Search NASA⌕ Search

SEARCH · Search NASA

Results for “FORECASTING MODEL”

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 721 records · Page 40
All sourcesNASA NTRSDOE OSTI
DOE OSTI2018

Measures of model performance based on the log accuracy ratio

Quantitative assessment of modeling and forecasting of continuous quantities uses a variety of approaches. We review existing literature describing metrics for forecast accuracy and bias, concentrating on those based on relative errors and percentage errors. Of these accuracy metrics, the mean absolute percentage error (MAPE) is one of the most common across many fields and has been widely applied in recent space science literature and we highlight the benefits and drawbacks of MAPE and proposed alternatives. We then introduce the log accuracy ratio, and derive from it two metrics: the median symmetric accuracy; and the symmetric signed percentage bias. Robust methods for estimating the spread of a multiplicative linear model using the log accuracy ratio are also presented. The developed metrics are shown to be easy to interpret, robust, and to mitigate the key drawbacks of their more widely-used counterparts based on relative errors and percentage errors. Their use is illustrated with radiation belt electron flux modeling examples.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Equation‐Free Surrogate Modeling of Geophysical Flows at the Intersection of Machine Learning and Data Assimilation

Abstract There is a growing interest in developing data‐driven reduced‐order models for atmospheric and oceanic flows that are trained on data obtained either from high‐resolution simulations or satellite observations. The data‐driven models are non‐intrusive in nature and offer significant computational savings compared to large‐scale numerical models. These low‐dimensional models can be utilized to reduce the computational burden of generating forecasts and estimating model uncertainty without losing the key information needed for data assimilation (DA) to produce accurate state estimates. This paper aims at exploring an equation‐free surrogate modeling approach at the intersection of machine learning and DA in Earth system modeling. With this objective, we introduce an end‐to‐end non‐intrusive reduced‐order modeling (NIROM) framework equipped with contributions in modal decomposition, time series prediction, optimal sensor placement, and sequential DA. Specifically, we use proper orthogonal decomposition (POD) to identify the dominant structures of the flow, and a long short‐term memory network to model the dynamics of the POD modes. The NIROM is integrated within the deterministic ensemble Kalman filter (DEnKF) to incorporate sparse and noisy observations at optimal sensor locations obtained through QR pivoting. The feasibility and the benefit of the proposed framework are demonstrated for the NOAA Optimum Interpolation Sea Surface Temperature (SST) V2 data set. Our results indicate that the NIROM is stable for long‐term forecasting and can model dynamics of SST with a reasonable level of accuracy. Furthermore, the prediction accuracy of the NIROM gets improved by almost one order of magnitude by the DEnKF algorithm.

Pawar, Suraj↗
DOE OSTI2017

A data model of the Climate and Forecast metadata conventions (CF-1.6) with a software implementation (cf-python v2.1)

Abstract. The CF (Climate and Forecast) metadata conventions are designed to promote the creation, processing, and sharing of climate and forecasting data using Network Common Data Form (netCDF) files and libraries. The CF conventions provide a description of the physical meaning of data and of their spatial and temporal properties, but they depend on the netCDF file encoding which can currently only be fully understood and interpreted by someone familiar with the rules and relationships specified in the conventions documentation. To aid in development of CF-compliant software and to capture with a minimal set of elements all of the information contained in the CF conventions, we propose a formal data model for CF which is independent of netCDF and describes all possible CF-compliant data. Because such data will often be analysed and visualised using software based on other data models, we compare our CF data model with the ISO 19123 coverage model, the Open Geospatial Consortium CF netCDF standard, and the Unidata Common Data Model. To demonstrate that this CF data model can in fact be implemented, we present cf-python, a Python software library that conforms to the model and can manipulate any CF-compliant dataset.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Scoping review of Culex mosquito life history trait heterogeneity in response to temperature

