the complex interplay of different features.

36 MATERIALS SCIENCE↗

DOE OSTI2025

A Measurement of Atmospheric Circular Polarization with POLARBEAR

At millimeter wavelengths, the atmospheric emission is circularly polarized owing to the Zeeman splitting of molecular oxygen by the Earth's magnetic field. We report a measurement of the signal in the 150 GHz band using 3 yr of observational data with the POLARBEAR project. Nonidealities of a continuously rotating half-wave plate (HWP) partially convert circularly polarized light to linearly polarized light. While POLARBEAR detectors are sensitive to linear polarization, this effect makes them sensitive to circular polarization. Although this was not the intended use, we utilized this conversion to measure circular polarization. We reconstruct the azimuthal gradient of the circular polarization signal and measure its dependency from the scanning direction and the detector bandpass. We compare the signal with a simulation based on atmospheric emission theory, the detector bandpass, and the HWP leakage spectrum model. We find the ratio of the observed azimuthal slope to the simulated slope is 0.92 ± 0.01(stat) ± 0.07(sys). This ratio corresponds to a brightness temperature of 3.8 mK at the effective band center of 121.8 GHz and bandwidth of 3.5 GHz estimated from representative detector bandpass and the spectrum of Zeeman emission. This result validates our understanding of the instrument and reinforces the feasibility of measuring the circular polarization using the imperfection of continuously rotating HWP. Continuously rotating HWP is popular in ongoing and future cosmic microwave background experiments to modulate the polarized signal. This work shows a method for signal extraction and leakage subtraction that can help measure circular polarization in such experiments.

79 ASTRONOMY AND ASTROPHYSICS↗

DOE OSTI2025

Enhanced Boundary Layer Height Detection Using Ceilometer, Surface Meteorology, and Radiation Products With a Random Forest Ensemble Method

This study develops and evaluates a Random Forest (RF) model for estimating planetary boundary layer height (PBLH) using 9 years of data from the Atmospheric Radiation Measurement Southern Great Plains (ARM SGP) user facility, with potential application in the NOAA Surface Radiation (SURFRAD) Network. The model integrates ceilometer, surface meteorology, and radiation measurements, and is trained using thermodynamic PBLH estimates derived from radiosondes. This approach aims to bridge gaps between aerosol-based and thermodynamic-based PBLH estimates. The RF model outperformed traditional methods during daytime and better captured transition periods, demonstrating improved accuracy and robustness. At ARM SGP, it showed a substantial reduction in both bias and RMSE, with a bias near zero (−4.9 m) compared with traditional Haar Wavelet (HW) (70.9 m) and Vaisala BL-View software (124.1 m), and an RMSE of 303.2 m, lower than both BL-View (566.9 m) and HW (404.6 m). During daytime hours, RF consistently outperformed both alternatives, maintaining lower bias and RMSE across all periods. At a second evaluation site, RF achieved the lowest overall RMSE (323.7 m), similar to HW (326.4 m) and significantly better than BL-View (738.3 m). However, all models showed reduced accuracy under stable nighttime conditions, limiting the reliability of PBLH estimates. Key predictors for the model included the lifting condensation level height (LCLH), aerosol gradients, and month for seasonal variability. The study underscores the potential of integrating machine learning with multiple data sets such as surface energy and thermodynamic data to advance PBLH estimation.

boundary layer height↗

DOE OSTI2023

Using Ultrasound Image Augmentation and Ensemble Predictions to Prevent Machine-Learning Model Overfitting

Deep learning predictive models have the potential to simplify and automate medical imaging diagnostics by lowering the skill threshold for image interpretation. However, this requires predictive models that are generalized to handle subject variability as seen clinically. Here, we highlight methods to improve test accuracy of an image classifier model for shrapnel identification using tissue phantom image sets. Using a previously developed image classifier neural network—termed ShrapML—blind test accuracy was less than 70% and was variable depending on the training/test data setup, as determined by a leave one subject out (LOSO) holdout methodology. Introduction of affine transformations for image augmentation or MixUp methodologies to generate additional training sets improved model performance and overall accuracy improved to 75%. Further improvements were made by aggregating predictions across five LOSO holdouts. This was done by bagging confidences or predictions from all LOSOs or the top-3 LOSO confidence models for each image prediction. Top-3 LOSO confidence bagging performed best, with test accuracy improved to greater than 85% accuracy for two different blind tissue phantoms. This was confirmed by gradient-weighted class activation mapping to highlight that the image classifier was tracking shrapnel in the image sets. Overall, data augmentation and ensemble prediction approaches were suitable for creating more generalized predictive models for ultrasound image analysis, a critical step for real-time diagnostic deployment.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2022

