Forecasting Subsurface Data Value for Greenfield Geothermal Development
Explore the source record for details and available documents.
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.
Explore the source record for details and available documents.
Puerto Rico Grid Resilience and Transitions to 100% Renewable Energy Study (PR100) provided a comprehensive analysis of pathways for Puerto Rico to achieve its goal of 100% renewable energy by 2050, based on extensive stakeholder input. The study illuminates immediate and longer-term investments needed to achieve reliability while pursuing Puerto Rico's energy goals and addressing critical energy needs. Expected benefits of these investments include improvements in safety, security, health, and economic opportunity.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Buildings typically have a 60-to-75-year lifespan before they require significant maintenance or modifications. However, most builders evaluate the performance of their new buildings using whole-building energy simulation tools based on the current typical meteorological year (TMY) file or actual meteorological year. The energy use consumption and sensible heat release pattern observed from buildings could potentially change based on shifting global climates. Therefore, the recommended energy-efficiency design parameters might also change during these periods. In this study, we evaluate the role of different building design parameters, such as material reflectivity, HVAC COP, and insulation values, on building energy usage and sensible heat release from buildings with different building coverage ratios (BCR), based on the current and future weather file TMY (fTMY) for the middle of the century (2040–2060). The role of sensible heat release from buildings is not accounted for accurately while estimating building energy usage in most whole-building energy simulations. The study conducts a series of whole-building energy simulation analyses using EnergyPlus to evaluate the role of different design parameters based on TMY and fTMY weather conditions. The analysis is conducted for two hot desert climatic cities: Phoenix (USA) and Abu Dhabi (UAE). The results show that, for the base case in a future climate, the sensible heat release is reduced by an average of 30% due to the reduced delta T between the surface and ambient air. Further, the results show an increase in total energy consumption by 5% annually. The results also show that, for buildings with traditional coatings, shorter buildings release more heat than taller buildings. On the other hand, for buildings with reflective paints, shorter buildings release less heat than taller buildings. The findings from this study can be used by policymakers, utility companies, and builders to better understand the relative role of different building design parameters while constructing new and retrofitting existing buildings.
Explore the source record for details and available documents.
Weather events cause most power outages. Often, we even get notifications on our phones to take cover or be prepared for an imminent event. If electric grid utilities had a similar warning that also included probable scenarios and the equipment involved, they could prepare and minimize the effects. Idaho National Laboratory had a project with the U.S. Department of Energy’s Cybersecurity, Energy Security, and Emergency Response program to develop a grid alert application that receives messages from the existing emergency alert system, filters and determines components possibly affected by the emergency event, calculates probable scenarios using MASTERRI (Modeling And Simulation for Targeted Reliability and Resilience Improvement). For high-risk events, the application can then send alert links to subscribed electric distribution utility operations staff to allow them to see and evaluate the scenarios and the impact in a web based interactive map tool. This proof of concept application used data from utilities and organizations, such as the international regulatory body North American Electric Reliability Corporation, which have complied historical failure data of elements that comprise the U.S. electric grid. Nominal failure rates are obtained from this data. To make this tool possible, estimated failure rates were calculated for different component types given the alert type, severity, and location. Historic weather-related grid element failures were correlated with historic weather events from the Integrated Public Alert & Warning System. These correlated events and failures are used along with Bayesian updates from the historical norms to provide a modified failure rate for grid elements in the alert areas and calculate probable scenarios. Working with an industry collaborator, actual grid models and data were used for demonstration cases. This report outlines the work performed for this project.
Here we examine the skill of three, five, and seven-category monthly ENSO probability forecasts (1982–2015) from single and multi-model ensemble integrations of the North American Multimodel Ensemble (NMME) project. Three-category forecasts are typical and provide probabilities for the ENSO phase (El Niño, La Niña or neutral). Additional forecast categories indicate the likelihood of ENSO conditions being weak, moderate or strong. The level of skill observed for differing numbers of forecast categories can help to determine the appropriate degree of forecast precision. However, the dependence of the skill score itself on the number of forecast categories must be taken into account. For reliable forecasts with same quality, the ranked probability skill score (RPSS) is fairly insensitive to the number of categories, while the logarithmic skill score (LSS) is an information measure and increases as categories are added. The ignorance skill score decreases to zero as forecast categories are added, regardless of skill level. For all models, forecast formats and skill scores, the northern spring predictability barrier explains much of the dependence of skill on target month and forecast lead. RPSS values for monthly ENSO forecasts show little dependence on the number of categories. However, the LSS of multimodel ensemble forecasts with five and seven categories show statistically significant advantages over the three-category forecasts for the targets and leads that are least affected by the spring predictability barrier. These findings indicate that current prediction systems are capable of providing more detailed probabilistic forecasts of ENSO phase and amplitude than are typically provided.
