Intervariable Correlations Research Articles

Magnetic Resonance Imaging Features and Functional Score in Patients Requiring Total Knee Arthroplasty.

Introduction Knee osteoarthritis (KOA) is a progressive degenerative disease, with an increasing prevalence among the population. The degenerative changes in KOA affect the cartilage, menisci, synovial tissue, and subchondral bone. The treatment for patients in advanced stages of the disease is total knee arthroplasty (TKA). The purpose of this descriptive study is to identify the MRI features in the case of patients with KOA who did not obtain an improvement in symptoms and joint function after the non-surgical treatmentsand who applied for surgical treatment, i.e. TKA. Also, we aimed to identify the correlations between the MRI changes and the functional score of the patients, as well as the inter-variable correlations. Materials and methods The study was conducted in the Department of Orthopedics and Traumatology at the University Emergency Hospital of Bucharest between January 1, 2023, and January 31, 2024. It included 50 patients who required TKA. This study is a prospective, observational, and descriptive analysis focusing on patients scheduled for TKA. Results The patients in the study group who required TKA had a Knee Society Score (KSS) ranging from 35 to 70and a KSS function score between 24 and 60. Among them, 98% had tricompartmental lesions of the articular cartilage, and 100% presented with osteophytes, even when they were not identifiable radiologically. Additionally, 58% of the patients had changes in the infrapatellar fat pad, 66% presented with joint effusion without any traumatic history, and 86% of the patients had partial or complete lesions of the anterior cruciate ligament. Conclusion TheMRI pattern of the patient who requires TKA consists of the presence of articular cartilage lesions in more than two compartments with exposure of the subchondral bone and diameter greater than 1 cm, meniscus lesions (>grade 2), meniscus extrusion (>grade 1), the presence of bone marrow lesions (BMLs) in the medial or lateral compartment of the femur or tibia, hyperintense signal of the infrapatellar fat pad, anterior cruciate ligament (ACL) lesions (>grade 2), and the presence of osteophytes together with the presence of effusion in the suprapatellar bursa. BMLs and changes in the infrapatellar fat pad may lead to the opening of new research perspectives explaining the complex changes in KOA in relation to the inflammatory process and gene expression.

Graph Transformer Network Incorporating Sparse Representation for Multivariate Time Series Anomaly Detection

Cyber–physical systems (CPSs) serve as the pivotal core of Internet of Things (IoT) infrastructures, such as smart grids and intelligent transportation, deploying interconnected sensing devices to monitor operating status. With increasing decentralization, the surge in sensor devices expands the potential vulnerability to cyber attacks. It is imperative to conduct anomaly detection research on the multivariate time series data that these sensors produce to bolster the security of distributed CPSs. However, the high dimensionality, absence of anomaly labels in real-world datasets, and intricate non-linear relationships among sensors present considerable challenges in formulating effective anomaly detection algorithms. Recent deep-learning methods have achieved progress in the field of anomaly detection. Yet, many methods either rely on statistical models that struggle to capture non-linear relationships or use conventional deep learning models like CNN and LSTM, which do not explicitly learn inter-variable correlations. In this study, we propose a novel unsupervised anomaly detection method that integrates Sparse Autoencoder with Graph Transformer network (SGTrans). SGTrans leverages Sparse Autoencoder for the dimensionality reduction and reconstruction of high-dimensional time series, thus extracting meaningful hidden representations. Then, the multivariate time series are mapped into a graph structure. We introduce a multi-head attention mechanism from Transformer into graph structure learning, constructing a Graph Transformer network forecasting module. This module performs attentive information propagation between long-distance sensor nodes and explicitly models the complex temporal dependencies among them to enhance the prediction of future behaviors. Extensive experiments and evaluations on three publicly available real-world datasets demonstrate the effectiveness of our approach.

Machine learning prediction of one-year mortality after percutaneous coronary intervention in acute coronary syndrome patients

BackgroundMachine learning (ML) models have the potential to accurately predict outcomes and offer novel insights into inter-variable correlations. In this study, we aimed to design ML models for the prediction of 1-year mortality after percutaneous coronary intervention (PCI) in patients with acute coronary syndrome. MethodsThis study was performed on 13,682 patients at Tehran Heart Center from 2015 to 2021. Patients were split into 70:30 for testing and training. Four ML models were designed: a traditional Logistic Regression (LR) model, Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Ada Boost models. The importance of features was calculated using the RF feature selector and SHAP based on the XGBoost model. The Area Under the Receiver Operating Characteristic Curve (AUC-ROC) for the prediction on the testing dataset was the main measure of the model's performance. ResultsFrom a total of 9,073 patients with >1-year follow-up, 340 participants died. Higher age and higher rates of comorbidities were observed in these patients. Body mass index and lipid profile demonstrated a U-shaped correlation with the outcome. Among the models, RF had the best discrimination (AUC 0.866), while the highest sensitivity (80.9%) and specificity (88.3%) were for LR and XGBoost models, respectively. All models had AUCs of >0.8. ConclusionML models can predict 1-year mortality after PCI with high performance. A classic LR statistical approach showed comparable results with other ML models. The individual-level assessment of inter-variable correlations provided new insights into the non-linear contribution of risk factors to post-PCI mortality.

