Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “variable selection”

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 487 records · Page 27
All sourcesNASA NTRSDOE OSTI
NASA NTRS1989

A Rocket Engine Design Expert System

The overall structure and capabilities of an expert system designed to evaluate rocket engine performance are described. The expert system incorporates a JANNAF standard reference computer code to determine rocket engine performance and a state of the art finite element computer code to calculate the interactions between propellant injection, energy release in the combustion chamber, and regenerative cooling heat transfer. Rule-of-thumb heuristics were incorporated for the H2-O2 coaxial injector design, including a minimum gap size constraint on the total number of injector elements. One dimensional equilibrium chemistry was used in the energy release analysis of the combustion chamber. A 3-D conduction and/or 1-D advection analysis is used to predict heat transfer and coolant channel wall temperature distributions, in addition to coolant temperature and pressure drop. Inputting values to describe the geometry and state properties of the entire system is done directly from the computer keyboard. Graphical display of all output results from the computer code analyses is facilitated by menu selection of up to five dependent variables per plot.

Davidian, Kenneth J.↗
NASA NTRS1989

An interim reference model for the variability of the middle atmosphere H2O vapor distribution

Water vapor is an important minor constituent in the studies of the middle atmosphere for a variety of reasons, including its role as a source for active HO(y) chemicals and its use in analysis of transport processes. A number of in situ and remote techniques were employed in the determination of water vapor distributions. Two of the more complete data sets were used to develop an interim reference profile. First, there are the seven months of Nimbus 7 limb infrared monitor of the stratosphere (LIMS) data obtained during Nov. 1978 to May 1979 over the range 64S to 84N latitude and from about 100 to 1 mb in the mid-mesosphere at several fixed Northern Hemisphere mid-latitude sites. These two data sets were combined to give a mid-lattitude, interim reference water vapor profile for the entire vertical range of the middle atmosphere and with accuracies of better than 25 percent. The daily variability of stratospheric water vapor profiles about the monthly mean was also established from these data sets for selected months. Information is also provided on the longitudinal variability of LIMS water vapor profiles about the daily, weekly, and monthly zonal means. Generally, the interim reference water vapor profile and its variability are consistent with prevailing ideas about chemistry and transport.

Remsberg, E. E.↗
NASA NTRS1990

Separability of soils in a tallgrass prairie using SPOT and DEM data

An investigation is conducted which uses a canonical transformation technique to reduce the features from SPOT and DEM data and evaluates the statistical separability of several prairie soils from the canonically transformed variables. Both SPOT and DEM data was gathered for a tallgrass prairie near Manhattan, Kansas, and high resolution SPOT satellite images were integrated with DEM data. Two canonical variables derived from training samples were selected and it is suggested that canonically transformed data were superior to combined SPOT and DEM data. High resolution SPOT images and DEM data can be used to aid second-order soil surveys in grasslands.

Su, Haiping↗
NASA NTRS1993

Task planning with uncertainty for robotic systems

In a practical robotic system, it is important to represent and plan sequences of operations and to be able to choose an efficient sequence from them for a specific task. During the generation and execution of task plans, different kinds of uncertainty may occur and erroneous states need to be handled to ensure the efficiency and reliability of the system. An approach to task representation, planning, and error recovery for robotic systems is demonstrated. Our approach to task planning is based on an AND/OR net representation, which is then mapped to a Petri net representation of all feasible geometric states and associated feasibility criteria for net transitions. Task decomposition of robotic assembly plans based on this representation is performed on the Petri net for robotic assembly tasks, and the inheritance of properties of liveness, safeness, and reversibility at all levels of decomposition are explored. This approach provides a framework for robust execution of tasks through the properties of traceability and viability. Uncertainty in robotic systems are modeled by local fuzzy variables, fuzzy marking variables, and global fuzzy variables which are incorporated in fuzzy Petri nets. Analysis of properties and reasoning about uncertainty are investigated using fuzzy reasoning structures built into the net. Two applications of fuzzy Petri nets, robot task sequence planning and sensor-based error recovery, are explored. In the first application, the search space for feasible and complete task sequences with correct precedence relationships is reduced via the use of global fuzzy variables in reasoning about subgoals. In the second application, sensory verification operations are modeled by mutually exclusive transitions to reason about local and global fuzzy variables on-line and automatically select a retry or an alternative error recovery sequence when errors occur. Task sequencing and task execution with error recovery capability for one and multiple soft components in robotic systems are investigated.

