Laboratory Data Sets Research Articles

A probabilistic model on streamwise velocity profile in open channels using Tsallis relative entropy theory

The present study uses the concept of information entropy to derive a velocity distribution model in open channel flow. The Tsallis relative entropy theory is explored for that purpose, where the prior probability density function (PDF) is chosen from the maximum Tsallis entropy distribution. Both the cases of the fixed and varying entropy index are considered from the literature. The velocity equations are derived analytically using the homotopy analysis method (HAM) together with the Padé approximation technique in a general framework. The approximations and the HAM-based series solution are verified against laboratory and field data sets. Also, the prediction accuracies of the proposed models are assessed through the measures of statistical errors. It is seen that the model corresponding to the varying index is superior to the other model. The entropy index of Tsallis relative entropy is considered a variable, and the non-unity values of this index justify the applicability of the entropy function. Moreover, it is observed that one of the Lagrange multipliers corresponding to the velocity model with a varying index becomes negligible subject to the data sets considered, and hence, it simplifies the model development mathematically. The proposed approach can be further extended to develop models for estimating the suspended sediment concentration and shear stress distribution.

A new correlation to determine the Lockhart-Martinelli parameter from vertical differential pressure for horizontal venturi tube over-reading correction

Venturi tubes are widely used to measure the gas flow rate in a wet-gas stream. To calculate the gas flow rate, wet-gas Venturi tubes require the liquid loading (i.e. Lockhart-Martinelli parameter) to be known as an input to a so-called over-reading correction correlation. It is challenging to derive the Lockhart-Martinelli parameter without installing additional devices in series to the Venturi tube thus adding significant cost and complexity. A relatively low cost and easy to implement method is to derive the Lockhart-Martinelli parameter by measuring the pressure drop across the Venturi tube or along a pipe section. A correlation that predicts the Lockhart-Martinelli parameter by measuring the pressure drop along a vertical pipe section downstream of the Venturi tube is presented. The correlation was obtained by fitting experimental results from TUV-SUD National Engineering Laboratory's wet-gas test facility with nitrogen-water in a 4″ vertical pipe, Lockhart-Martinelli from 0.04 to 0.29, and gas Froude number from 1 to 2.7. Further experiments were conducted at Spirax Sarco Engineering plc to evaluate the performance of the proposed correlation under steam-water conditions. Tests were conducted with steam-water in a vertical 2″ pipe, Lockhart-Martinelli from 0.005 to 0.064, and gas Froude number from 0.55 to 1.35. The Lockhart-Martinelli parameter is predicted within ±0.035 absolute value with the proposed new correlation for both the National Engineering Laboratory and Spirax Sarco datasets. Gas flow rate values within approximately ±5% error are obtained with the proposed correlation together with ISO/TR 11583 over-reading correlation for the NEL dataset. Although further improvements to the proposed method are required, this work demonstrates the efficacy of an agile, simple, and low-cost differential pressure method to aid the Venturi meter to determine the gas mass flow rate in wet-gas flow.

Evaluating Impacts between Laboratory and Field-Collected Datasets for Plant Disease Classification

Deep learning with convolutional neural networks represents the most used approach in recent years in the classification of leaves’ diseases. The literature has extensively addressed the problem using laboratory-acquired datasets with a homogeneous background. In this article, we explore the variability factors that influence the classification of plant diseases by analyzing the same plant and disease under different conditions, i.e., in the field and in the laboratory. Two plant species and five biotic stresses are analyzed using different architectures, such as EfficientB0, MobileNetV2, InceptionV2, ResNet50 and VGG16. Experiments show that model performance drops drastically when using representative datasets, and the features learned from the network to determine the class do not always belong to the leaf lesion. In the worst case, the accuracy drops from 92.67% to 54.41%. Our results indicate that while deep learning is an effective technique, there are some technical issues to consider when applying it to more representative datasets collected in the field.

