SEARCH · Search NASA
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.
This dataset consists of 1000 records in a single HDF5 file generated from the simulation code available at: https://github.com/lezahlie/heat_diffusion_simulation. The specific heat equations implemented in that simulation are explained here https://visualpde.com/basic- pdes/heat- equation.html and here https://en.wikipedia.org/wiki/Heat_equation.
Climate model projections have previously been used to compute ice shelf basal melt rates in ice sheet models, but the strategies employed – e.g., ocean input, parameterization, calibration technique, and corrections – have varied widely and are often ad hoc. Here, a methodology is proposed for the calculation of circum-Antarctic basal melt rates for floating ice, based on climate models, that is suitable for ISMIP6, the Ice Sheet Model Intercomparison Project for CMIP6 (6th Coupled Model Intercomparison Project). The past and future evolution of ocean temperature and salinity is derived from a climate model by estimating anomalies with respect to the modern day, which are added to a present-day climatology constructed from existing observational datasets. Temperature and salinity are extrapolated to any position potentially occupied by a simulated ice shelf. A simple formulation is proposed for a basal melt parameterization in ISMIP6, constrained by the observed temperature climatology, with a quadratic dependency on either the nonlocal or local thermal forcing. Two calibration methods are proposed: (1) based on the mean Antarctic melt rate (MeanAnt) and (2) based on melt rates near Pine Island's deep grounding line (PIGL). Future Antarctic mean melt rates are an order of magnitude greater in PIGL than in MeanAnt. The PIGL calibration and the local parameterization result in more realistic melt rates near grounding lines. PIGL is also more consistent with observations of interannual melt rate variability underneath Pine Island and Dotson ice shelves. This work stresses the need for more physics and less calibration in the parameterizations and for more observations of hydrographic properties and melt rates at interannual and decadal timescales.
Carbon monoxide concentrations and aerosol properties provide the tracers of global tropospheric perturbation by humankind that are most easily observed from satellite platforms. Aircraft field observations are additionally needed for us to simulate and understand the patterns observed. We report on our accumulating experience in making detailed, situation-specific, 3-D simulations of these tropospheric constituents as observed from the MAPS CO sensor, the Advanced Very High Resolution Radiometer (AVHRR) aerosol scattering data, and the Total Ozone Mapping Spectrometer (TOMS) absorbing-aerosol information. We have found a strong complimentarity of satellite and aircraft data in this effort, and recommend modelers use: (1) satellite data to pose initial questions; (2) aircraft data for modelers' first detailed simulations; and (3) a return to satellite data for global generalization. This work provides one example. Our main tools are the MM5 numerical model for meteorological assimilation and GRACES, our NASA Ames tracer-chemistry model, which incorporates emissions estimates. We report on several six-week simulations of global biomass burning effects as observed in the MAPS October 1994 dataset, and show the usefulness of two aircraft datasets, the TRACE-A (1992) and PEM-Tropics missions of NASA's Global Tropospheric Experiment. Additional information is contained in the original extended abstract.
The 26 December 2004 Indian Ocean tsunami was the first earthquake tsunami of its magnitude to occur since the advent of both digital seismometry and satellite radar altimetry. Both have independently recorded the event from different physical aspects. The seismic data has then been used to estimate the earthquake fault parameters, and a three-dimensional ocean-general-circulation-model (OGCM) coupled with the fault information has been used to simulate the satellite-observed tsunami waves. Here we show that these two datasets consistently provide the tsunami source using independent methodologies of seismic waveform inversion and ocean modeling. Cross-examining the two independent results confirms that the slip function is the most important condition controlling the tsunami strength, while the geometry and the rupture velocity of the tectonic plane determine the spatial patterns of the tsunami.
With the launch of the European Space Agency's Aeolus Mission in 2013, direct spaceborne measurements of vertical wind profiles are imminent via Doppler wind lidar technology. Part of the preparedness for such missions is the development of the proper data assimilation methodology for handling such observations. Since no heritage measurements exist in space, the Joint Observing System Simulation Experiment (Joint OSSE) framework has been utilized to generate a realistic proxy dataset as a precursor to flight. These data are being used for the development of the Gridpoint Statistical Interpolation (GSI) data assimilation system utilized at a number of centers through the United States including the Global Modeling and Assimilation Office (GMAO) at NASA/Goddard Space Flight Center and at the National Centers for Environmental Prediction (NOAA/NWS/NCEP) as an activity through the Joint Center for Satellite Data Assimilation. An update of this ongoing effort will be presented, including the methodology of proxy data generation, the limitations of the proxy data, the handling of line-of-sight wind measurements within the GSI, and the impact on both analyses and forecasts with the addition of the new data type.