Abstract Background Mosquitoes in the genus Culex are primary vectors in the US for West Nile virus (WNV) and other arboviruses. Climatic drivers such as temperature have differential effects on species-specific changes in mosquito range, distribution, and abundance, posing challenges for population modeling, disease forecasting, and subsequent public health decisions. Understanding these differences in underlying biological dynamics is crucial in the face of climate change. Methods We collected empirical data on thermal response for immature development rate, egg viability, oviposition, survival to adulthood, and adult lifespan for Culex pipiens, Cx. quinquefasciatus, Cx. tarsalis , and Cx. restuans from existing literature according to the PRISMA scoping review guidelines. Results We observed linear relationships with temperature for development rate and lifespan, and nonlinear relationships for survival and egg viability, with underlying variation between species. Optimal ranges and critical minima and maxima also appeared varied. To illustrate how model output can change with experimental input data from individual Culex species, we applied a modified equation for temperature-dependent mosquito type reproduction number for endemic spread of WNV among mosquitoes and observed different effects. Conclusions Current models often input theoretical parameters estimated from a single vector species; we show the need to implement the real-world heterogeneity in thermal response between species and present a useful data resource for researchers working toward that goal. Graphical Abstract

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2019

Simultaneous Dependence of the Earthquake-Size Distribution on Faulting Style and Depth

We analyze two high-quality Southern Californian earthquake catalogues, one with focal mechanisms, to statistically model and test for dependencies of the earthquake-size distribution, the b values, on both faulting style and depth. In our null hypothesis, b is assumed constant. We then develop and calibrate one model based only on faulting style, another based only on depth dependence and two models that assume a simultaneous dependence on both parameters. We develop a new maximum-likelihood estimator corrected for the degrees of freedom to assess models' performances. Our results show that all models significantly reject the null hypothesis. The best performing is the one that simultaneously takes account of depth and faulting style. Our results suggest that differential stress variations in the Earth's crust systematically influence b values and that this variability should be considered for contemporary seismic hazard studies.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Temporal Forecasting of Distributed Temperature Sensing in a Thermal Hydraulic System With Machine Learning and Statistical Models

We benchmark performance of long-short term memory (LSTM) network machine learning model and autoregressive integrated moving average (ARIMA) statistical model in temporal forecasting of distributed temperature sensing (DTS). Data in this study consists of fluid temperature transient measured with two co-located Rayleigh scattering fiber optic sensors (FOS) in a forced convection mixing zone of a thermal tee. We treat each gauge of a FOS as an independent temperature sensor. We first study prediction of DTS time series using Vanilla LSTM and ARIMA models trained on prior history of the same FOS that is used for testing. The results yield maximum absolute percentage error (MaxAPE) and root mean squared percentage error (RMSPE) of 1.58% and 0.06% for ARIMA, and 3.14% and 0.44% for LSTM, respectively. Next, we investigate zero-shot forecasting (ZSF) with LSTM and ARIMA trained on history of the co-located FOS only, which is advantageous when limited training data is available. The ZSF MaxAPE and RMSPE values for ARIMA are comparable to those of the Vanilla use case, while the error values for LSTM increase. We show that in ZSF, performance of LSTM network can be improved by training on most correlated gauges between the two FOS, which are identified by calculating the Pearson correlation coefficient. The improved ZSF MaxAPE and RMSPE for LSTM are 4.4% and 0.33%, respectively. Performance of ZSF LSTM can be further enhanced through transfer learning (TL), where LSTM is re-trained on a subset of the FOS that is the target of forecasting. We show that LSTM pre-trained on correlated dataset and re-trained on 30% of testing target dataset achieves MaxAPE and RMSPE values of 2.32% and 0.28%, respectively.

ARIMA↗
DOE OSTI2024

How well are hazards associated with derechos reproduced in regional climate simulations?