The ion–ion recombination coefficient α: comparison of temperature- and pressure-dependent parameterisations for the troposphere and stratosphere

Abstract. Many different atmospheric, physical, and chemical processes are affected by ions. An important sink for atmospheric ions is the reaction and mutual neutralisation of a positive and negative ion, also called ion–ion recombination. While the value for the ion–ion recombination coefficient α is well-known for standard conditions (namely 1.7 × 10−6 cm3 s−1), it needs to be calculated for deviating temperature and pressure conditions, especially for applications at higher altitudes of the atmosphere. In this work, we review the history of theories and parameterisations of the ion–ion recombination coefficient, focussing on the temperature and pressure dependencies as well as the altitude range between 0 and 50 km. Commencing with theories based on J. J. Thomson's work, we describe important semi-empirical adjustments as well as field, model, and laboratory data sets, followed by short reviews of binary recombination theories, model simulations, and the application of ion–aerosol theories to ion–ion recombination. We present a comparison between theories, parameterisations, and field, model, and laboratory data sets to conclude favourable parameterisations. While many theories agree well with field data above an altitude of approximately 10 km, the nature of the recombination coefficient is still widely unknown between Earth's surface and an altitude of 10 km. According to the current state of knowledge, it appears reasonable to assume an almost constant value for the recombination coefficient for this region, while it is necessary to use values that are adjusted for pressure and temperature for altitudes above 10 km. Suitable parameterisations for different altitude ranges are presented and the need for future research, be it in the laboratory or by means of modelling, is identified.

Developing a Lagrangian sediment transport model for open channel flows

A three-dimensional stochastic Lagrangian particle tracking sediment transport model is developed to solve the discrete advection–dispersion equation using a combination of empirical dispersion equations. The performance of three widely-used longitudinal dispersion coefficient equations was examined to select one of them as the primary dispersion equation term in the developed model. Also, a conditional empirical equation was used to consider the effect of vertical dispersion term in top layers near the water surface. The performance of particle tracking model (PTM) to calculate the sediment concentration was evaluated for various sediment classes (very fine sand, fine sand, and medium sand) using available laboratory dataset. Multiple statistical measures were calculated using the Taylor diagram for each dispersion equation. Based on the result, the developed particle tracking model estimated the suspended sediment concentration in a rectangular open channel with a correlation coefficient (R) of 0.96, standard deviation (STD) of 0.262, and root mean square deviation (RMSD) of 0.06 for three different sediment classes proving the acceptable accuracy of the developed model for different range of sediment gradations. In addition, the accuracy of PTM was compared to recently developed models. Comparison between the particle tracking model and the analytical solution of the advection–dispersion equation showed that the developed model predicted the maximum concentration of suspended sediment 6% and 9.4% lower than the analytical solution result for 30 and 50 s simulation time in the straight channel, respectively. However, there was a good agreement between the longitudinal and transverse sediment concentration distribution from the model and the analytical solution approach. The result of the sensitivity analysis and validation process showed that the model could be used to simulate sediment transport in open channel flows.

A Customer Behavior Recognition Method for Flexibly Adapting to Target Changes in Retail Stores.

To provide analytic materials for business management for smart retail solutions, it is essential to recognize various customer behaviors (CB) from video footage acquired by in-store cameras. Along with frequent changes in needs and environments, such as promotion plans, product categories, in-store layouts, etc., the targets of customer behavior recognition (CBR) also change frequently. Therefore, one of the requirements of the CBR method is the flexibility to adapt to changes in recognition targets. However, existing approaches, mostly based on machine learning, usually take a great deal of time to re-collect training data and train new models when faced with changing target CBs, reflecting their lack of flexibility. In this paper, we propose a CBR method to achieve flexibility by considering CB in combination with primitives. A primitive is a unit that describes an object’s motion or multiple objects’ relationships. The combination of different primitives can characterize a particular CB. Since primitives can be reused to define a wide range of different CBs, our proposed method is capable of flexibly adapting to target CB changes in retail stores. In experiments undertaken, we utilized both our collected laboratory dataset and the public MERL dataset. We changed the combination of primitives to cope with the changes in target CBs between different datasets. As a result, our proposed method achieved good flexibility with acceptable recognition accuracy.

Intelligent Intrusion Detection Scheme for Smart Power-Grid Using Optimized Ensemble Learning on Selected Features

The smart grid has gained a reputation as the advanced paradigm of the power grid. It is a complicated cyber-physical system that combines information and communication technology (ICT) with a traditional grid that can remotely control operations. It provides the medium for exchanging real-time data between the company and users through the advanced metering infrastructure (AMI) and smart meters. However, smart grids have many security and privacy concerns, such as intruding sensitive data, firmware hijacking, and modifying data due to the high reliance on ICT. To protect the power-grid system from these counteracts and for reliable and efficient power distribution, early and accurate identification of these issues needs to be addressed. The intrusion detection in a smart grid system plays an essential role in providing a secure service and transmitting the high priority alert message to the system admin about the detection of adversary attacks. This paper proposes an intelligent intrusion detection scheme to accurately classify various attacks on smart power grid systems. The proposed scheme used the binary grey wolf optimization-based feature selection. It optimized the ensemble classification approach to learn the non-linear, overlapping, and complex electrical grid features taken from publicly available Mississippi State University and Oak Ridge National Laboratory (MSU-ORNL) dataset. The experimental results using a 10-fold cross-validation setup and selected feature subset for two class and three class problems reveal the proposed method's promising performance. Further, the significantly superior performance compared to the existing benchmark methods justified the robustness of the proposed scheme.