Cao, Tiehua↗
NASA NTRS1993

Lunar Base Sitting

Speculation with regard to a permanent lunar base has been with us since Robert Goddard was working on the first liquid-fueled rockets in the 1920's. With the infusion of data from the Apollo Moon flights, a once speculative area of space exploration has become an exciting possibility. A Moon base is not only a very real possibility, but is probably a critical element in the continuation of our piloted space program. This article, originally drafted by World Space Foundation volunteers in conjuction with various academic and research groups, examines some of the strategies involved in selecting an appropriate site for such a lunar base. Site selection involves a number of complex variables, including raw materials for possible rocket propellant generation, hot an cold cycles, view of the sky (for astronomical considerations, among others), geological makeup of the region, and more. This article summarizes the key base siting considerations and suggests some alternatives. Availability of specific resources, including energy and certain minerals, is critical to success.

Staehle, Robert L.↗
NASA NTRS2013

Comparison of Four Precipitation Forcing Datasets in Land Information System Simulations over the Continental U.S.

The NASA Short ]term Prediction Research and Transition (SPoRT) Center in Huntsville, AL is running a real ]time configuration of the NASA Land Information System (LIS) with the Noah land surface model (LSM). Output from the SPoRT ]LIS run is used to initialize land surface variables for local modeling applications at select National Weather Service (NWS) partner offices, and can be displayed in decision support systems for situational awareness and drought monitoring. The SPoRT ]LIS is run over a domain covering the southern and eastern United States, fully nested within the National Centers for Environmental Prediction Stage IV precipitation analysis grid, which provides precipitation forcing to the offline LIS ]Noah runs. The SPoRT Center seeks to expand the real ]time LIS domain to the entire Continental U.S. (CONUS); however, geographical limitations with the Stage IV analysis product have inhibited this expansion. Therefore, a goal of this study is to test alternative precipitation forcing datasets that can enable the LIS expansion by improving upon the current geographical limitations of the Stage IV product. The four precipitation forcing datasets that are inter ]compared on a 4 ]km resolution CONUS domain include the Stage IV, an experimental GOES quantitative precipitation estimate (QPE) from NESDIS/STAR, the National Mosaic and QPE (NMQ) product from the National Severe Storms Laboratory, and the North American Land Data Assimilation System phase 2 (NLDAS ]2) analyses. The NLDAS ]2 dataset is used as the control run, with each of the other three datasets considered experimental runs compared against the control. The regional strengths, weaknesses, and biases of each precipitation analysis are identified relative to the NLDAS ]2 control in terms of accumulated precipitation pattern and amount, and the impacts on the subsequent LSM spin ]up simulations. The ultimate goal is to identify an alternative precipitation forcing dataset that can best support an expansion of the real ]time SPoRT ]LIS to a domain covering the entire CONUS.

