Search NASA⌕ Search

SEARCH · Search NASA

Results for “biasing”

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 109 records · Page 6
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Mechanisms behind the Springtime North Pacific ENSO Teleconnection Bias in Climate Models

Previous studies have shown that models overestimate the strength of ENSO teleconnections to the North Pacific during springtime, but the underlying reasons for this bias remain unknown. Here, in this work, the relative contributions from basic-state and thermodynamic/dynamic forcing factors are disentangled through idealized experiments with the Community Earth System Model and a range of stationary wave modeling experiments. It is revealed that in CESM1 the diabatic heating biases over the tropical Indian Ocean and tropical central-western Pacific jointly favor a cyclonic (anticyclonic) circulation bias to occur in the North Pacific during the springtime of El Niño (La Niña) events. On one hand, the difference in the modeled and observed climatological basic state does not lead to the bias formation directly, as the diabatic heating biases are the primary cause. On the other hand, the springtime basic state is conducive to a more vigorous stationary wave response to the biased diabatic heating than the wintertime state, and this explains why the teleconnection bias occurs during springtime but not in winter. An iterative bias-correction approach is then implemented in the atmospheric model component of CESM1 to verify the linkage between the tropical diabatic heating bias and the teleconnection bias. Moreover, this explanation is shown to be relevant in other models of phase 5 of the Coupled Model Intercomparison Project (CMIP5) as a strong relationship is found between biases in ENSO-related tropical central-western Pacific/Indian Ocean precipitation and North Pacific circulation across models in spring.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Updating and Evaluating Anthropogenic Emissions for NOAA’s Global Ensemble Forecast Systems for Aerosols (GEFS-Aerosols): Application of an SO 2 Bias-Scaling Method

We updated the anthropogenic emissions inventory in NOAA’s operational Global Ensemble Forecast for Aerosols (GEFS-Aerosols) to improve the model’s prediction of aerosol optical depth (AOD). We used a methodology to quickly update the pivotal global anthropogenic sulfur dioxide (SO 2 ) emissions using a speciated AOD bias-scaling method. The AOD bias-scaling method is based on the latest model predictions compared to NASA’s Modern-Era Retrospective analysis for Research and Applications, version 2 (MERRA2). The model bias was subsequently applied to the CEDS 2019 SO 2 emissions for adjustment. The monthly mean GEFS-Aerosols AOD predictions were evaluated against a suite of satellite observations (e.g., MISR, VIIRS, and MODIS), ground-based AERONET observations, and the International Cooperative for Aerosol Prediction (ICAP) ensemble results. The results show that transitioning from CEDS 2014 to CEDS 2019 emissions data led to a significant improvement in the operational GEFS-Aerosols model performance, and applying the bias-scaled SO 2 emissions could further improve global AOD distributions. The biases of the simulated AODs against the observed AODs varied with observation type and seasons by a factor of 3~13 and 2~10, respectively. The global AOD distributions showed that the differences in the simulations against ICAP, MISR, VIIRS, and MODIS were the largest in March–May (MAM) and the smallest in December–February (DJF). When evaluating against the ground-truth AERONET data, the bias-scaling methods improved the global seasonal correlation (r), Index of Agreement (IOA), and mean biases, except for the MAM season, when the negative regional biases were exacerbated compared to the positive regional biases. The effect of bias-scaling had the most beneficial impact on model performance in the regions dominated by anthropogenic emissions, such as East Asia. However, it showed less improvement in other areas impacted by the greater relative transport of natural emissions sources, such as India. The accuracies of the reference observation or assimilation data for the adjusted inputs and the model physics for outputs, and the selection of regions with less seasonal emissions of natural aerosols determine the success of the bias-scaling methods. A companion study on emission scaling of anthropogenic absorbing aerosols needs further improved aerosol prediction.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1995