A comparison of distributed machine learning methods for the support of "many labs" collaborations in computational modeling of decision making.

Deep learning models are powerful tools for representing the complex learning processes and decision-making strategies used by humans. Such neural network models make fewer assumptions about the underlying mechanisms thus providing experimental flexibility in terms of applicability. However, this comes at the cost of involving a larger number of parameters requiring significantly more data for effective learning. This presents practical challenges given that most cognitive experiments involve relatively small numbers of subjects. Laboratory collaborations are a natural way to increase overall dataset size. However, data sharing barriers between laboratories as necessitated by data protection regulations encourage the search for alternative methods to enable collaborative data science. Distributed learning, especially federated learning (FL), which supports the preservation of data privacy, is a promising method for addressing this issue. To verify the reliability and feasibility of applying FL to train neural networks models used in the characterization of decision making, we conducted experiments on a real-world, many-labs data pool including experiment data-sets from ten independent studies. The performance of single models trained on single laboratory data-sets was poor. This unsurprising finding supports the need for laboratory collaboration to train more reliable models. To that end we evaluated four collaborative approaches. The first approach represents conventional centralized learning (CL-based) and is the optimal approach but requires complete sharing of data which we wish to avoid. The results however establish a benchmark for the other three approaches, federated learning (FL-based), incremental learning (IL-based), and cyclic incremental learning (CIL-based). We evaluate these approaches in terms of prediction accuracy and capacity to characterize human decision-making strategies. The FL-based model achieves performance most comparable to that of the CL-based model. This indicates that FL has value in scaling data science methods to data collected in computational modeling contexts when data sharing is not convenient, practical or permissible.

Estimation of the respiratory rate from ballistocardiograms using the Hilbert transform

BackgroundMeasuring the respiratory rate is usually associated with discomfort for the patient due to contact sensors or a high time demand for healthcare personnel manually counting it.MethodsIn this paper, two methods for the continuous extraction of the respiratory rate from unobtrusive ballistocardiography signals are introduced. The Hilbert transform is used to generate an amplitude-invariant phase signal in-line with the respiratory rate. The respiratory rate can then be estimated, first, by using a simple peak detection, and second, by differentiation.ResultsBy analysis of a sleep laboratory data set consisting of nine records of healthy individuals lasting more than 63 h and including more than 59,000 breaths, a mean absolute error of as low as 0.7 BPM for both methods was achieved.ConclusionThe results encourage further assessment for hospitalised patients and for home-care applications especially with patients suffering from diseases of the respiratory system like COPD or sleep apnoea.

A new wave height distribution for intermediate and shallow water depths

The present paper addresses the short-term distribution of zero-crossing wave heights in intermediate and shallow water depths. New physical insights are provided regarding the effects of nonlinearity, directionality, reduced effective water depth and finite spectral bandwidth. These are demonstrated through the analysis of a large database of experimental simulations of short-crested sea-states on flat bed bathymetries. A new wave height model is proposed building upon these physical insights and is calibrated using the experimental data. Independent comparisons between field measurements and the proposed model indicate that it is appropriate to a wide range of incident wave conditions and that it provides considerable improvement over existing models. • Extensive laboratory and field datasets in intermediate and shallow water depths are analysed. • The effects of nonlinearity, directionality, spectral bandwidth and effective water depth on the wave height distribution are systematically investigated. • A new wave height model for intermediate and shallow water depths is developed. • The proposed model is shown to provide accurate predictions over a broad range of sea-states.