Case, Jonathan L.↗
NASA NTRS2015

Multi-Objective Hybrid Optimal Control for Multiple-Flyby Interplanetary Mission Design using Chemical Propulsion

Preliminary design of high‐thrust interplanetary missions is a highly complex process. The mission designer must choose discrete parameters such as the number of flybys and the bodies at which those flybys are performed. For some missions, such as surveys of small bodies, the mission designer also contributes to target selection. In addition, real‐valued decision variables, such as launch epoch, flight times, maneuver and flyby epochs, and flyby altitudes must be chosen. There are often many thousands of possible trajectories to be evaluated. The customer who commissions a trajectory design is not usually interested in a point solution, but rather the exploration of the trade space of trajectories between several different objective functions. This can be a very expensive process in terms of the number of human analyst hours required. An automated approach is therefore very desirable. This work presents such an approach by posing the impulsive mission design problem as a multi‐objective hybrid optimal control problem. The method is demonstrated on several real‐world problems. Two assumptions are frequently made to simplify the modeling of an interplanetary high‐thrust trajectory during the preliminary design phase. The first assumption is that because the available thrust is high, any maneuvers performed by the spacecraft can be modeled as discrete changes in velocity. This assumption removes the need to integrate the equations of motion governing the motion of a spacecraft under thrust and allows the change in velocity to be modeled as an impulse and the expenditure of propellant to be modeled using the time‐independent solution to Tsiolkovsky's rocket equation [1]. The second assumption is that the spacecraft moves primarily under the influence of the central body, i.e. the sun, and all other perturbing forces may be neglected in preliminary design. The path of the spacecraft may then be modeled as a series of conic sections. When a spacecraft performs a close approach to a planet, the central body switches from the sun to that planet and the trajectory is modeled as a hyperbola with respect to the planet. This is known as the method of patched conics. The impulsive and patched‐conic assumptions significantly simplify the preliminary design problem.

Optimization↗
NASA NTRSDate not supplied

Application of the Concepts of Load State, Load Space, and Output Space to Strain-Gage Balances

The concepts of load state, load space, and output space are used to describe elements of the development of a load prediction process for a six–component strain–gage balance. The use of these terms adds clarity during the selection of the independent and dependent variables for the development of the load prediction process. First, the control volume model of a strain–gage balance is reviewed as it helps better understand the role of loads, outputs, and state variables during the load prediction. Then, the description of the absolute load datum in both the load space and the output space is discussed. It is concluded that the use of the difference between a raw output and the natural zero of a balance gage has a key advantage for the definition of the output space. In that case, the origin of the output space will also describe the absolute load datum of the balance. Finally, details of the application of the concepts of load state, load space, and output space to two commonly used balance load prediction methods are discussed.

strain-gage balance↗
NASA NTRSDate not supplied

Assessing Uncertainties and Variability 34 Years of Surface Radiative Fluxes Using GEWEX SRB Release4-IP

The NASA/GEWEX Surface Radiation Budget (SRB) project produces 3-hrly shortwave and longwave surface and top of atmosphere radiative fluxes for the 1983-near present time period. The new Release 4 Integrated Product (IP) uses the newly recalibrated and processed ISCCP HXS product as its primary input for cloud and radiance data, replacing ISCCP DX with a ninefold increase in pixel count (10 km instead of 30 km). This first version retains a 1°x1° resolution enabling intercomparison against previous versions and other data sets such as CERES, but spans 34 years from July 1983 through June 2017 and was announced by Kummerow et al. 2019 (GEWEX News). This new IP product also uses an atmospheric temperature and moisture dataset known as nnHIRS and other parameters such as near surface and skin temperatures from SeaFlux and LandFlux data sets. In addition to the input data improvements, several important algorithm improvements have been made since Release 3. These include recalculated SW atmospheric transmissivities and reflectivities, updated ocean and snow/ice albedos, and variable total solar irradiance consistent with SORCE measurements. The LW code has been updated to improve the optical property treatment for clouds and aerosols are included in this version. Radiative treatment of ice clouds is also improved in the LW. The variable aerosol optical properties for the SW and LW are specified using a detailed aerosol history from the Max Planck Institute Aerosol Climatology (MAC). Here we present an assessment of the LW radiative fluxes and the uncertainty of those fluxes relative to the various inputs to surface SW/LW flux measurements from BSRN and PMEL buoys measurements. We review the validation of the SW and LW fluxes and then in terms of time series and then assess the products in terms of their long-term variability of the surface SW and LW net fluxes compared to multiple other data products including atmospheric reanalysis products. The comparisons of radiative estimates to observations are performed at various temporal scales and aimed at investigation of agreement at longer time averages but accessing potential change in diurnal magnitude and daily variability. Utilizing this uncertainty information, to access long-term variability of surface radiation components at selected region and global scales, considering satellite sampling/calibration “artifacts” as necessary. At the longer time scales, the net SW and net LW the TOA and surface have implications toward closure of the energy budgets at the surface, we assess these compared to other studies on energy budget closure for the same selected global and regional scales.