Measurements of electromagnetic bias at Ku and C bands

The electromagnetic (EM) bias epsilon is an error present in radar altimetry of the ocean surface due to nonuniform reflection with surface displacement. The electromagnetic bias is defined as the difference in height between the mean reflecting surface and the mean sea surface. A knowledge of the electromagnetic bias is required for reducing errors in mean sea level measurements by satellite radar altimeters. Direct measurements of the EM bias at 14 GHz (Ku band) and 5 GHz (C band) were made from an oil production platform in the Gulf of Mexico over a 6-month period during 1989 and 1990. A total of 1280 hours of usable data was collected. During the experiment the significant wave height (SWH) varied from 0.6 to 3.2 m; the wind speed at 25 m above the surface varied from 0.1 to 14.3 m/s; the Ku band bias varied from -1.0 to -13.8 cm, or from -1.6% to -5.3% of the SWH; and the C band bias varied from -0.4 to -19.9 cm, or from -0.6% to -6.3% of the SWH. The biases has mean values of -3.7% and -3.6% of SWH with standard deviations of the variability about the mean of 0.7% and 1.0% of the SWH for Ku and C bands, respectively. We found a nonlinear relationship between dimensionless bias (bias/SWH) and wind speed at both low and high wind speeds. For wind speeds less than 3-4 m/s but less than 10 m/s, both biases were found to increase linearly with wind speed. For wind speeds greater than 11-12 m/s, the C band bias reaches a maximum. The Ku band bias reaches a maximum and then begins to decrease for wind speeds greater than 9-10 m/s.

Arnold, D. V.↗
NASA NTRS2014

Estimating Sampling Biases and Measurement Uncertainties of AIRS-AMSU-A Temperature and Water Vapor Observations Using MERRA Reanalysis

We use MERRA (Modern Era Retrospective-Analysis for Research Applications) temperature and water vapor data to estimate the sampling biases of climatologies derived from the AIRS/AMSU-A (Atmospheric Infrared Sounder/Advanced Microwave Sounding Unit-A) suite of instruments. We separate the total sampling bias into temporal and instrumental components. The temporal component is caused by the AIRS/AMSU-A orbit and swath that are not able to sample all of time and space. The instrumental component is caused by scenes that prevent successful retrievals. The temporal sampling biases are generally smaller than the instrumental sampling biases except in regions with large diurnal variations, such as the boundary layer, where the temporal sampling biases of temperature can be +/- 2 K and water vapor can be 10% wet. The instrumental sampling biases are the main contributor to the total sampling biases and are mainly caused by clouds. They are up to 2 K cold and greater than 30% dry over mid-latitude storm tracks and tropical deep convective cloudy regions and up to 20% wet over stratus regions. However, other factors such as surface emissivity and temperature can also influence the instrumental sampling bias over deserts where the biases can be up to 1 K cold and 10% wet. Some instrumental sampling biases can vary seasonally and/or diurnally. We also estimate the combined measurement uncertainties of temperature and water vapor from AIRS/AMSU-A and MERRA by comparing similarly sampled climatologies from both data sets. The measurement differences are often larger than the sampling biases and have longitudinal variations.

MERRA↗
DOE OSTI2016

A new method to measure galaxy bias by combining the density and weak lensing fields

We present a new method to measure redshift-dependent galaxy bias by combining information from the galaxy density field and the weak lensing field. This method is based on the work of Amara et al., who use the galaxy density field to construct a bias-weighted convergence field κg. The main difference between Amara et al.'s work and our new implementation is that here we present another way to measure galaxy bias, using tomography instead of bias parametrizations. The correlation between κg and the true lensing field κ allows us to measure galaxy bias using different zero-lag correlations, such as <κgκ>/<κκ> or <κgκg>/<κgκ>. Our method measures the linear bias factor on linear scales, under the assumption of no stochasticity between galaxies and matter. We use the Marenostrum Institut de Ciències de l'Espai (MICE) simulation to measure the linear galaxy bias for a flux-limited sample (i < 22.5) in tomographic redshift bins using this method. This article is the first that studies the accuracy and systematic uncertainties associated with the implementation of the method and the regime in which it is consistent with the linear galaxy bias defined by projected two-point correlation functions (2PCF). We find that our method is consistent with a linear bias at the per cent level for scales larger than 30 arcmin, while non-linearities appear at smaller scales. This measurement is a good complement to other measurements of bias, since it does not depend strongly on σ8 as do the 2PCF measurements. We will apply this method to the Dark Energy Survey Science Verification data in a follow-up article.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Bias-Variance Trade-Off in Physics-Informed Neural Networks with Randomized Smoothing for High-Dimensional PDEs