Powernet: A novel method for wind power predictive analytics using Powernet deep learning model

Sustainable energy is a significant power generation resource for a cleaner and CO2 free environment. Out of different renewable energies out there, wind energy is a rapidly growing sector and integrated into the power grid. However, uncertainty, stochastic, and non-stationary nature of meteorological features, on which wind power depends, makes it difficult to predict accurately. The efficiency of wind farms and the power grid is directly proportional to efficient wind power predictive analytics. This study describes a hybrid model named Powernet for improving the predicted accuracy in the field of wind power analytics. The improved hybrid model is a combination of Convolution 1 Dimensional (Conv-1D) and Bidirectional Long Short-Term Memory (BiLSTM) models. First, Conv-1D layers extract the spatial features of timestamped data sequentially. Then, the output generated by multiple convolution operations at the nested layers is embedded with BiLSTM to work on the temporal characteristics of wind power data. The nesting of spatial and temporal extractors generates a novel architecture, Powernet for wind power forecasting from raw data. The effectiveness of Powernet has been validated on the real-time wind power National Renewable Energy Laboratory dataset. Also, error and computational analysis have been conducted for short-term wind power forecasting with an ensemble of long short-term memory-based models. The comparative analysis demonstrates that the proposed model Powernet achieves better prediction than traditional deep learning standalone and hybrid models. Also, the statistical models are compared to show that the raw data need to be pre-processed when conventional models are applied. However, Powernet does not need the overhead of pre-processing for generating better predictions.

Personalized neurorehabilitative precision medicine: from data to therapies (MWKNeuroReha) – a multi-centre prospective observational clinical trial to predict long-term outcome of patients with acute motor stroke

BackgroundStroke is one of the most frequent diseases, and half of the stroke survivors are left with permanent impairment. Prediction of individual outcome is still difficult. Many but not all patients with stroke improve by approximately 1.7 times the initial impairment, that has been termed proportional recovery rule. The present study aims at identifying factors predicting motor outcome after stroke more accurately than before, and observe associations of rehabilitation treatment with outcome.MethodsThe study is designed as a multi-centre prospective clinical observational trial. An extensive primary data set of clinical, neuroimaging, electrophysiological, and laboratory data will be collected within 96 h of stroke onset from patients with relevant upper extremity deficit, as indexed by a Fugl-Meyer-Upper Extremity (FM-UE) score ≤ 50. At least 200 patients will be recruited. Clinical scores will include the FM-UE score (range 0–66, unimpaired function is indicated by a score of 66), Action Research Arm Test, modified Rankin Scale, Barthel Index and Stroke-Specific Quality of Life Scale. Follow-up clinical scores and applied types and amount of rehabilitation treatment will be documented in the rehabilitation hospitals. Final follow-up clinical scoring will be performed 90 days after the stroke event. The primary endpoint is the change in FM-UE defined as 90 days FM-UE minus initial FM-UE, divided by initial FM-UE impairment. Changes in the other clinical scores serve as secondary endpoints. Machine learning methods will be employed to analyze the data and predict primary and secondary endpoints based on the primary data set and the different rehabilitation treatments.DiscussionIf successful, outcome and relation to rehabilitation treatment in patients with acute motor stroke will be predictable more reliably than currently possible, leading to personalized neurorehabilitation. An important regulatory aspect of this trial is the first-time implementation of systematic patient data transfer between emergency and rehabilitation hospitals, which are divided institutions in Germany.Trial registrationThis study was registered at ClinicalTrials.gov (NCT04688970) on 30 December 2020.

High-performance steel-framed wall and floor/ceiling assemblies

Lightweight steel-framed assemblies can present significant advantages for construction: they are non-combustible, they can be faster to build, and they can be significantly lighter than concrete assemblies. However, steel-framed assemblies often present significant challenges when trying to obtain high levels of acoustical performance. In this paper the author presents sets of laboratory data for two assemblies. The first is a floor/ceiling assembly comprised of steel C-section joists with a spring-isolated gypsum board ceiling. This assembly achieved ratings of STC and IIC 60+ with hard-surface flooring. The second is a cross-braced double-stud wall with resilient sound isolation clips and damped gypsum board panels. This assembly achieved ratings up to STC 81. The test series also looked at the performance of the wall with and without cross-bracing, sound isolation clips, and damped gypsum board panels.