Paul W Stackhouse↗
DOE OSTI2026

Xanthos-Lake Dataset

The Xanthos-Lake v1.0 dataset provides the input data, trained machine-learning models, and simulation outputs needed to characterize lake water balance, snow and ice conditions, and mixing-layer temperature within the Xanthos global hydrological modeling framework. The dataset supports lake representation across a wide range of lake sizes and hydroclimatic conditions by combining xLSIM, a basin-specific machine-learning emulator of lake snow, ice, ice-cover fraction, and mixing-layer temperature, with the Xanthos-Lake water-balance model. The archive contains NetCDF datasets used to train and evaluate xLSIM, trained model weights, processed meteorological and lake-property inputs, and basin- and lake-category-specific simulation outputs. These materials are organized into four primary data groups, described below. Snowice_model_inputs: Contains the NetCDF input data used to train xLSIM. The xLSIM machine-learning framework uses three lake-based datasets. The meteorological forcing dataset provides monthly relative humidity, specific humidity, surface wind speed, maximum and minimum air temperature, downward longwave and shortwave radiation, snowfall, surface air pressure, and total precipitation. Lake surface area is included as an additional static predictor. The target-state dataset provides lake ice thickness, snow depth, snow cover, and lake mixing-layer temperature, while a companion lake-surface dataset provides the lake ice-cover fraction. Before training, ice thickness and snow depth are converted from meters to centimeters, mixing-layer temperature is converted from kelvin to degrees Celsius and constrained to nonnegative values, and ice-cover fraction is converted from a fraction to a percentage. The predictor variables are normalized using statistics calculated across the selected lakes and time steps. Snowice_model_outputs: Contains the NetCDF outputs generated by xLSIM. For each basin, xLSIM produces a file containing observed and predicted lake-state variables for the training, validation, and testing periods. The modeled variables include lake ice thickness, snow depth, snow cover, mixing-layer temperature, and lake ice-cover fraction. For basins without a sufficiently persistent snow-and-ice signal, the emulator predicts only mixing-layer temperature. The outputs also include training and validation loss histories, the selected model configuration, identifiers of the lakes used in training, and SHAP-based feature-importance information at the global, lake, and seasonal-regime levels. The trained machine-learning model weights are provided separately within the dataset archive. Together, these files support model evaluation and subsequent coupling with the Xanthos-Lake water-balance framework. XanthosLAKES: Contains the NetCDF input data used by the Xanthos-Lake framework. Monthly meteorological inputs include relative and specific humidity, downward shortwave and longwave radiation, mean, maximum, and minimum air temperature, wind speed, precipitation, snowfall, and surface air pressure. Static lake-property datasets provide lake identifiers, geographic locations, surface area, volume, mean depth, elevation, drainage area, fetch, outlet-routing information, and associated Xanthos grid-cell attributes. Separate bathymetric datasets provide the coefficients of the area–depth and volume–depth relationships for each aggregated lake unit. GLEV-based records provide observed lake surface area and evaporation data used to initialize lake states, define reference conditions, and calibrate and evaluate the model. Xanthos-Lake Outputs: Contains the basin- and lake-category-specific NetCDF outputs generated by Xanthos-Lake. Monthly variables include lake surface area, storage volume, outlet discharge, evaporation rate, evaporation volume, lake–groundwater exchange, lake inflow, ice thickness, snow depth, snow-cover fraction, ice-cover fraction, and mixing-layer temperature. The files also contain lake-specific calibration and validation statistics, including normalized root-mean-square error, mean absolute error, Nash–Sutcliffe efficiency, Kling–Gupta efficiency, and percent bias. Stored calibrated and derived parameters include the weir discharge coefficient, fractional freeboard, groundwater exchange coefficient, reference water level, corresponding reference surface area and storage volume, weir-width adjustment factor, and the fraction of routed inflow entering the lake. Basin identifiers, lake category, simulation period, calibration and validation periods, and parameter-schema information are retained as NetCDF metadata.