Physics-Informed Neural Networks (PINNs) have triggered a paradigm shift in scientific computing, leveraging mesh-free properties and robust approximation capabilities. While proving effective for low-dimensional partial differential equations (PDEs), the computational cost of PINNs remains a hurdle in high-dimensional scenarios. This is particularly pronounced when computing high-order and high-dimensional derivatives in the physics-informed loss. Randomized Smoothing PINN (RS-PINN) introduces Gaussian noise for stochastic smoothing of the original neural net model, enabling the use of Monte Carlo methods for derivative approximation, which eliminates the need for costly automatic differentiation. Despite its computational efficiency, especially in the approximation of high-dimensional derivatives, RS-PINN introduces biases in both loss and gradients, negatively impacting convergence, especially when coupled with stochastic gradient descent (SGD) algorithms. We present a comprehensive analysis of biases in RS-PINN, attributing them to the nonlinearity of the Mean Squared Error (MSE) loss as well as the intrinsic nonlinearity of the PDE itself. We propose tailored bias correction techniques, delineating their application based on the order of PDE nonlinearity. The derivation of an unbiased RS-PINN allows for a detailed examination of its advantages and disadvantages compared to the biased version. Specifically, the biased version has a lower variance and runs faster than the unbiased version, but it is less accurate due to the bias. To optimize the bias-variance trade-off, we combine the two approaches in a hybrid method that balances the rapid convergence of the biased version with the high accuracy of the unbiased version. In addition to methodological contributions, we present an enhanced implementation of RS-PINN. Extensive experiments on diverse high-dimensional PDEs, including Fokker-Planck, Hamilton-Jacobi-Bellman (HJB), viscous Burgers’, Allen-Cahn, and Sine-Gordon equations, illustrate the bias-variance trade-off and highlight the effectiveness of the hybrid RS-PINN. Empirical guidelines are provided for selecting biased, unbiased, or hybrid versions, depending on the dimensionality and nonlinearity of the specific PDE problem.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Analysis of Integrated Vapor Transport Biases

Abstract High-fidelity analyses and forecasts of integrated vapor transport (VT) are central to the study of Earth’s hydrological cycle as well as high-impact phenomena such as monsoons and atmospheric rivers. The impact of the in-line analysis correction-based additive inflation (ACAI) on IVT biases and forecast errors is examined within the Navy Earth System Prediction Capability (Navy ESPC) global coupled system. The ACAI technique uses atmospheric analysis corrections from the data assimilation system to approximate model bias and as a representation of stochastic model error to simultaneously reduce systematic and random errors and improve ensemble performance. ACAI reduces the global average magnitude of the 7- and 14-day IVT bias by 16%–17% during Northern Hemisphere summer, reaching 70% reductions in some tropical regions. The global average IVT bias reduction is similar to the bias reduction for low-level wind speed bias and considerably smaller than the bias reduction in total precipitable water. The localized regions where ACAI increases IVT bias occur where the control IVT biases change sign and structure with increasing forecast lead time, such as the South Asian monsoon region. Substituting analyzed wind or moisture fields for the forecast fields when calculating the forecast IVT confirms that, on average, wind errors dominate the IVT error calculation in the tropics, although wind and moisture error contributions are comparable in the extratropics. The existence of regions where using either analyzed winds or analyzed moisture increases IVT bias or mean absolute error reveals areas with compensating errors.