Abstract. A 15-member ensemble of convection-permitting regional simulations of the fast-moving and destructive derecho of 29–30 June 2012 that impacted the northeastern urban corridor of the USA is presented. This event generated 1100 reports of damaging winds, generated significant wind gusts over an extensive area of up to 500 000 km2, caused several fatalities, and resulted in widespread loss of electrical power. Extreme events such as this are increasingly being used within pseudo-global-warming experiments to examine the sensitivity of historical, societally important events to global climate non-stationarity and how they may evolve as a result of changing thermodynamic and dynamic contexts. As such it is important to examine the fidelity with which such events are described in hindcast experiments. The regional simulations presented herein are performed using the Weather Research and Forecasting (WRF) model. The resulting ensemble is used to explore simulation fidelity relative to observations for wind gust magnitudes, spatial scales of convection (as is manifest in high composite reflectivity, cREF), and both rainfall and hail production as a function of model configuration (microphysics parameterization, lateral boundary conditions (LBCs), start date, use of nudging, compiler choice, damping, and number of vertical levels). We also examine the degree to which each ensemble member differs with respect to key mesoscale drivers of convective systems (e.g., convective available potential energy and vertical wind shear) and critical manifestations of deep convection, e.g., vertical velocities, cold-pool generation, and how those properties relate to the correct characterization of the associated atmospheric hazards (wind gusts and hail). Use of a double-moment, seven-class scheme with number concentrations for all species (including hail and graupel) results in the greatest fidelity of model-simulated wind gusts and convective structure to the observations of this event. All ensemble members, however, fail to capture the intensity of the event in terms of the spatial extent of convection and the production of high near-surface wind gusts. We further show very high sensitivity to the LBCs employed and specifically that simulation fidelity is higher for simulations nested within ERA-Interim compared to ERA5. Excess convective available potential energy (CAPE) in all ensemble members after the derecho passage leads to excess production of convective cells, wind gusts, cREF > 40 dBZ, and precipitation during a frontal passage on the subsequent day. This event proved very challenging to forecast in real time and to reproduce in the 15-member hindcast simulation ensemble presented here. Future work could examine if simulations with other initial and lateral boundary conditions can achieve greater fidelity.

Shepherd, Tristan (ORCID:0000000186276419)↗
DOE OSTI2021

Time Evolution and Diurnal Variability of the Parametric Sensitivity of Turbine-Height Winds in the MYNN-EDMF Parameterization

The Mellor-Yamada-Nakanishi-Niino (MYNN) parameterization applied in the Weather Research and Forecasting (WRF) model has been augmented to include the Eddy-Diffusion Mass-Flux (EDMF) approach to better represent transport by boundary-layer eddies. This change includes the addition of new parameters associated with convective updrafts and boundary-layer clouds. These additions lead to new parametric sensitivities in the hub-height wind speed compared to simulations using the standard MYNN parameterization that were explored previously. This work builds on efforts focused on WRF’s MYNN parameterization by examining the sensitivity of hub-height wind speed to parameters in the MYNN-EDMF parameterization as a function of simulation lead time. Summer and winter periods were selected from the second Wind Forecast Improvement Project (WFIP2). Five sets of simulations were completed for each season, with lead times ranging from 2 to 6 days. The results show that the sensitivity to the new parameters associated with the EDMF scheme is generally small compared to other parameters, with the largest impact in convective boundary layers. The spread in the perturbed parameter ensembles was found to grow quickly over the first 8 to 19 hours in the summer simulations and 17 to 24 hours in the winter simulations with little change after that, regardless of the simulation length. A strong diurnal cycle in the parameter sensitivity was also found associated with changes in the atmospheric stability, as well as an increase in the sensitivity to the entrainment parameter used in the EDMF parameterization that is associated with increasing fractional areas covered by plumes.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

PS3: The Pheno-Synthesis software suite for integration and analysis of multi-scale, multi-platform phenological data