The subseasonal-to-seasonal (S2S) predictive time scale, encompassing lead times ranging from 2 weeks to a season, is at the frontier of forecasting science. Forecasts on this time scale provide opportunities for enhanced application-focused capabilities to complement existing weather and climate services and products. There is, however, a “knowledge–value” gap, where a lack of evidence and awareness of the potential socioeconomic benefits of S2S forecasts limits their wider uptake. To address this gap, here we present the first global community effort at summarizing relevant applications of S2S forecasts to guide further decision-making and support the continued development of S2S forecasts and related services. Focusing on 12 sectoral case studies spanning public health, agriculture, water resource management, renewable energy and utilities, and emergency management and response, we draw on recent advancements to explore their application and utility. These case studies mark a significant step forward in moving from potential to actual S2S forecasting applications. We show that by placing user needs at the forefront of S2S forecast development—demonstrating both skill and utility across sectors—this dialogue can be used to help promote and accelerate the awareness, value, and cogeneration of S2S forecasts. We also highlight that while S2S forecasts are increasingly gaining interest among users, incorporating probabilistic S2S forecasts into existing decision-making operations is not trivial. Nevertheless, S2S forecasting represents a significant opportunity to generate useful, usable, and actionable forecast applications for and with users that will increasingly unlock the potential of this forecasting time scale.
The ability of the WRF atmospheric model to forecast wind speed over the Nysted wind park was investigated as a function of time. It was found that in the time period we considered (August 1-19, 2008), the model is able to predict wind speeds reasonably accurately for 48 hours ahead, but that its forecast skill deteriorates rapidly after 48 hours. In addition, a preliminary analysis was carried out to investigate the impact of vertical grid resolution on the forecast skill. Our preliminary finding is that increasing vertical grid resolution does not have a significant impact on the forecast skill of the WRF model over Nysted wind park during the period we considered. Additional simulations during this period, as well as during other time periods, will be run in order to validate the results presented here. Wind speed is a difficult parameter to forecast due the interaction of large and small length scale forcing. To accurately forecast the wind speed at a given location, the model must correctly forecast the movement and strength of synoptic systems, as well as the local influence of topography / land use on the wind speed. For example, small deviations in the forecast track or strength of a large-scale low pressure system can result in significant forecast errors for local wind speeds. The purpose of this study is to provide a preliminary baseline of a high-resolution limited area model forecast performance against observations from the Nysted wind park. Validating the numerical weather prediction model performance for past forecasts will give a reasonable measure of expected forecast skill over the Nysted wind park. Also, since the Nysted Wind Park is over water and some distance from the influence of terrain, the impact of high vertical grid spacing for wind speed forecast skill will also be investigated.
The wind-energy (WE) industry relies on numerical weather prediction (NWP) forecast models as foundational or base models for many purposes, including wind-resource assessment and wind-power forecasting. During the Second Wind Forecast Improvement Project (WFIP2) in the Columbia River Basin of Oregon and Washington, a significant effort was made to improve NWP forecasts through focused model development, to include experimental refinements to the High Resolution Rapid Refresh (HRRR) model physics and horizontal grid spacing. In this study, the performance of an experimental version of HRRR that includes these refinements is tested against a control version, which corresponds to that of the operational HRRR run by National Oceanic and Atmospheric Administration/National Centers for Environmental Protection at the outset of WFIP2. Furthermore, the effects of horizontal grid resolution were also tested by comparing wind forecasts from the HRRR (with 3-km grid spacing) with those from a finer-resolution HRRR nest with 750-m grid spacing. Model forecasts are validated against accurate wind-profile measurements by three scanning, pulsed Doppler lidars at sites separated by a total distance of 71 km. Model skill and improvements in model skill, attributable to physics refinements and improved horizontal grid resolution, varied by season, by site, and during periods of atmospheric phenomena relevant to WE. In general, model errors were the largest below 150 m above ground level (AGL). Experimental HRRR refinements tended to reduce the mean absolute error (MAE) and other error metrics for many conditions, but degradation in skill (increased MAE) was noted below 150 m AGL at the two lowest-elevation sites at night. Finer resolution was found to produce the most significant reductions in the error metrics.