The Method of Rolling Bearing Fault Diagnosis Based on Multi-Domain Supervised Learning of Convolution Neural Network

The rolling bearing is a critical part of rotating machinery and its condition determines the performance of industrial equipment; it is necessary to detect rolling bearing faults as early as possible. The traditional methods of fault diagnosis are not efficient and are time-consuming. With the help of deep learning, the convolution neural network (CNN) plays a huge role in the data-driven methods of bearing fault diagnosis. However, the vibration signal is non-stationary, contains high noise, and is one-dimensional, which is difficult to analyze directly by the CNN model. Considering the multi-domain learning as an advantage of deep learning, this paper proposes a novel rolling bearing fault diagnosis approach using an improved one-dimensional (1D) and two-dimensional (2D) convolution neural network (CNN) of two-domain information learning. The constructed fault diagnosis model combining 1D and 2D CNN extracts the fault features from the two-domain information of bearing fault samples. The padding and dropout technology are utilized to fully extract features from the raw data and reduce over-fitting. To prove the validity of the proposed method, this paper performs two tests with two bearing datasets, the Case Western Reserve University (CWRU) bearing dataset and the Dalian University of Technology (DUT) vibration laboratory dataset. The experimental results show that our proposed method achieves high recognition accuracy of rolling bearing fault states via two-domain learning of monitoring data, and there is no manual experience necessary. Vibration data under strong noise were also used to test the method, and the results show the superiority and robustness of the proposed method.

Utilization of Cystatin C in the Outpatient Setting

Introduction: Serum creatinine is the traditional biomarker for estimating glomerular filtration rate (eGFR). Cystatin C is an alternative biomarker for which estimating equations exist. The use of cystatin C testing, and the interrelationships among the recently revised Chronic Kidney Disease Epidemiology (CKD-EPI) 2021 estimating equations, was evaluated in a national outpatient laboratory dataset. Methods: Cystatin C results reported on adults between November 2011 and June 2018 by Laboratory Corporation of America Holdings were examined, with classification of ordering providers and diagnostic codes. Updated eGFR results were calculated using the CKD-EPI 2021 equations for each sample with both cystatin C and creatinine values available. The Spearman correlation coefficients were calculated. Reclassification at clinically relevant cut-off values was examined. Results: There were 87,803 serum cystatin C levels among 55,360 patients; mean age 58 ± 17 years; 50% women. Cystatin C usage increased over time and was ordered for many indications. Among 73,367 samples with simultaneous creatinine and cystatin C, r = 0.84 between eGFR-creatinine and eGFR-cystatin. Correlations of eGFR-creatinine, eGFR-cystatin, and the averaged result of the two equations to the new combined equation were r = 0.94, r = 0.97, and r = 0.998, respectively (p < 0.001 for all). Use of combined/averaged equations tended to result in a higher eGFR and upclassification, compared to eGFR-creatinine. Conclusion/Discussion: Use of Cystatin C is increasing and has moved beyond the nephrology community and the original indications from the 2012 KDIGO guidelines. Community utilization of cystatin C measurement is likely to expand, and understanding of the relationships between estimating equations will help clinicians optimize their use in the outpatient setting.

An Endoscope Image Enhancement Algorithm Based on Image Decomposition

The visual quality of endoscopic images is a significant factor in early lesion inspection and surgical procedures. However, due to the interference of light sources, hardware, and other configurations, the endoscopic images collected clinically have uneven illumination, blurred details, and contrast. This paper proposed a new endoscopic image enhancement algorithm. The image decomposes into a detail layer and a base layer based on noise suppression. The blood vessel information is stretched by channel in the detail layer, and adaptive brightness correction is performed in the base layer. Finally, Fusion obtained a new endoscopic image. This paper compares the algorithm with six other algorithms in the laboratory dataset. The algorithm is in the leading position in all five objective evaluation metrics, further indicating that the algorithm is ahead of other algorithms in contrast, structural similarity, and peak signal-to-noise ratio. It can effectively highlight the blood vessel information in endoscopic images while avoiding the influence of noise and highlight points. The proposed algorithm can well solve the existing problems of endoscopic images.