Phenology is the study of recurring plant and animal life-cycle stages which can be observed across spatial and temporal scales that span orders of magnitude (e.g., organisms to landscapes). The variety of scales at which phenological processes operate is reflected in the range of methods for collecting phenologically relevant data, and the programs focused on these collections. Consideration of the scale at which phenological observations are made, and the platform used for observation, is critical for the interpretation of phenological data and the application of these data to both research questions and land management objectives. However, there is currently little capacity to facilitate access, integration and analysis of cross-scale, multi-platform phenological data. This paper reports on a new suite of software and analysis tools – the “Pheno-Synthesis Software Suite,” or PS3 – to facilitate integration and analysis of phenological and ancillary data, enabling investigation and interpretation of phenological processes at scales ranging from organisms to landscapes and from days to decades. We use PS3 to investigate phenological processes in a semi-aride, mixed shrub-grass ecosystem, and find that the apparent importance of seasonal precipitation to vegetation activity (i.e., “greenness”) is affected by the scale and platform of observation. We end by describing potential applications of PS3 to phenological modeling and forecasting, understanding patterns and drivers of phenological activity in real-world ecosystems, and supporting agricultural and natural resource management and decision-making.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Privacy-preserving federated learning: Application to behind-the-meter solar photovoltaic generation forecasting

Here, the growing usage of decentralized renewable energy sources has made accurate estimation of their aggregated generation crucial for maintaining grid flexibility and reliability. However, the majority of distributed photovoltaic (PV) systems are behind-the-meter (BTM) and invisible to utilities, leading to three challenges in obtaining an accurate forecast of their aggregated output. Firstly, traditional centralized prediction algorithms used in previous studies may not be appropriate due to privacy concerns. There is therefore a need for decentralized forecasting methods, such as federated learning (FL), to protect privacy. Secondly, there has been no comparison between localized, centralized, and decentralized forecasting methods for BTM PV production, and the trade-off between prediction accuracy and privacy has not been explored. Lastly, the computational time of data-driven prediction algorithms has not been examined. This article presents a FL power forecasting method for PVs, which uses federated learning as a decentralized collaborative modeling approach to train a single model on data from multiple BTM sites. The machine learning network used to design this FL-based BTM PV forecasting model is a multi-layered perceptron, which ensures privacy and security of the data. Comparing the suggested FL forecasting model to non-private centralized and entirely private localized models revealed that it has a high level of accuracy, with an RMSE that is 18.17% lower than localized models and 9.9% higher than centralized models.

14 SOLAR ENERGY↗
DOE OSTI2023

Cross-Scale Land-Atmosphere Experiment (CSLAEX) (Final Report)

It is now well recognized that land-atmosphere interactions play an important role in both weather and climate. These interactions take place over a wide range of spatial and temporal scales. Most of the current formulations of the momentum, heat and moisture transport laws are informed by observations with a limited temporal or spatial scale, and hence may have limited applicability. The objective of this work is to improve the multi-scale representation of the near-surface heat transfer in the ground and in the atmosphere using a combination of modeling and field observations. Our knowledge and predictive capacity in land-atmosphere coupling is limited by our capacity to observe the full-scale spectrum at play. In order to bridge this fundamental gap in our observation capacity I will work to: create the first long-term Integrated Cross-Scale land-Atmosphere Experiment (CSLAEX) since I will work to: measure the components of the surface energy balance at a spatial scale relevant to remote sensing and meteorological applications (5km), along with observations of the nested subscale processes with Distributed Temperature Sensors (DTS). This proposal aims to fill the gaps in modeling and observing land-atmosphere interactions across scales via a novel Cross-Scale Land-Atmosphere Experiment (CSLAEX) to be implemented as part of the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) site near Lamont, Oklahoma. The proposed research effort aims at exploring land-atmosphere processes across multiple spatial scales and to investigate the effect of landscape heterogeneity using Distributed Temperature Sensor measurements. The novel instrumentation of the Surface Energy Balance and simultaneous observations of the atmospheric state (boundary layer, cloud, convection) available at the DOE ARM SGP site will be used to create CSLAEX. This dataset will be used to frame and evaluate new transport laws and parameterizations for the boundary layer and the land surface, which can be implemented into the Weather and Research Forecasting (WRF) model and Community Atmospheric Model (CAM). The result of this study should impact multiple fields: meteorology, climate, and hydrology.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