Abeshu, Guta [Pacific Northwest National Laborator↗
DOE OSTI2024

Geographical Insights into Suicide Mortality Through Spatial Machine Learning

Suicide mortality is a leading cause of death in the United States, with an upward trend that emphasizes its significance as a public health issue. Previous research has employed global models like ordinary least squares (OLS) regression and local models such as geographically weighted regression (GWR). While local models are useful for analyzing spatial variations in suicide mortality, they share limitations with traditional global models, particularly about their inability to handle multi-collinearity and non-linear relationships. Machine learning approaches, like random forests (RF), can address some of these limitations but often fail to account for spatial variability. This gap highlights the need for spatial ML models specifically designed to tackle suicide mortality. This research seeks to fill this void by using a geographically weighted random forest model (GWRF) to examine the associations between county-level suicide mortality in the U.S. from 2010 to 2020 and various social and environmental determinants of health. A key aspect of our methodology is disciplined feature selection, which reduces the pool of explanatory variables by about 90%. This refinement enhances the explanatory power of both global (R2 improved from 0.59 to 0.67) and local (R2 improved from 0.64 to 0.67) RF models while reducing their run times. An analysis of the importance scores for these selected features reveals that the drivers of suicide mortality vary by context. Thus, to effectively address regional disparities and inform targeted public health interventions, a holistic approach that incorporates multiple county-level characteristics is essential.

Lebakula, Viswadeep [ORNL] (ORCID:0000000152935914↗
NASA NTRS1993

The use of designed experiments in the process development of continuous propellant mixing

A continuous mix pilot plant was constructed at Aerojet Propulsion Division in Sacramento, California to develop a robust propellant mixing process for the full scale plant that was to be built at the NASA Advanced Solid Rocket Motor facility Yellow Creek, Mississippi. The plant was used to conduct dozens of subsystem and full system mixing tests for evaluation of equipment, processing methods, and control schemes for later use at the production plant. As a culmination to this work, a series of designed experiments were conducted using an eight run Taguchi analysis with four factors at two levels each to determine the primary effect of processing parameters on propellant ballistic and mechanical properties. The factors examined in these runs included the propellant production rate (454 (1000) and 622 kg/hr (1371 Ib/hr)), the product temperature out of the mixer (49 (120) and 63 deg C (145 deg F)), mixer screw speed (75 and 90 rpm), and the deaerator excess capacity (20 and 80 percent). Measured response variables included the uncured and cured density, Crawford Bomb liquid strand burning rates, and selected mechanical properties. The experiment revealed that several of the response variables displayed significant changes from run-to-run with the product temperature being the single most important factor. After concluding this experiment, a twenty-six hour confirmation run was conducted to verify the conclusions reached in the designed experiment. The extended run produced over 12,250 kgs (27,000 lbs) of propellant meeting all of the pre-run targeted properties including density (1.803 g/cc (0.065 lb/in(exp 3)) with a 0.12 percent coefficient of variation (CV) at 25 deg C (77 deg F)), liquid strand burn rate (0.889 cm/s (0.350 in/s) with a 0.69 percent CV at 4210 KPa (610 psig), 15.6 deg C (60 deg F)), nominal maximum stress (828 KPa (120 psig) with a 2.84 percent CV, S&E at 25 deg C (77 deg F), 5.08 cm/min (2 in/min)), strain at nominal maximum (47.4 percent with a 3.96 percent CV), and initial tangent modulus (5349 KPa (775 psig) with a 7.26 percent CV).