Background: Early insights into the timing of the start, peak, and intensity of the influenza season could be useful in planning influenza prevention and control activities. To encourage development and innovation in influenza forecasting, the Centers for Disease Control and Prevention (CDC) organized a challenge to predict the 2013–14 Unites States influenza season. Methods: Challenge contestants were asked to forecast the start, peak, and intensity of the 2013–2014 influenza season at the national level and at any or all Health and Human Services (HHS) region level(s). The challenge ran from December 1, 2013–March 27, 2014; contestants were required to submit 9 biweekly forecasts at the national level to be eligible. The selection of the winner was based on expert evaluation of the methodology used to make the prediction and the accuracy of the prediction as judged against the U.S. Outpatient Influenza-like Illness Surveillance Network (ILINet). Results: Nine teams submitted 13 forecasts for all required milestones. The first forecast was due on December 2, 2013; 3/13 forecasts received correctly predicted the start of the influenza season within one week, 1/13 predicted the peak within 1 week, 3/13 predicted the peak ILINet percentage within 1 %, and 4/13 predicted the season duration within 1 week. For the prediction due on December 19, 2013, the number of forecasts that correctly forecasted the peak week increased to 2/13, the peak percentage to 6/13, and the duration of the season to 6/13. As the season progressed, the forecasts became more stable and were closer to the season milestones. Conclusion: Forecasting has become technically feasible, but further efforts are needed to improve forecast accuracy so that policy makers can reliably use these predictions. CDC and challenge contestants plan to build upon the methods developed during this contest to improve the accuracy of influenza forecasts.
Over the next three years, the Public Service Company of New Mexico (PNM) plans to increase utility-scale solar photovoltaic (PV) capacity from today’s roughly 330MW to about 1600MW. This massive increase in variable generation—from about 15% to 75% of peak load—will require changes in how PNM operates their system. We characterize the 5 and 30-minute solar and wind forecast errors that the system is likely to experience in order to determine the level of reserves needed to counteract such events. Our focus in this study is on negative forecast error (in other words, shortfalls relative to forecast) – whereas excess variable generation can be curtailed if needed, a shortfall must be compensated for to avoid loss of load. Calculating forecast error requires the use of the same forecasting methods that PNM uses or a reasonable approximation thereof. For wind, we use a persistence forecast on actual 5-minute 2019 wind output data (scaled up to reflect the amount of wind capacity planned for 2025). For solar, we use a formula incorporating the clear sky index (CSI) for the forecast. As the solar on the grid now is a small fraction of what is planned for 2025, we generated 5-minute solar data using 2019 weather inputs. We find that to handle 99.9% of the 5-minute negative forecast errors, a maximum of 275MW of variable generation reserve during daylight hours, and a maximum of 75MW during non-daylight hours, should be sufficient. Note that this variable generation reserve is an additional reserve category that specifies reserves over and above what are currently carried for contingency reserve. This would require a significant increase in reserve relative to what PNM currently carries or can call upon from other utilities per reserve sharing agreements. This variable generation reserve specification may overestimate the actual level needed to deal with PNM’s planned variable generation in 2025. The forecasting methodologies used in this study likely underperform PNM’s forecasting – and better forecasting allows for less reserve. To obtain more precise estimates, it is necessary to consider load and use the same forecasting inputs and methods used by PNM.