First Ever Polymer Flood Field Pilot To Enhance The Recovery Of Heavy Oils On Alaska’s North Slope - Pushing Ahead One Year Later

In June 2018 the team embarked on an ambitious project to address the slow development pace of Alaska’s 20+ billion barrels heavy oil resource via the first ever polymer flood pilot. Following the successful commencement of the pilot in August 2018, the field demonstration, supporting laboratory experiments and numerical simulation have steadily progressed. A significant amount of valuable data and lessons learned have been collected, and are reported in this paper. The ongoing pilot and the research activities is making headway toward the primary objective of validating the use of polymer flooding for extracting heavy oil in Alaska’s challenging environment. The pilot is conducted in two pre-existing pairs of horizontal injectors and producers in an isolated fault block of the Schrader Bluff heavy oil reservoir at the Milne Point Field. A customized polymer blending and pumping unit injects HPAM polymer at a concentration of 1,750 ppm to achieve a target viscosity of 45 cP. Supporting coreflood laboratory experiments have focused on quantification of polymer retention in the rock, and effect of injection water salinity, polymer, and their combinations on oil recovery. The injection and production response of the pilot flood pattern is utilized to develop a history matched reservoir simulation model for forecasting oil recovery beyond the pilot. Finally, specially designed laboratory experiments address anticipated operating concerns regarding post-polymer breakthrough such as oil-water separation efficiency and polymer induced fouling of heater tubes. Polymer has been injected continuously since startup except for two short equipment modification shutdowns, and more recently a prolonged disruption due to polymer hydration issues at the J-pad field site. Cumulatively, over 600,000 lbs. of polymer has been injected, corresponding to ~7%PV. The two producers show significant decrease in the water cut, gradually increasing oil rate, and no polymer breakthrough. Two main observations from the coreflood are a significant uncertainty in polymer retention values, and positive oil recovery response to low salinity water (2,600 mg/liter TDS). The heterogeneity in the flood pattern presents some challenges in obtaining a robust history matched simulation model. Experimental results on produced fluids treatment indicate the formation of a dense polymer deposit, at certain conditions, on heating tubes that can negatively impact the heat transfer efficiency. The scientific knowledge, including the lessons learned during unanticipated shutdowns, quality control, logistics and field data that is being acquired from this effort has referential value for other planned EOR projects. Finally, by all indications, the polymer field pilot is steadily progressing toward achieving the ultimate goal of unlocking the massive heavy oil resources on Alaska North Slope (ANS).

Dandekar, Abhijit↗
DOE OSTI2017

Development and Implementation of an Optimization Model for Hydropower and Total Dissolved Gas in the Mid-Columbia River System

Managing energy, water, and environmental priorities and constraints within a cascade hydropower system is a challenging multiobjective optimization effort that requires advanced modeling and forecasting tools. Within the mid-Columbia River system, there is currently a lack of specific solutions for predicting how coordinated operational decisions can mitigate the impacts of total dissolved gas (TDG) supersaturation while satisfying multiple additional policy and hydropower generation objectives. In this study, a reduced-order TDG uptake equation is developed that predicts tailrace TDG at seven hydropower facilities on the mid-Columbia River. The equation is incorporated into a general multiobjective river, reservoir, and hydropower optimization tool as a prioritized operating goal within a broader set of system-level objectives and constraints. A test case is presented to assess the response of TDG and hydropower generation when TDG supersaturation is optimized to remain under state water-quality standards. Satisfaction of TDG as an operating goal is highly dependent on whether constraints that limit TDG uptake are implemented at a higher priority than generation requests. According to the model, an opportunity exists to reduce TDG supersaturation and meet hydropower generation requirements by shifting spillway flows to different time periods. In conclusion, a coordinated effort between all project owners is required to implement systemwide optimized solutions that satisfy the operating policies of all stakeholders.