With the launch of the European Space Agency's Atmospheric Dynamics Mission (ADM-Aeolus) in 2011 and the call for the 3D-Winds mission in National Research Council's decadal survey, direct spaceborne measurements of vertical wind profiles are imminent via Doppler wind lidar technology. Part of the preparedness for such missions is the development of the proper data assimilation methodology for handling such observations. Since no heritage measurements exist in space, the Joint Observing System Simulation Experiment (Joint OSSE) framework is being utilized to generate a realistic proxy dataset as a precursor to flight. These data are being used for the development of the Gridpoint Statistical Interpolation (GSI) data assimilation system utilized at a number of centers through the United States including the Global Modeling and Assimilation Office (GMAO) at NASA/Goddard Space Flight Center and at the National Centers for Environmental Prediction (NOAA/NWS/NCEP). This effort will be presented, including the methodology of proxy data generation, the handling of line-of-sight wind measurements within the GSI, and the impact on both analyses and forecasts with the addition of the new data type.
Around half the exposed land of the Northern Hemisphere experiences seasonal freezing and thawing. The changes to the soil thermal, hydraulic, and mechanical properties are large and these, in turn, exert significant controls on the water, energy, and carbon cycles of the affected regions. Wide-area freeze/thaw discrimination via microwave remote sensing is possible using both passive and active techniques. We explore freeze/thaw remote sensing using observations from NASA’s Scanning L-band Active Passive (SLAP) airborne sensor during the SLAPex Freeze/Thaw campaign near Winnipeg, Canada in November, 2015. SLAP is an airborne simulator of the Soil Moisture Active Passive (SMAP) satellite mission. The SLAPex Freeze/Thaw dataset is a unique dataset that allows comparison of airborne-derived freeze/thaw state (plus soil moisture for thawed areas) against in-situ ground truth, satellite-derived freeze/thaw from SMAP, as well as tower-based passive microwave freeze/thaw observations. The sensitivity of passive and active microwave sensing to different depths of freezing and different dielectric models (when soil moisture is retrieved) is examined, and the spatial scaling of heterogeneous frozen/thawed pixels is also explored. Implications for the accuracy of satellite-based freeze/thaw discrimination are discussed.
The ability to monitor and understand natural and anthropogenic variability in atmospheric carbon dioxide(CO2) is a growing need of many stakeholders across the world. Systems that assimilate satellite observations, given their short latency and dense spatial coverage, into high resolution global models are valuable, if not essential, tools for addressing this need. A notable drawback of modern assimilation systems is the long latency of many vital in-put datasets, e.g., inventories, in situ measurements, and re-processed remote-sensing data can trail the current date by months to years. This paper describes techniques for bias-correcting surface fluxes derived from satellite observations of the Earth’s surface to be consistent with constraints from inventories and in situ CO2 datasets. The techniques are applicable in both short-term forecasts and retrospective simulations, thus taking advantage of the coverage and short latency of satellite data while reproducing the major features of long-term inventory and in situ records. Our approach begins with a standard collection of diagnostic fluxes which incorporate a variety of remote-sensing driver data, viz. vegetation indices, fire radiative power, and nighttime lights. We then apply an empirical sink so that global budgets of the diagnostic fluxes match given atmospheric and oceanic growth rates for each year. This step removes coherent, systematic flux errors that produce biases inCO2which mask the signals an assimilation system hopes to capture. Depending on the simulation mode, the empirical sink uses different choices of atmospheric growth rates: estimates based on observations in retrospective mode and projections based on seasonal fore-casts of sea surface temperature in forecasting mode. The retrospective fluxes, when used in simulations with NASA’s Goddard Earth Observing System (GEOS), reproduce marine boundary layer measurements with comparable skill to those using fluxes from a modern inversion system. The forecasted fluxes show promising accuracy in their application to the analysis of changes in the carbon cycle as they occur.