Campbell, J. A.↗
NASA NTRS1985

Supercomputing on massively parallel bit-serial architectures

Research on the Goodyear Massively Parallel Processor (MPP) suggests that high-level parallel languages are practical and can be designed with powerful new semantics that allow algorithms to be efficiently mapped to the real machines. For the MPP these semantics include parallel/associative array selection for both dense and sparse matrices, variable precision arithmetic to trade accuracy for speed, micro-pipelined train broadcast, and conditional branching at the processing element (PE) control unit level. The preliminary design of a FORTRAN-like parallel language for the MPP has been completed and is being used to write programs to perform sparse matrix array selection, min/max search, matrix multiplication, Gaussian elimination on single bit arrays and other generic algorithms. A description is given of the MPP design. Features of the system and its operation are illustrated in the form of charts and diagrams.

Iobst, Ken↗
NASA NTRS2013

Photosynthesis-dependent Isoprene Emission from Leaf to Planet in a Global Carbon-chemistry-climate Model

We describe the implementation of a biochemical model of isoprene emission that depends on the electron requirement for isoprene synthesis into the FarquharBallBerry leaf model of photosynthesis and stomatal conductance that is embedded within a global chemistry-climate simulation framework. The isoprene production is calculated as a function of electron transport-limited photosynthesis, intercellular and atmospheric carbon dioxide concentration, and canopy temperature. The vegetation biophysics module computes the photosynthetic uptake of carbon dioxide coupled with the transpiration of water vapor and the isoprene emission rate at the 30 min physical integration time step of the global chemistry-climate model. In the model, the rate of carbon assimilation provides the dominant control on isoprene emission variability over canopy temperature. A control simulation representative of the present-day climatic state that uses 8 plant functional types (PFTs), prescribed phenology and generic PFT-specific isoprene emission potentials (fraction of electrons available for isoprene synthesis) reproduces 50 of the variability across different ecosystems and seasons in a global database of 28 measured campaign-average fluxes. Compared to time-varying isoprene flux measurements at 9 select sites, the model authentically captures the observed variability in the 30 min average diurnal cycle (R2 6496) and simulates the flux magnitude to within a factor of 2. The control run yields a global isoprene source strength of 451 TgC yr1 that increases by 30 in the artificial absence of plant water stress and by 55 for potential natural vegetation.

photosynthesis↗
DOE OSTI2025

Optimal Complementarity Analysis of Potential Floating Solar Co-Located With Existing Hydropower Assets Across the Contiguous United States

The U.S. is expected to double its rate of renewable capacity from 2024 to 2030. However, the stochastic nature of renewable energy poses challenges to the operation and reliability of our power grid. The combined generation from renewable energy sources, with dispatchable sources (such as hydropower) operating as a hybrid energy plant, could mitigate this variability. In this paper, the complementarity analysis of selected U.S. reservoirs with existing hydropower assets (EHAs) and potential floating photovoltaics (FPVs) is conducted for the continuous U.S. (CONUS). The optimal FPV capacity for each site is determined by minimizing the variability of the combined output, while adhering to the FPV potential. Our results indicate that over 50% of the analyzed reservoirs achieve a stability coefficient exceeding 0.5, leading to a less-variable output after optimization. Finally, we analyze the complementary hydro-FPV hybrid reservoirs by considering both the Pearson correlation coefficient and the stability coefficient on daily, monthly, and yearly scales. Summaries are included of locations of theoretical FPVs co-located with hydropower plants that exhibit high complementarity based on the selected metrics.