13 HYDRO ENERGY↗
DOE OSTI2018

Impacts of air pollutants from fire and non-fire emissions on the regional air quality in Southeast Asia

Abstract. Severe haze events in Southeast Asia caused by particulate pollution have become more intense and frequent in recent years. Widespread biomass burning occurrences and particulate pollutants from human activities other than biomass burning play important roles in degrading air quality in Southeast Asia. In this study, numerical simulations have been conducted using the Weather Research and Forecasting (WRF) model coupled with a chemistry component (WRF-Chem) to quantitatively examine the contributions of aerosols emitted from fire (i.e., biomass burning) versus non-fire (including fossil fuel combustion, and road dust, etc.) sources to the degradation of air quality and visibility over Southeast Asia. These simulations cover a time period from 2002 to 2008 and are driven by emissions from (a) fossil fuel burning only, (b) biomass burning only, and (c) both fossil fuel and biomass burning. The model results reveal that 39 % of observed low-visibility days (LVDs) can be explained by either fossil fuel burning or biomass burning emissions alone, a further 20 % by fossil fuel burning alone, a further 8 % by biomass burning alone, and a further 5 % by a combination of fossil fuel burning and biomass burning. Analysis of an 24 h PM 2.5 air quality index (AQI) indicates that the case with coexisting fire and non-fire PM 2.5 can substantially increase the chance of AQI being in the moderate or unhealthy pollution level from 23 to 34 %. The premature mortality in major Southeast Asian cities due to degradation of air quality by particulate pollutants is estimated to increase from ∼ 4110 per year in 2002 to ∼ 6540 per year in 2008. In addition, we demonstrate the importance of certain missing non-fire anthropogenic aerosol sources including anthropogenic fugitive and industrial dusts in causing urban air quality degradation. An experiment of using machine learning algorithms to forecast the occurrence of haze events in Singapore is also explored in this study. All of these results suggest that besides minimizing biomass burning activities, an effective air pollution mitigation policy for Southeast Asia needs to consider controlling emissions from non-fire anthropogenic sources.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2017

Impacts of air pollutants from fire and non-fire emissions on the regional air quality in Southeast Asia

Severe haze events in Southeast Asia caused by particulate pollution have become more intense and frequent in recent years, degrading air quality, threatening human health, and interrupting economic and societal activities. Widespread biomass burning activities are a major source of severe haze events in Southeast Asia. On the other hand, particulate pollutants from human activities other than biomass burning also play an important role in degrading air quality in Southeast Asia. In this study, numerical simulations have been conducted using the Weather Research and Forecasting (WRF) model coupled with a chemistry component (WRF-Chem) to quantitatively examine the contributions of aerosols emitted from fire (i.e., biomass burning) versus non-fire (including fossil fuel combustion, road and industrial dust, land use, and land change, etc.) sources to the degradation of air quality and visibility over Southeast Asia. These simulations cover a time period from 2002 to 2008 and were respectively driven by emissions from: (a) fossil fuel burning only, (b) biomass burning only, and (c) both fossil fuel and biomass burning. Across ASEAN 50 cities, these model results reveal that 39 % of observed low visibility days can be explained by either fossil fuel burning or biomass burning emissions alone, a further 20 % by fossil fuel burning alone, a further 8 % by biomass burning alone, and a further 5 % by a combination of fossil fuel burning and biomass burning. The remaining 28 % of observed low visibility days remain unexplained, likely due to emissions sources that have not been accounted for. Further analysis of 24-hr PM 2.5 Air Quality Index (AQI) indicates that comparing to the simulated result of the case with stand-alone non-fire emissions, the case with coexisting fire and non-fire PM 2.5 can substantially increase the chance of AQI being in the moderate or unhealthy pollution level from 23 % to 34 %. The premature mortality among major Southeast Asian cities due to degradation of air quality by particulate pollutants is estimated to increase from ~ 4110 per year in 2002 to ~ 6540 per year in 2008. In addition, we demonstrate the importance of certain missing non-fire anthropogenic aerosol sources including anthropogenic fugitive and industrial dusts in causing urban air quality degradation. An exploratory experiment of using machine learning algorithms to forecasting the occurrence of haze events in Singapore is also demonstrated in this study. All these results suggest that besides minimizing biomass burning activities, an effective air pollution mitigation policy for Southeast Asia needs to consider controlling emissions from non-fire anthropogenic sources.