This report reviews the scientific literature regarding the human system risks to the microgravity environment of space flight in relation to human performance. The primary human performance-related risks involve deconditioning of the cardiovascular and skeletal muscles systems due to prolonged exposure to the reduced gravitational input. The chronological history of U.S. space flight is reviewed as a starting point to inform and understand the gaps in the knowledge to these risks. Maintenance of physical performance capabilities involves understanding the health of many organ systems (peripheral [vascular, heart, blood volume, skeletal muscle] and central [brain]) that ultimately contribute to the submaximal and maximal capacity of the aerobic (VO2peak), skeletal muscle (strength and endurance) systems. Maintaining astronaut VO2peak, muscle mass, strength, and endurance before, during, and after space flight is a significant priority to NASA for the current International Space Station (ISS) era, as well as for future exploration missions. A growing research database from both space flight and ground-based analog studies finds that the cardiorespiratory system is compromised and skeletal muscles (predominantly postural muscles of the lower limbs) undergo atrophy. These structural and metabolic responses to living in microgravity conditions contribute to physiological deconditioning during space flight that potentially increase the risks to astronauts returning to surface operations (i.e., Moon, Mars, or Earth). The time course changes from short to long-duration space flight and the relationships between in-flight performance deconditioning levels are not well characterized. Moreover, there are large interindividual variabilities that may be dependent on genetics, age, sex, preflight fitness levels, and individual exercise prescriptions that need further careful evaluations. Efforts should be made to understand the current status of preflight, in-flight, and postflight exercise performance capability and to define the operational goals and target areas for protection with the in-flight exercise program. There is a bi-directional relationship between exercise prescription and hardware countermeasures that need further understanding in-flight. For example, hardware with limited capabilities/modalities may be counterbalanced by changes in exercise prescription (i.e., frequency, time, intensity, volume) for providing effective responses to maintain fitness. Importantly, the minimal requirements for exercise prescription on ISS hardware may not translate to lower capability hardware on exploration missions. Due to limited volume on exploration vehicles, future Artemis missions to the Lunar surface will not have similar exercise hardware capabilities as ISS. This may alter the effectiveness of hardware to provide adequate physiological stress on bodily systems allowing for adaptations to maintain aerobic capacity, strength, and bone density. Thus, it will be important to understand the exercise responses of current ISS countermeasures to develop individualized exercise prescriptions that minimize aerobic and muscular risks, accounting for the large variability of responses among crewmembers. Newer exploration exercise hardware is currently being evaluated that is more compact (i.e., E4D and Orion Flywheel) and will require careful evaluation of the hardware on the stressor (i.e., metabolic rate, oxygen uptake, and heart rate work relationships, and force plate load profiles) needed the human body to protect and maintain crew health and performance. Moreover, exercise responses on the hardware need careful evaluation on the chronic adaptations. Lastly, in-flight evaluation of hardware exercise response may differ in 0-g or partial-g compared to 1-g. Therefore, it cannot be assumed that the stress on the body will be the same in each environment. Understanding this has a direct impact on exercise prescriptions. This document provides an overview of key scientific investigations that have been conducted before, during, and after human space flight missions, as well as from human ground-based analog studies that contribute to the evidence base on changes in aerobic capacity and muscle mass, strength, and endurance. Additional data from rodent and nonhuman primate experiments of skeletal muscle unloading completed during space flight or ground-based flight-simulations provide supportive information about this risk topic. Most importantly, a recent, large dataset from long-duration ISS crew has been added to give improved insight into the variability of exercise response of crew, demonstrating that a large portion of the crew population return to Earth with greater than 10-20% loss of aerobic capacity and muscle strength and endurance. Data from human space flight and ground-based studies are narrowing in on the required exercise paradigms but thus far still provide an incomplete answer to an effective approach for maintaining skeletal muscle function and aerobic fitness of all human space travelers. Finally, the relationship of this risk topic to various space flight operational scenarios is examined and discussed.