Regulation of Alfvén Eigenmodes by Microturbulence in Fusion Plasmas

Global gyrokinetic simulations of mesoscale reversed shear Alfven eigenmodes (RSAE) excited by energetic particles (EP) in fusion plasmas find that RSAE amplitude and EP transport are much higher than experimental levels at nonlinear saturation, but quickly diminish to very low levels after the saturation when background microturbulence is artificially suppressed. In contrast, in simulations coupling micro-meso scales, the RSAE amplitude and EP transport decrease drastically at the initial saturation but later increases to the experimental levels in the quasi-steady state with bursty dynamics due to regulation by thermal ion temperature gradient (ITG) microturbulence. The quasi-steady state EP transport is larger for a stronger microturbulence. Here, the RSAE amplitude in the quasi-steady state ITG-RSAE turbulence from gyrokinetic simulations, for the first time, agrees very well with experimental measurements.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2025

Cathode Upcycling for Direct Recycling of Lithium‐Ion Batteries Using a Precipitation Approach

With the increased production of electric vehicles to reduce carbon emissions, the lithium-ion battery market to supply those vehicles has grown dramatically. To enhance battery sustainability and circularity, direct recycling methods aim to recover intact cathode materials. However, end-of-life cathode materials are typically 15–20 years old and often have lower energy density compared to current cathode materials. To address this challenge, a rapid precipitation process is developed to boost energy density by converting low Ni-compositions, LiNi 0.33 Co 0.33 Mn 0.33 O 2 (NMC111), into higher Ni-compositions (NMC622). This process forms a Ni-rich coating on cathode particles that diffuses into the core upon high-temperature relithiation, increasing compositional homogeneity. The upcycling process leverages existing infrastructure, offering low capital cost and minimal additional chemical input. Through ex situ tomographic transmission X-ray microscopy (TXM), 3D Ni:Co elemental mixing is quantified, confirming that elemental content evens at the secondary particle level with a mean NMC622 composition upon relithiation. However, elemental gradients remain in the single crystalline primary particles. Ex situ high-resolution and in situ wide-angle X-ray diffraction reveals concurrent structural changes during the relithiation process. These findings provide key insights into the structural and chemical mechanisms of elemental diffusion to obtain further improvements of increased capacity and retention through compositional conversion of cathode materials.

Cathode Upcycling↗

DOE OSTI2023

EXCHANGE Campaign 1: A Community-Driven Baseline Characterization of Soils, Sediments, and Water Across Coastal Gradients

The EXploration of Coastal Hydrobiogeochemistry Across a Network of Gradients and Experiments (EXCHANGE) program is a consortium of scientists working together to improve our understanding of how the two-way exchange of water between estuaries or large lake lacustuaries and the terrestrial landscape influence the state and function of ecosystems across the coastal interface. EXCHANGE Campaign 1 (EC1) focuses on the spatial variation in biogeochemical structure and function at the coastal terrestrial-aquatic interface (TAI). In the Fall of 2021, the EXCHANGE Consortium gathered samples from 52 TAIs. Samples collected from EC1 were analyzed for bulk geochemical parameters, bulk physicochemical parameters, organic matter characteristics, and redox-sensitive elements.Please download ec1_README.pdf for a complete list of available data in each .zip folder, package version history, and detailed information about the project. This README will serve as the central place for EC1 Data Package updates. Experimental setup and v1 methods are documented in Myers-Pigg and Pennington et al., 2023 (https://doi.org/10.1038/s41597-023-02548-7).EC1 Data Package Structure:ec1_README.pdfec1_methods.pdfec1_metadata_v3.zip...ec1_dd.csv...ec1_flmd.csv...ec1_sample_catalog.csv...ec1_metadata_kitlevel.csv...ec1_metadata_collectionlevel.csv...ec1_data_collectionlevel.csv...ec1_igsn_metadata.csvec1_soil_v3.zipec1_sediment_v3.zipec1_water_v3.zipec1_processingscripts_v3.zipThis data package is on v3 and was originally published May 2023 (v1). Subsequent updates will be published here with new version numbers. Please see the Change History section in ec1_README.pdf for detailed changes.---Acknowledging EXCHANGE: General Support and Data Product UseWe ask that users of EXCHANGE data add the following acknowledgement when publishing data in scholarly articles and data repositories:"This research is based on work supported by COMPASS-FME, a multi-institutional project supported by the U.S. Department of Energy, Office of Science, Biological and Environmental Research as part of the Environmental System Science Program."

54 ENVIRONMENTAL SCIENCES↗