Reynolds, Carolyn A.↗
DOE OSTI2022

Quantifying albedo susceptibility biases in shallow clouds

The evaluation of radiative forcing associated with aerosol–cloud interactions remains a significant source of uncertainty in future climate projections. The problem is confounded by the fact that aerosol particles influence clouds locally and that averaging to larger spatial and/or temporal scales carries biases that depend on the heterogeneity and spatial correlation of the interacting fields and the nonlinearity of the responses. Mimicking commonly applied satellite data analyses for calculation of albedo susceptibility S o , we quantify S o aggregation biases using an ensemble of 127 large eddy simulations of marine stratocumulus. We explore the cloud field properties that control this spatial aggregation bias and quantify the bias for a large range of shallow stratocumulus cloud conditions manifesting a variety of morphologies and ranges of cloud fractions. We show that S o spatial aggregation biases can be on the order of hundreds of percent, depending on the methodology. Key uncertainties emanate from the typically applied adiabatic drop concentration N d retrieval, the correlation between aerosol and cloud fields, and the extent to which averaging reduces the variance in cloud albedo A c and N d . So biases are more often positive than negative and are highly correlated with biases in the liquid water path adjustment. Temporal aggregation biases are shown to offset spatial aggregation biases. Both spatial and temporal biases have significant implications for observationally based assessments of aerosol indirect effects and our inferences of underlying aerosol–cloud–radiation effects.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS2013

Simultaneous Estimation of Model State Variables and Observation and Forecast Biases Using a Two-Stage Hybrid Kalman Filter

In this paper, we present a two-stage hybrid Kalman filter to estimate both observation and forecast bias in hydrologic models, in addition to state variables. The biases are estimated using the discrete Kalman filter, and the state variables using the ensemble Kalman filter. A key issue in this multi-component assimilation scheme is the exact partitioning of the difference between observation and forecasts into state, forecast bias and observation bias updates. Here, the error covariances of the forecast bias and the unbiased states are calculated as constant fractions of the biased state error covariance, and the observation bias error covariance is a function of the observation prediction error covariance. In a series of synthetic experiments, focusing on the assimilation of discharge into a rainfall-runoff model, it is shown that both static and dynamic observation and forecast biases can be successfully estimated. The results indicate a strong improvement in the estimation of the state variables and resulting discharge as opposed to the use of a bias-unaware ensemble Kalman filter. Furthermore, minimal code modification in existing data assimilation software is needed to implement the method. The results suggest that a better performance of data assimilation methods should be possible if both forecast and observation biases are taken into account.

Data Assimilation↗
NASA NTRS2021

Seasonal Bias in Global Ocean Color Observations

In this study we identify a seasonal bias in the ocean color satellite derived remote sensing reflectances (Rrs(λ); sr-1) at the ocean color validation site, MOBY (Marine Optical BuoY). The seasonal bias in Rrs(λ) is present to varying degrees in all ocean color satellites examined, including VIIRS (Visible Infrared Imaging Radiometer Suite), SeaWiFS (Sea-viewing Wide Field-of-view Sensor), and MODIS (Moderate Resolution Imaging Spectrometer). The relative bias in Rrs has spectral dependence. Products derived from Rrs(λ) are affected by the bias to varying degrees, with particulate backscattering varying up to 50% over a year, chlorophyll varying up to 25% over a year, and absorption from phytoplankton or dissolved material varying by up to 15%. The propagation of Rrs(λ) bias into derived products is broadly confirmed on regional and global scales using Argo floats and data from the CALIOP (Cloud-Aerosol Lidar with Orthogonal Polarization) instrument aboard the CALIPSO (Cloud-Aerosol Lidar and Infrared Pathfinder) satellite. The artifactual seasonality in ocean color is prominent in areas of low biomass (i.e., subtropical gyres) and is not easily discerned in areas of high biomass. While we have eliminated several candidates that could cause the biases in Rrs(λ), there are still outstanding questions regarding potential contributions from atmospheric corrections. Specifically, we provide evidence that the aquatic bidirectional reflectance distribution function may in part cause the observed seasonal bias, but this does not preclude an additional effect of the aerosol estimation. Our investigation highlights the contributions that atmospheric correction schemes can make in introducing biases in Rrs (λ) and we recommend more simulations to discern these influence Rrs (λ) biases. Community efforts are needed to find the root cause of the seasonal bias because all past, present, and future data are or will be affected until a solution is implemented.

satellite ocean color↗
NASA NTRSDate not supplied

Bias in Planning Algorithms

Does bias exist in planning algorithms? If so, how does bias manifest, and how important is this bias? Answering this question requires a formal, mathematical definition of bias. We formally define bias as the distance between the probability distributions of solutions returned by various algorithms, and the uniform distribution over solutions. We show in this paper that deterministic algorithms are inherently biased, as they don’t return all solutions, and that this property holds even when algorithms return a set of plans instead of just one plan. Exceptions are problem instances or problem classes for which only a single solution exists. We then discuss changing the definition of bias to compare the probability distributions of properties of sets of plans instead of individual plans. We show the property bias is smaller than the bias of actual plans. Finally, we show that entropy is a proxy for the more complex and more expensive distance measurement between pairs of probability distributions. We then describe a roadmap for future investigations of bias in planning.