Modeling of long-term trends of aerosols and their properties is important for constraining aerosol-climate forcing, and for characterizing changes in particulate matter pollution speciation and exposure. Here we study global and regional long-term trends in surface fine particulate matter (PM ) and aerosol optical properties for 60 years from 1958 to 2018, using simulations performed with the NASA Goddard Earth Observing System Chemistry Climate Model (GEOSCCM), and evaluate the model hindcast with observations for the last three decades. Comparing the modeled aerosols with a diverse set of observations helps interpret observed and simulated trends, and serves as a benchmark for future GEOSCCM developments and input datasets improvements. We first characterize modeled global and regional temporal changes in surface PM and its components, aerosol optical depth (AOD) and single scattering albedo (SSA) and we interpret their link with emissions drivers. We then compare modeled surface PM with ground-based observations from monitoring networks and with global reconstructed PM datasets from observations-model data fusion. Total AOD is compared with long-term satellite measurements from the Moderate Resolution Imaging Spectroradiometer (MODIS) and measurements from the ground-based Aerosol Robotic Network (AERONET). Additional aerosol optical properties, such as absorption and scattering coefficients, are evaluated with ground-based observations from the Global Atmosphere Watch (GAW) records.
Remote sensing atmospheric correction methods have primarily focused on imagery over land. However, accurate correction over water is important for monitoring and research of aquatic environments. More research in this area is ongoing, though one of the biggest challenges is enough quality data to develop and validate correction methods. This is especially true for neural network (NN) -based models which have shown promise in this area given enough quality data. To address this deficiency of data, we are leveraging a synthetic dataset produced by a model called SWIPE that uses radiative transfer modeling to simulate the atmospheric effects on water-leaving (WL) reflectance to estimate top-of-atmosphere (TOA) reflectance. This allows us to produce almost unlimited pairs of WL reflectance and corresponding TOA reflectance for model training across a variety of atmospheric conditions. We use two approaches for our atmospheric correction model. One uses a conditional variational autoencoder (VAE) to estimate a single WL reflectance value from a single TOA reflectance value. The second is based on a UNET architecture and estimates an array of WL reflectance values from an array of TOA reflectance values. The goal of the second method is to capture atmospheric effects that occur spatially between values within the array as compared to the first method.
A dedicated study on JET-ILW, deploying two types of ICRH antennas and spectroscopic observation spots at two outboard, beryllium limiters, has provided insight on long-range (up to 6m) RFenhanced plasma-surface interactions (RF-PSI) due to near-antenna electric fields. To aid in the interpretation of optical emission measurements of these effects, the antenna near-fields are computed using the TOPICA code, specifically run for the ITER-like antenna (ILA); similar modelling already existed for the standard JET antennas (A2). In the experiment, both antennas were operated in current drive mode, as RF-PSI tends to be higher in this phasing and at similar power (~0.5 MW). When sweeping the edge magnetic field pitch angle, peaked RF-PSI effects, in the form of 2-4 fold increase in the local Be source,are consistently measured with the observation spots magnetically connect to regions of TOPICAL-calculated high near-fields, particularly at the near-antenna limiters. It is also found that similar RF-PSI effects are produced by the two types of antenna on similarly distant limiters. Although this mapping of calculated near-fields to enhanced RF-PSI gives only qualitative interpretion of the data, the present dataset is expected to provide a sound experimental basis for emerging RF sheath simulation model validation.
NASA GeneLab is an open-access repository for omics datasets generated by biological experiments conducted in space or ground experiments relevant to spaceflight (e.g. simulated cosmic radiation, simulated microgravity, bed rest studies). The GeneLab Data Systems (GLDS) version 4.0 will be available on October 1st 2019, and will provide a state-of-the-art bioinformatics platform for the space biology and radiation communities to upload their data into an omics data commons, to process their data with vetted standard workflows and to compare with existing analyses. Started in 2015 as a repository designed to archive omics data from space experiments, GeneLab has expanded its scope to all ionizing radiation omics experiments conducted on the ground and has put considerable effort in providing carefully characterized radiation metadata on all datasets. GeneLab is also providing processed data derived from the raw data covering a large spectrum of omics (genome, epigenome, transcriptome, epitranscriptome, proteome, metabolome) to help users explore important questions: 1) Which genes or proteins are expressed differently in space for various living organisms? 2) What specific DNA mutations or epigenetic changes happen in space or after exposure to ionizing radiation? and 3) How does genetics affect these responses? Processed data available on GeneLab are derived by standard data analysis workflows vetted by hundreds of scientists who volunteered to join one of the four GeneLab Analysis Working Groups (Animal AWG, Plant AWG, Microbe AWG, Multi-Omics AWG). In this presentation, we will discuss how to bridge the gap between irradiation studies performed on earth and biological experiments conducted in space since the early 1990's. We will discuss how radiation dosimetry was estimated for datasets derived from samples collected during the Space Shuttle era on the International Space Station and on other orbiting platforms. Finally, we will address future strategies regarding dose monitoring in future missions into space, inter-agency efforts to unify data under one umbrella, and knowledge dissemination across the radiation research community and the space biology community.