13 - HYDRO ENERGY↗
DOE OSTI2022

The Quest for New Correlations in the Realm of the Gamma-Ray Burst—Supernova Connection

Gamma-ray bursts (GRBs) are very energetic cosmological transients. Long GRBs are usually associated with Type Ib/c supernovae (SNe), and we refer to them as GRB-SNe. Since the associated SN for a given GRB is observed only at low redshift, a possible selection effect exists when we consider intrinsically faint sources that cannot be observed at high redshift. Thus, it is important to explore the possible relationships between GRB and SN parameters after these have been corrected for astrophysical biases due to the instrumental selection effects and redshift evolution of the variables involved. So far, only GRB prompt emission properties have been checked against the SN Ib/c properties without considering the afterglow (AG). This work investigates the existence of relationships among GRBs’ prompt and AG and associated SN properties. We investigate 91 bidimensional correlations among the SN and GRB observables before and after their correction for selection biases and evolutionary effects. As a result of this investigation, we find hints of a new correlation with a Pearson correlation coefficient >0.50 and a probability of being drawn by chance <0.05. This correlation is between the luminosity at the end of the GRB optical plateau emission and the rest-frame peak time of the SN. According to this relation, the brightest optical plateaus are accompanied by the largest peak times. This correlation is corrected for selection biases and redshift evolution and may provide new constraints for the astrophysical models associated with the GRB-SN connection.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Seasonal drivers of dissolved oxygen across a tidal creek–marsh interface revealed by machine learning

Abstract Dissolved oxygen (DO) is a key biogeochemical control in coastal systems, and its concentration and drivers vary markedly through time and space. This makes it difficult to accurately represent coastal DO and associated biogeochemical processes in models, limiting our ability to predict how these systems will respond to global change. We obtained high‐frequency (5‐min) in situ measurements of DO collected at three locations across the interface of a tidal creek and coastal marsh in the Pacific Northwest, USA. Random Forest machine learning models quantified the importance of three categories of environmental drivers (Aquatic, Climatic, and Terrestrial) of DO variability across the creek–marsh interface. We selected two 4‐month datasets representing Summer and Winter seasonal periods to test two hypotheses on the dominant drivers of DO at the coastal interface. We found that the Terrestrial driver—characterized by long periods of anaerobic conditions and episodic pulses in DO after floods—was most important during the Winter, whereas the Aquatic driver—characterized by variability over tidal, diel, and lunar cycles—was most important during the Summer. We explored how future climate change scenarios could alter the drivers of DO variability using a cumulative sums driver–response framework. Our results suggest that under climate change, Aquatic and Climatic drivers may increase in importance during the Summer, potentially linked to changing metabolic regimes and sea level, with Terrestrial driver importance potentially increasing during the Winter. Our approach highlights useful methods for understanding the spatiotemporal complexity of oxygen across coastal interfaces and quantifying the relative importance of distinct environmental drivers.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Numerical characterization of support recovery in sparse regression with correlated design

Sparse regression is employed in diverse scientific settings as a feature selection method. A pervasive aspect of scientific data is the presence of correlations between predictive features. These correlations hamper both feature selection and estimation and jeopardize conclusions drawn from estimated models. On the other hand, theoretical results on sparsity-inducing regularized regression have largely addressed conditions for selection consistency via asymptotics, and disregard the problem of model selection, whereby regularization parameters are chosen. In this numerical study, we address these issues through exhaustive characterization of the performance of several regression estimators, coupled with a range of model selection strategies. These estimators and selection criteria were examined across correlated regression problems with varying degrees of signal to noise, distributions of non-zero model coefficients, and model sparsity. Our results reveal a fundamental tradeoff between false positive and false negative control in all regression estimators and model selection criteria examined. Additionally, we numerically explore a transition point modulated by the signal-to-noise ratio and spectral properties of the design covariance matrix at which the selection accuracy of all considered algorithms degrades. Overall, we find that SCAD coupled with BIC or empirical Bayes model selection performs the best feature selection across the regression problems considered.

97 MATHEMATICS AND COMPUTING↗