Planning Scheduling Algorithms↗
NASA NTRS2024

Quantifying Volumetric Scattering Bias in ICESat-2 and Operation IceBridge Altimetry Over Greenland Firn and Aged Snow

The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) mission has collected surface elevation measurements for over 5 years. ICESat-2 carries an instrument that emits laser light at 532 nm, and ice and snow absorb weakly at this wavelength. Previous modeling studies found that melting snow could induce significant bias to altimetry signals, but there is no formal assessment on ICESat-2 acquisitions during the melting season. We performed two case studies over the Greenland Ice Sheet to quantify bias in ICESat-2 signals over snow: one to validate Airborne Topographic Mapper (ATM) data against Next Generation Airborne Visible/Infrared Imaging Spectrometer (AVIRIS-NG) grain sizes, and a second to estimate ICESat-2 bias relative to ATM. We used snow optical grain sizes derived from ATM and AVIRIS-NG to attribute altimetry bias to snowpack properties. For the first case study, the mean and standard deviation of optical grain sizes were 340 ± 65 µm (AVIRIS-NG) and 670 ± 420 µm (ATM). A mean altimetry bias of 4.81 ± 1.76 cm was found for ATM, with larger biases linked to increases in grain size. In the second case study, we found a mean grain size of 910 ± 381 µm and biases of 6.42 ± 1.77 cm (ICESat-2) and 9.82 ± 0.97 cm (ATM). The grain sizes and densities needed to recreate biases with a model are uncommon in nature, so we propose that additional surface attributes must be considered to characterize ICESat-2 bias over snow. The altimetry biases are within the accuracy requirements of the ICESat-2 mission, but we cannot rule out more significant errors over coarse-grained snow.

Zachary Fair↗
DOE OSTI2024

Causes of and Solutions to Wind Speed Bias in NREL's 2020 Offshore Wind Resource Assessment for the California Pacific Outer Continental Shelf

This report provides the results of a detailed analysis into the causes of high wind speed bias in the 20-year wind resource data set for offshore California the National Renewable Energy Laboratory (NREL) released in 2020, herein called CA20. The data set was developed using the state-of-the-art Weather Research and Forecasting (WRF) model. Notably, no floating lidars were available at the time in offshore California to validate offshore hub-height wind speeds. In late 2020, the Pacific Northwest National Laboratory (PNNL) deployed two floating lidars in the California outer continental shelf (OCS), near the Bureau of Ocean Energy Management (BOEM) call areas of Humboldt and Morro Bay. Using these observations through 2021, NREL found considerable bias in modeled hub-height winds at both locations: up to +2 m/s at Humboldt over a 6-month period, and up to +1 m/s at Morro Bay over a one-year period. Upon the discovery of this bias, the Department of Energy (DOE) and BOEM funded NREL and PNNL to investigate the causes of, impacts of, and solutions to the bias in the CA20 data set. This report summarizes the findings of this research. We first investigated whether different WRF model setups could lead to reduced bias. We found that the choice of planetary boundary layer (PBL) scheme - which controls the vertical turbulent mixing of momentum, heat, and moisture in the lowermost part of the atmosphere - greatly affected hub-height wind speeds in the region. Specifically, switching from the Mellor-Yamada-Nakanishi-Niino (MYNN) scheme used in CA20 (and widely used across a range of operational and research weather models) to the less common Yonsei University (YSU) scheme nearly eliminated the bias at both the Humboldt and Morro Bay lidar locations. The large discrepancy between the MYNN- and YSU-modeled hub-height winds pointed towards the role of atmospheric stability. In general, PBL schemes agree well in conditions of high turbulence and mixing, normally referred to as "unstable" conditions. By contrast, PBL schemes start to diverge in "stable" conditions, where turbulence is low and thermal stratification (i.e., higher temperature air sitting on top of colder air) greatly suppresses vertical mixing. Under such conditions, winds aloft can decouple from surface effects and greatly accelerate, causing high wind speeds at hub-height and frequent low-level jets (LLJs). We determined that these stable conditions are in fact dominant in offshore California. The region is characterized by moderate-to-extreme stable stratification with a LLJ on average around 200 meters above sea-level. To our knowledge, no wind energy area globally has as strongly stable stratification as offshore California. Under these extreme conditions, we determined that the MYNN scheme models higher stability than YSU, resulting in less vertical turbulent mixing than YSU, allowing for the acceleration of hub-height winds, more intense LLJs, and higher-amplitude inertial oscillations. Using surface observations, we found that MYNN overestimates near-surface stability, whereas YSU tends to model stability better. We then considered several short-term case studies to assess additional meteorological drivers of the bias at Humboldt. We found that during synoptic scale northerly flows driven by the North Pacific High and inland thermal low, a coastal warm bias in the MYNN case studies contributes to the modeled wind speed bias by altering the boundary layer thermodynamics via a thermal wind mechanism. Given the strong performance of the YSU-based runs in offshore California, NREL has produced and published an updated version of the CA20 data set with YSU as the PBL scheme. This updated data set is now part of NREL's 2023 National Offshore Wind (NOW-23) data set, which covers all the U.S. offshore waters. The development and final validation of the NOW-23 data set in offshore California is documented in this report.