Mixed-phase cloud microphysical and dynamical processes are still poorly understood, and their representation in GCMs is a major source of uncertainties in overall cloud feedback in GCMs. Thus improving mixed-phase cloud parameterizations in climate models is critical to reducing the climate forecast uncertainties. This study aims at providing improved knowledge of mixed-phase cloud properties from the long-term ACRF observations and improving mixed-phase clouds simulations in the NCAR Community Atmosphere Model version 5 (CAM5). The key accomplishments are: 1) An improved retrieval algorithm was developed to provide liquid droplet concentration for drizzling or mixed-phase stratiform clouds. 2) A new ice concentration retrieval algorithm for stratiform mixed-phase clouds was developed. 3) A strong seasonal aerosol impact on ice generation in Arctic mixed-phase clouds was identified, which is mainly attributed to the high dust occurrence during the spring season. 4) A suite of multi-senor algorithms was applied to long-term ARM observations at the Barrow site to provide a complete dataset (LWC and effective radius profile for liquid phase, and IWC, Dge profiles and ice concentration for ice phase) to characterize Arctic stratiform mixed-phase clouds. This multi-year stratiform mixed-phase cloud dataset provides necessary information to study related processes, evaluate model stratiform mixed-phase cloud simulations, and improve model stratiform mixed-phase cloud parameterization. 5). A new in situ data analysis method was developed to quantify liquid mass partition in convective mixed-phase clouds. For the first time, we reliably compared liquid mass partitions in stratiform and convective mixed-phase clouds. Due to the different dynamics in stratiform and convective mixed-phase clouds, the temperature dependencies of liquid mass partitions are significantly different due to much higher ice concentrations in convective mixed phase clouds. 6) Systematic evaluations of mixed-phase cloud simulations by CAM5 were performed. Measurement results indicate that ice concentrations control stratiform mixed-phase cloud properties. The improvement of ice concentration parameterization in the CAM5 was done in close collaboration with Dr. Xiaohong Liu, PNNL (now at University of Wyoming).
Occupancy schedule is one of the key inputs in Building Energy Modeling (BEM) to reflect the interaction between buildings and occupants. Over the past decades, standardized occupancy schedules, developed mainly by engineering rule-of-thumb, have been widely used in BEM due to its simplicity and lack of real measured occupancy data. However, the BEM community has recognized their association with uncertainty and reliability in simulation results from BEM. This study introduces representative occupancy schedules in the U.S. residential buildings, derived from a large smart thermostat dataset and time-series K-means clustering, and an open-source tool to generate a stochastic residential occupancy schedule. Over 90,000 residential occupancy schedules were estimated from the ecobee Donate Your Data dataset. Then, the representative occupancy schedules were identified through clustering. This study further investigated the impacts of three parameters (day, house type, and state) on residential occupancy schedules. Then, a tool, the Residential Occupancy Schedule Simulator (ROSS), is developed using the representative occupancy schedules derived in this study. Details of this tool are presented in this paper. In conclusion, the derived representative occupancy schedules and the ROSS tool can help improve the energy modeling of residential buildings.