Technique for recognizing faces using a hybrid of moments and a local binary pattern histogram

&lt;p&gt;&lt;span&gt;The face recognition process is widely studied, and the researchers made great achievements, but there are still many challenges facing the applications of face detection and recognition systems. This research contributes to overcoming some of those challenges and reducing the gap in the previous systems for identifying and recognizing faces of individuals in images. The research deals with increasing the precision of recognition using a hybrid method of moments and local binary patterns (LBP). The moment technique computed several critical parameters. Those parameters were used as descriptors and classifiers to recognize faces in images. The LBP technique has three phases: representation of a face, feature extraction, and classification. The face in the image was subdivided into variable-size blocks to compute their histograms and discover their features. Fidelity criteria were used to estimate and evaluate the findings. The proposed technique used the standard Olivetti Research Laboratory dataset in the proposed system training and recognition phases. The research experiments showed that adopting a hybrid technique (moments and LBP) recognized the faces in images and provide a suitable representation for identifying those faces. The proposed technique increases accuracy, robustness, and efficiency. The results show enhancement in recognition precision by 3% to reach 98.78%.&lt;/span&gt;&lt;/p&gt;

Scour depth prediction at bridge piers using metaheuristics-optimized stacking system

This paper proposes a novel artificial intelligence-based approach, called the metaheuristics-optimized stacking system (MOSS), to assist civil engineers in estimating scour depth at bridge piers. MOSS integrates the advantages of hybrid and stacking ensemble schemes by combining a metaheuristic optimization algorithm with efficient machine learning (ML) models. The metaheuristic algorithm simultaneously finds the optimal values of all hyperparameters of constituent ML models in the stacking ensemble to generate the most effective system. The developed MOSS was established by integrating a forensic-based investigation (FBI) algorithm, least squares support vector regression (LSSVR), and radial basis function neural network (RBFNN). The efficiency of the MOSS is verified comprehensively with reference to three case studies of both laboratory data and field data that cover various levels of complexity and types of pier foundations. The performance of MOSS is compared to that of other single ML models, conventional voting ensemble, hybrid models, empirical methods, and mathematical approach. The analytical results of cross-validation reveal that MOSS was the most reliable approach, achieving the best values of all performance evaluation metrics. MOSS exhibited mean absolute percentage errors of 7.127%, 29.195% and 13.131% in the prediction of scour depth using a laboratory dataset, a field dataset, and a complex pier foundations dataset, respectively, and these values are at least 36%, 19% and 41% lower than those obtained using other approaches. The automated predictive analytics revealed the robustness, efficiency, and stability of MOSS. The novelty and contributions of this study include (1) providing a general stacking framework and re-defining the establishment of an ensemble ML model, and (2) developing a highly promising tool that greatly outperforms currently available methods to help civil engineers estimate the scour depth around bridge piers.

ISW-LM: An intensive symptom weight learning mechanism for early COVID-19 diagnosis

The novel coronavirus disease 2019 (COVID-19) pandemic has severely impacted the world. The early diagnosis of COVID-19 and self-isolation can help curb the spread of the virus. Besides, a simple and accurate diagnostic method can help in making rapid decisions for the treatment and isolation of patients. The analysis of patient characteristics, case trajectory, comorbidities, symptoms, diagnosis, and outcomes will be performed in the model. In this paper, a symptom-based machine learning (ML) model with a new learning mechanism called Intensive Symptom Weight Learning Mechanism (ISW-LM) is proposed. The proposed model designs three new symptoms’ weight functions to identify the most relevant symptoms used to diagnose and classify COVID-19. To verify the efficiency of the proposed model, multiple laboratory and clinical datasets containing epidemiological symptoms and blood tests are used. Experiments indicate that the importance of COVID-19 infection symptoms varies between countries and regions. In most datasets, the most frequent and significant predictive symptoms for diagnosing COVID-19 are fever, sore throat, and cough. The experiment also compares the state-of-the-art methods with the proposed method, which shows that the proposed model has a high accuracy rate of up to 97.1711%. The positive results indicate that the proposed learning mechanism can help clinicians quickly diagnose and screen patients for COVID-19 at an early stage.

A new method for parameterization of wave dissipation by sea ice

We present a method for predicting wave dissipation by sea ice that is based on the dimensional analysis of data with a scaling defined by ice thickness. Applying the method to an extensive dataset from the measurements during the “Polynyas, Ice Production, and seasonal Evolution in the Ross Sea” (PIPERS) cruise in 2017, we derive a new model of wave dissipation which describes a nonlinear dependence on ice thickness, and reveals the interrelation between the dependences on ice thickness and on wave frequency. This nonlinear dependence on ice thickness can have important implications on predicting low-frequency waves. The root-mean-square error of the prediction is significantly reduced using the new model, compared with other existing parametric models that are also calibrated for the PIPERS dataset. The new model also explicitly describes a condition of similarity between large- and small-scale observations, which is shown to exist when various laboratory datasets collapse onto the prediction. Thus, the new model improves the estimate of wave dissipation by ice across multiple scales.