17 WIND ENERGY↗
DOE OSTI2018

CAUSES: Diagnosis of the Summertime Warm Bias in CMIP5 Climate Models at the ARM Southern Great Plains Site

Abstract All the weather and climate models participating in the Clouds Above the United States and Errors at the Surface project show a summertime surface air temperature (T2 m) warm bias in the region of the central United States. To understand the warm bias in long‐term climate simulations, we assess the Atmospheric Model Intercomparison Project simulations from the Coupled Model Intercomparison Project Phase 5, with long‐term observations mainly from the Atmospheric Radiation Measurement program Southern Great Plains site. Quantities related to the surface energy and water budget, and large‐scale circulation are analyzed to identify possible factors and plausible links involved in the warm bias. The systematic warm season bias is characterized by an overestimation of T2 m and underestimation of surface humidity, precipitation, and precipitable water. Accompanying the warm bias is an overestimation of absorbed solar radiation at the surface, which is due to a combination of insufficient cloud reflection and clear‐sky shortwave absorption by water vapor and an underestimation in surface albedo. The bias in cloud is shown to contribute most to the radiation bias. The surface layer soil moisture impacts T2 m through its control on evaporative fraction. The error in evaporative fraction is another important contributor to T2 m. Similar sources of error are found in hindcast from other Clouds Above the United States and Errors at the Surface studies. In Atmospheric Model Intercomparison Project simulations, biases in meridional wind velocity associated with the low‐level jet and the 500 hPa vertical velocity may also relate to T2 m bias through their control on the surface energy and water budget.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2019

Spillover as a cause of bias in baseline evaluation methods for demand response programs

Prior research has shown load reduction estimates from residential event-driven demand response programs (e.g., Critical Peak Pricing)using X of the highest Y days with a weather adjustment method are the best performing within the class of currently used baseline methods. However, they are still biased relative to estimates produced from randomized control trials (RCTs), the unbiased “gold standard” evaluation method. In this paper we identify underlying factors that cause some of the bias found in one commonly used baseline method. Rather than simply quantifying bias, our research provides a deeper understanding of what causes the bias and thus can be used to develop more accurate methods that are not subject to these underlying factors. Previous studies have compared various baseline methods relative to each other; however, because all baseline methods are biased, it is impossible to determine the true bias that exists in them. We have access to a unique RCT dataset: the Sacramento Municipal Utility District's study of critical peak pricing. Our analysis of 23 event days over two summers allows us to identify the true bias on load reduction estimates by using the RCT estimates as the unbiased gold standard against which we compare the estimates from the baseline methods. We found that spillover of energy reductions, from hours targeted by a program onto other hours, is one underlying factor that is a major cause of bias in baseline methods. We discuss alternative baseline methods that may not be subject to this same bias.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