We present a probabilistic framework tailored for solar energy applications referred to as the Weather Research and Forecasting-Solar ensemble prediction system (WRF-Solar EPS). WRF-Solar EPS has been developed by introducing stochastic perturbations into the most relevant physical variables for solar irradiance predictions. In this study, we comprehensively discuss the impact of the stochastic perturbations of WRF-Solar EPS on solar irradiance forecasting compared to a deterministic WRF-Solar prediction (WRF-Solar DET), a stochastic ensemble using the stochastic kinetic energy backscatter scheme (SKEBS), and a WRF-Solar multi-physics ensemble (WRF-Solar PHYS). The performances of the four forecasts are evaluated using irradiance retrievals from the National Solar Radiation Database (NSRDB) over the contiguous United States. We focus on the predictability of the day-ahead solar irradiance forecasts during the year of 2018. The results show that the ensemble forecasts improve the quality of the forecasts, compared to the deterministic prediction system, by accounting for the uncertainty derived by the ensemble members. However, the three ensemble systems are under-dispersive, producing unreliable and overconfident forecasts due to a lack of calibration. In particular, WRF-Solar EPS produces less optically thick clouds than the other forecasts, which explains the larger positive bias in WRF-Solar EPS (31.7 W/m 2 ) than in the other models (22.7–23.6 W/m 2 ). This study confirms that the WRF-Solar EPS reduced the forecast error by 7.5% in terms of the mean absolute error (MAE) compared to WRF-Solar DET, and provides in-depth comparisons of forecast abilities with the conventional scientific probabilistic approaches (i.e., SKEBS and a multi-physics ensemble). Guidelines for improving the performance of WRF-Solar EPS in the future are provided.
Abstract The rapid rise of deep learning (DL) in numerical weather prediction (NWP) has led to a proliferation of models which forecast atmospheric variables with comparable or superior skill than traditional physics-based NWP. However, among these leading DL models, there is a wide variance in both the training settings and architecture used. Further, the lack of thorough ablation studies makes it hard to discern which components are most critical to success. In this work, we show that it is possible to attain high forecast skill even with relatively off-the-shelf architectures, simple training procedures, and moderate compute budgets. Specifically, we train a minimally modified Swin Transformer V2 (SwinV2) on ERA5 data and find that it attains superior skill in terms of mean-square errors of deterministic forecasts when compared against the European Centre for Medium-Range Weather Forecasts’ Integrated Forecasting System (IFS). Almost all DL–NWP systems share a core set of hyperparameters and design decisions. To aid and expedite future DL–NWP research, we present an in-depth, systematic exploration of different loss functions, model sizes and depths, patch sizes, and multistep training objectives. We also examine the model performance with metrics beyond the typical accuracy (ACC) and RMSE and investigate how the performance scales with model size. Through our open-source code, scoring pipelines, and models, we share our findings on key aspects of the training pipeline. These ablations reduce the necessity for expensive hyperparameter tuning and lower the barrier to entry for future DL–NWP research. Significance Statement This study investigates the potential of using large-scale transformer-based models for weather prediction, showing that it is possible to achieve high forecast accuracy with simpler, off-the-shelf architectures. By training a minimally modified SwinV2 transformer on ERA5 data, we show that the model achieves competitive forecast skill in terms of mean-square error for key variables, outperforming the European Centre for Medium-Range Weather Forecasts’ Integrated Forecasting System (IFS) at all lead times. Our findings suggest that effective training strategies, such as multistep fine-tuning and channel-weighted losses, significantly enhance the model’s performance. However, we also highlight that these improvements come with trade-offs in other areas, such as ensemble spread and high-frequency spatial detail. This work highlights the promise of deep learning in improving weather forecasts, which could lead to better preparedness and response to weather events, ultimately benefiting society by providing more reliable weather predictions.