Purpose: Respiratory motion may affect the accuracy of image guidance of radiation treatment of lung cancer. A cone beam computed tomography (CBCT) image spans several breathing cycles, resulting in a blurred object with a theoretical size equal to the sum of tumor size and breathing motion. However, several factors may affect this theoretical relationship. The objective of this study was to analyze the effect of tumor motion on megavoltage (MV)-CBCT images, by comparing target sizes on simulation and pretreatment images of a large cohort of lung cancer patients. Methods: Ninety-three MV-CBCT images from 17 patients were analyzed. Internal target volumes were contoured on each MV-CBCT dataset [internal target volume (ITV{sub CB})]. Their extent in each dimension was compared to that of two volumes contoured on simulation 4-dimensional computed tomography (4D-CT) images: the combination of the tumor contours of each phase of the 4D-CT (ITV{sub 4D}) and the volume contoured on the average CT calculated from the 4D-CT phases (ITV{sub ave}). Tumor size and breathing amplitude were assessed by contouring the tumor on each CBCT raw projection where it could be unambiguously identified. The effect of breathing amplitude on the quality of the MV-CBCT image reconstruction was analyzed. Results: The mean differences between the sizes of ITV{sub CB} and ITV{sub 4D} were −1.6 ± 3.3 mm (p < 0.001), −2.4 ± 3.1 mm (p < 0.001), and −7.2 ± 5.3 mm (p < 0.001) in the anterior/posterior (AP), left/right (LR), and superior/inferior (SI) directions, respectively, showing that MV-CBCT underestimates the full target size. The corresponding mean differences between ITV{sub CB} and ITV{sub ave} were 0.3 ± 2.6 mm (p = 0.307), 0.0 ± 2.4 mm (p = 0.86), and −4.0 ± 4.3 mm (p < 0.001), indicating that the average CT image is more representative of what is visible on MV-CBCT in the AP and LR directions. In the SI directions, differences between ITV{sub CB} and ITV{sub ave} could be separated into two groups based on tumor motion: −3.2 ± 3.2 mm for tumor motion less than 15 mm and −10.9 ± 6.3 mm for tumor motion greater than 15 mm. Deviations of measured target extents from their theoretical values derived from tumor size and motion were correlated with motion amplitude similarly for both MV-CBCT and average CT images, suggesting that the two images were subject to similar motion artifacts for motion less than 15 mm. Conclusions: MV-CBCT images are affected by tumor motion and tend to under-represent the full target volume. For tumor motion up to 15 mm, the volume contoured on average CT is comparable to that contoured on the MV-CBCT. Therefore, the average CT should be used in image registration for localization purposes, and the standard 5 mm PTV margin seems adequate. For tumor motion greater than 15 mm, an additional setup margin may need to be used to account for the increased uncertainty in tumor localization.
The objectives of the HOST Turbine Heat Transfer subproject were to obtain a better understanding of the physics of the aerothermodynamic phenomena occurring in high-performance gas turbine engines and to assess and improve the analytical methods used to predict the fluid dynamics and heat transfer phenomena. At the time the HOST project was initiated, an across-the-board improvement in turbine design technology was needed. Therefore, a building-block approach was utilized, with research ranging from the study of fundamental phenomena and analytical modeling to experiments in simulated real-engine environments. Experimental research accounted for 75 percent of the project, and analytical efforts accounted for approximately 25 percent. Extensive experimental datasets were created depicting the three-dimensional flow field, high free-stream turbulence, boundary-layer transition, blade tip region heat transfer, film cooling effects in a simulated engine environment, rough-wall cooling enhancement in a rotating passage, and rotor-stator interaction effects. In addition, analytical modeling of these phenomena was initiated using boundary-layer assumptions as well as Navier-Stokes solutions.
Increased demand for high-performance permanent magnets in the electric vehicle and wind-turbine industries has prompted the search for cost-effective alternatives. Discovering magnetic materials with the desired intrinsic and extrinsic permanent magnet properties presents a significant challenge to researchers because of issues with the global supply of rare-earth elements, material stability, and a low maximum magnetic energy product BH max . While first-principles density functional theory (DFT) predicts materials’ magnetic moments, magnetocrystalline anisotropy constants, and exchange interactions, it cannot compute extrinsic properties such as coercivity (H c ). Although it is possible to calculate H c theoretically with micromagnetic simulations, the predicted value is larger than the experiment by almost an order of magnitude due to the Brown paradox. To circumvent these issues, we employ machine-learning (ML) methods on an extensive database obtained from experiments, DFT calculations, and micromagnetic modeling. The use of a large experimental dataset enables realistic H c predictions for materials such as Ce-doped Nd 2 Fe 14 B, comparing favorably against micromagnetically simulated coercivities. Remarkably, our ML model accurately identifies uniaxial magneto-crystalline anisotropy as the primary contributor to H c . With DFT calculations, we predict the Nd-site-dependent magnetic anisotropy behavior in Nd 2 Fe 14 B, confirming that Nd 4g sites mainly contribute to uniaxial magnetocrystalline anisotropy, and also calculate the Curie temperature (T c ). Finally, both calculated results are in good agreement with the experiments. The coupled experimental dataset and ML modeling with DFT input predict H c with far greater accuracy and speed than was previously possible using micromagnetic modeling. Further, we reverse engineer the grain-boundary and intergrain exchange coupling with micromagnetic simulations by employing the ML predictions.