Reduction of Pacific Double‐ITCZ Bias by Convection Parameterization in NCAR CESM2.2

Abstract The impact of convective closure on the double‐ITCZ bias in the NCAR CESM2.2 is investigated in this study. The standard CESM2.2 simulates a remarkable double‐ITCZ bias in the central and eastern Pacific, especially in boreal winter and spring. Modifications to the closure in convection parameterization scheme greatly reduce the double‐ITCZ bias in all seasons, demonstrating that convection parameterization can substantially influence the double‐ITCZ bias in CESM2.2. Further analyses suggest that convection parameterization can modulate the tropical atmosphere‐ocean feedback processes, through which it influences the SST in the southern ITCZ region and hence the double‐ITCZ bias. The changes in the upper ocean temperature advection induced by modified convective closure plays important roles in reducing the warm SST bias and double‐ITCZ precipitation bias in the southern ITCZ region. The modified convective closure improves the low‐level cloud and shortwave cloud radiative forcing in the southeastern Pacific. However, surface heat flux plays only a limited role in reducing warm SST bias and double ITCZ bias because the impacts of shortwave radiation changes are largely canceled by changes in longwave radiation and latent heat flux.

Song, Xiaoliang [Scripps Institution of Oceanograp↗
DOE OSTI2022

Photometric redshift estimation with convolutional neural networks and galaxy images: Case study of resolving biases in data-driven methods

Deep-learning models have been increasingly exploited in astrophysical studies, but these data-driven algorithms are prone to producing biased outputs that are detrimental for subsequent analyses. In this work, we investigate two main forms of biases: class-dependent residuals, and mode collapse. We do this in a case study, in which we estimate photometric redshift as a classification problem using convolutional neural networks (CNNs) trained with galaxy images and associated spectroscopic redshifts. We focus on point estimates and propose a set of consecutive steps for resolving the two biases based on CNN models, involving representation learning with multichannel outputs, balancing the training data, and leveraging soft labels. The residuals can be viewed as a function of spectroscopic redshift or photometric redshift, and the biases with respect to these two definitions are incompatible and should be treated individually. We suggest that a prerequisite for resolving biases in photometric space is resolving biases in spectroscopic space. Experiments show that our methods can better control biases than benchmark methods, and they are robust in various implementing and training conditions with high-quality data. Our methods hold promises for future cosmological surveys that require a good constraint of biases, and they may be applied to regression problems and other studies that make use of data-driven models. Nonetheless, the bias-variance tradeoff and the requirement of sufficient statistics suggest that we need better methods and optimized data usage strategies.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

Transient plasma potential in pulsed dual frequency inductively coupled plasmas and effect of substrate biasing

An electron emitting probe in saturated floating potential mode has been used to investigate the temporal evolution of plasma potential and the effect of substrate RF biasing on it for pulsed dual frequency (2 MHz/13.56 MHz) inductively coupled plasma (ICP) source. The low frequency power (P{sub 2MHz}) has been pulsed at 1 KHz and a duty ratio of 50%, while high frequency power (P{sub 13.56MHz}) has been used in continuous mode. The substrate has been biased with a separate bias power at (P{sub 12.56MHz}) Argon has been used as a discharge gas. During the ICP power pulsing, three distinct regions in a typical plasma potential profile, have been identified as ‘initial overshoot’, pulse ‘on-phase’ and pulse ‘off-phase’. It has been found out that the RF biasing of the substrate significantly modulates the temporal evolution of the plasma potential. During the initial overshoot, plasma potential decreases with increasing RF biasing of the substrate, however it increases with increasing substrate biasing for pulse ‘on-phase’ and ‘off-phase’. An interesting structure in plasma potential profile has also been observed when the substrate bias is applied and its evolution depends upon the magnitude of bias power. The reason of the evolution of this structure may be the ambipolar diffusion of electron and its dependence on bias power.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