The overall goal of this multi-phased research project known as WindSENSE is to develop an observation system deployment strategy that would improve wind power generation forecasts. The objective of the deployment strategy is to produce the maximum benefit for 1- to 6-hour ahead forecasts of wind speed at hub-height (~80 m). In Phase III of the project, the focus was on the Mid-Columbia Basin region which encompasses the Bonneville Power Administration (BPA) wind generation area shown in Figure 1 that includes Klondike, Stateline, and Hopkins Ridge wind plants. The typical hub height of a wind turbine is approximately 80-m above ground level (AGL). So it would seem that building meteorological towers in the region upwind of a wind generation facility would provide data necessary to improve the short-term forecasts for the 80-m AGL wind speed. However, this additional meteorological information typically does not significantly improve the accuracy of the 0- to 6-hour ahead wind power forecasts because processes controlling wind variability change from day-to-day and, at times, from hour-to-hour. It is also important to note that some processes causing significant changes in wind power production function principally in the vertical direction. These processes will not be detected by meteorological towers at off-site locations. For these reasons, it is quite challenging to determine the best type of sensors and deployment locations. To address the measurement deployment problem, Ensemble Sensitivity Analysis (ESA) was applied in the Phase I portion of the WindSENSE project. The ESA approach was initially designed to produce spatial fields that depict the sensitivity of a forecast metric to a set of prior state variables selected by the user. The best combination of variables and locations to improve the forecast was determined using the Multiple Observation Optimization Algorithm (MOOA) developed in Phase I. In Zack et al. (2010a), the ESA-MOOA approach was applied and evaluated for the wind plants in the Tehachapi Pass region for a period during the warm season. That research demonstrated that forecast sensitivity derived from the dataset was characterized by well-defined, localized patterns for a number of state variables such as the 80-m wind and the 25-m to 1-km temperature difference prior to the forecast time. The sensitivity patterns produced as part of the Tehachapi Pass study were coherent and consistent with the basic physical processes that drive wind patterns in the Tehachapi area. In Phase II of the WindSENSE project, the ESA-MOOA approach was extended and applied to the wind plants located in the Mid-Columbia Basin wind generation area of Washington-Oregon during the summer and to the Tehachapi Pass region during the winter. The objective of this study was to identify measurement locations and variables that have the greatest positive impact on the accuracy of wind forecasts in the 0- to 6-hour look-ahead periods for the two regions and to establish a higher level of confidence in ESA-MOOA for mesoscale applications. The detailed methodology and results are provided in separate technical reports listed in the publications section below. Ideally, the data assimilation scheme used in the Phase III experiments would have been based upon an ensemble Kalman filter (EnKF) that was similar to the ESA method used to diagnose the Mid-Columbia Basin sensitivity patterns in the previous studies. However, running an EnKF system at high resolution is impractical because of the very high computational cost. Thus, it was decided to use a three-dimensional variational (3DVAR) analysis scheme that is less computationally intensive. The objective of this task is to develop an observation system deployment strategy for the mid Columbia Basin (i.e. the BPA wind generation region) that is designed to produce the maximum benefit for 1- to 6-hour ahead forecasts of hub-height (~80 m) wind speed with a focus on periods of large changes in wind speed. There are two tasks in the current project effort designed to validate the ESA observational system deployment approach in order to move closer to the overall goal: (1) Perform an Observing System Experiment (OSE) using a data denial approach. (2) Conduct a set of Observing System Simulation Experiments (OSSE) for the Mid-Columbia basin region.
The implementation of the Department of Energy (DOE) mandated accelerated cleanup program created significant potential technical impediments. The schedule compression required close coordination and a comprehensive review and prioritization of the barriers that impeded treatment and disposition of the waste streams at each site. Many issues related to site waste treatment and disposal were potential critical path issues under the accelerated schedules. In order to facilitate accelerated cleanup initiatives, waste managers at DOE field sites and at DOE Headquarters in Washington, D.C., needed timely waste forecast information regarding the volumes and types of waste that would be generated by DOE sites over the next 30 years. Each local DOE site has historically collected, organized, and displayed site waste forecast information in separate and unique systems. However, waste information from all sites needed a common application to allow interested parties to understand and view the complete complex-wide picture. A common application allows identification of total waste volumes, material classes, disposition sites, choke points, and technological or regulatory barriers to treatment and disposal. The Applied Research Center (ARC) at Florida International University (FIU) in Miami, Florida, has completed the deployment of this fully operational, web-based forecast system. New functional modules and annual waste forecast data updates have been added to ensure the long-term viability and value of this system. In conclusion: WIMS continues to successfully accomplish the goals and objectives set forth by DOE for this project. WIMS has replaced the historic process of each DOE site gathering, organizing, and reporting their waste forecast information utilizing different database and display technologies. In addition, WIMS meets DOE's objective to have the complex-wide waste forecast information available to all stakeholders and the public in one easy-to-navigate system. The enhancements to WIMS made over the year since its web deployment include the addition of new DOE sites, an updated data set, and the ability to easily print the forecast data tables, the disposition maps, and the GIS maps. Future enhancements will include a high-level waste summary, a display of waste forecast by mode of transportation, and a user help module. The waste summary display module will provide a high-level summary view of the waste forecast data based on the selection of sites, facilities, material types, and forecast years. The waste summary report module will allow users to build custom filtered reports in a variety of formats, such as MS Excel, MS Word, and PDF. The user help module will provide a step-by-step explanation of various modules, using screen shots and general tutorials. The help module will also provide instructions for printing and margin/layout settings to assist users in using their local printers to print maps and reports. (authors)