Lee, Hsiang-He (ORCID:0000000251407324)↗
DOE OSTI2022

The Shortwave Radiative Flux Response to an Injection of Sea Salt Aerosols in the Gulf of Mexico

We report marine cloud brightening (MCB) has been proposed as a potential means of geoengineering the climate, temporarily providing cooling to offset some of the effects of climate change. Marine sky brightening (MSB), involving the direct scattering of sunlight from sea salt injection into the marine boundary layer, has been proposed as an additional geoengineering method that could work in areas that are not regularly cloudy. Here, we use a regional atmospheric model to simulate MCB and MSB over the Gulf of Mexico and nearby land, a highly populated and economically important region that is not characterized by persistent marine stratocumulus cloud cover. Injection of sea salt in the Aitken mode from a region in the central Gulf of Mexico equivalent to 10.8 Tg yr -1 produces an upwards 8.4 W m -2 radiative flux change across the region at the top of the atmosphere, largely due to cloud property changes. Comparatively, a similar mass injection in the accumulation mode produces a 3.1 W m -2 radiative flux change driven primarily by direct scattering. Injection of even larger particles produces a much smaller radiative flux change. Shortwave flux changes due to clouds are largely driven by an increase in cloud droplet number concentration and an increase in cloud liquid water path (each contributing about 45% to the flux change), with a much lower contribution from cloud fraction changes (10%).

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Discrete Graph Structure Learning for Forecasting Multiple Time Series

Time series forecasting is an extensively studied subject in statistics, economics, and computer science. Exploration of the correlation and causation among the variables in a multivariate time series shows promise in enhancing the performance of a time series model. When using deep neural networks as forecasting models, we hypothesize that exploiting the pairwise information among multiple (multivariate) time series also improves their forecast. If an explicit graph structure is known, graph neural networks (GNNs) have been demonstrated as powerful tools to exploit the structure. In this work, we propose learning the structure simultaneously with the GNN if the graph is unknown. We cast the problem as learning a probabilistic graph model through optimizing the mean performance over the graph distribution. The distribution is parameterized by a neural network so that discrete graphs can be sampled differentiably through reparameterization. Empirical evaluations show that our method is simpler, more efficient, and better performing than a recently proposed bilevel learning approach for graph structure learning, as well as a broad array of forecasting models, either deep or non-deep learning based, and graph or non-graph based.

Shang, Chao↗
DOE OSTI2020

Development of a Time-Height Profile Assimilation Technique for Large-Eddy Simulation

Mesoscale-to-microscale coupling (MMC) aims to address the limited scope of traditional large-eddy simulations by driving the microscale flow with information concerning large-scale weather patterns provided by mesoscale models. We present a new offline MMC technique for horizontally homogeneous microscale flow conditions, in which internal forcing terms are computed based on mesoscale time–height profiles of mean-flow quantities. The advantage of such an approach is that it can be used to drive a microscale simulation with either mesoscale or observational data, and that it does not rely on specific terms in the mesoscale budget equations, which are typically not part of the default output of a mesoscale solver. The performance of the proposed profile assimilation technique is assessed based on the simulation of a typical diurnal cycle over the Scaled Wind Farm Technology site in west Texas. Results indicate that simple data assimilation techniques lead to unphysically high levels of shear and turbulence caused by the algorithm’s inability to cope with inaccuracies in the mesoscale time–height profiles. Modifying the algorithm to account for vertical coherence in the mesoscale source terms gives the microscale solver a greater ability to correct the provided mesoscale time–height profiles, leading to improved predictions of shear and turbulence statistics. The resulting turbulence statistics are in good agreement with meteorological tower observations and simulation results obtained with state-of-the-art coupling techniques using mesoscale budget components.

data assimilation↗