Generative Ensemble Deep Learning Severe Weather Prediction from a Deterministic Convection-Allowing Model

Abstract An ensemble postprocessing method is developed for the probabilistic prediction of severe weather (tornadoes, hail, and wind gusts) over the conterminous United States (CONUS). The method combines conditional generative adversarial networks (CGANs), a type of deep generative model, with a convolutional neural network (CNN) to postprocess convection-allowing model (CAM) forecasts. The CGANs are designed to create synthetic ensemble members from deterministic CAM forecasts, and their outputs are processed by the CNN to estimate the probability of severe weather. The method is tested using High-Resolution Rapid Refresh (HRRR) 1–24-h forecasts as inputs and Storm Prediction Center (SPC) severe weather reports as targets. The method produced skillful predictions with up to 20% Brier skill score (BSS) increases compared to other neural-network-based reference methods using a testing dataset of HRRR forecasts in 2021. For the evaluation of uncertainty quantification, the method is overconfident but produces meaningful ensemble spreads that can distinguish good and bad forecasts. The quality of CGAN outputs is also evaluated. Results show that the CGAN outputs behave similarly to a numerical ensemble; they preserved the intervariable correlations and the contribution of influential predictors as in the original HRRR forecasts. This work provides a novel approach to postprocess CAM output using neural networks that can be applied to severe weather prediction. Significance Statement We use a new machine learning (ML) technique to generate probabilistic forecasts of convective weather hazards, such as tornadoes and hailstorms, with the output from high-resolution numerical weather model forecasts. The new ML system generates an ensemble of synthetic forecast fields from a single forecast, which are then used to train ML models for convective hazard prediction. Using this ML-generated ensemble for training leads to improvements of 10%–20% in severe weather forecast skills compared to using other ML algorithms that use only output from the single forecast. This work is unique in that it explores the use of ML methods for producing synthetic forecasts of convective storm events and using these to train ML systems for high-impact convective weather prediction.

Multivariable correlation feature network construction and health condition assessment for unlabeled single-sample data

The construction of effective health indicators is crucial for assessing system degradation, enabling anomaly detection and health condition assessment, which contribute to reducing costs, improving productivity, and enhancing system availability. However, there are notable challenges in the health condition assessment for unlabeled single-sample data. In this paper, a multivariable correlation feature network—the Nested Autoencoder (NAE) network, is proposed, which incorporates inter-variable correlations to constrain the construction of latent space, thereby enhancing the accuracy of health condition assessment. Moreover, the autoencoder parameters calculated with each discrete interval data are leveraged to construct latent feature graphs, enabling component anomaly detection. Furthermore, a comprehensive indicator is introduced to describe the topological changes of the graphs, facilitating the assessment of the health conditions of the system. Finally, the effectiveness of our method is validated on the N-CMAPSS dataset as well as a real satellite dataset.

THE MUSCULOSKELETAL PROFILE OF FEMALE ADOLESCENT ELITE WATER POLO PLAYERS AND THE FACTORS AFFECTING THROWING PERFORMANCE

The musculoskeletal (MSK) profiles of water polo players and other overhead athletes has been shown to relate to injury and throwing performance (TP). There have been no robust studies conducted on the MSK profiles and the variables affecting TP amongst female, adolescent, elite water polo players.A prospective quantitative cohort design was conducted amongst eighty-three female adolescent, elite water polo players (range 14–19 years). All participants filled out the Kerlan-Jobe Orthopaedic Clinic questionnaire, followed by a battery of screening tests aimed to identify possible MSK factors affecting TP. Pain provocation tests, range of motion (ROM), upward scapula rotation (USR), strength and pectoralis minor length measurements were all included. Participants also performed throwing speed (TS) and throwing accuracy (TA) tests.All the data collected were grouped together and analysed using SPSS 28.0. The condition for statistical significance was set as p <0.05. Multi-collinearity was tested for among variables to find out inter-variable correlations. Finally, a multiple regression analysis was performed.The mean KJOC score was 82.55 ± 14.96. 26.5% tested positive for at least one of the impingement tests. The MSK profile revealed decreased internal rotation ROM, increased external rotation ROM, a downwardly rotated scapula, weak external rotators, weak serratus anterior strength, strong lower trapezius and gluteus medius strength and a shorter pectoralis minor length all on the dominant side.Age, pectoralis minor length, upper trapezius and serratus anterior strength as well as upward scapula rotation were all positively correlated with TS, while sitting height, upper trapezius and serratus anterior strength and glenohumeral internal rotation ROM were positively correlated with TA.Multiple MSK parameters were found to be related to TS and TA in elite, adolescent water polo players.

Novel Generative Recurrent Neural Network Framework to Produce Accurate, Applicable, and Deidentified Synthetic Medical Data for Patients With Metastatic Cancer

Sensitive patient data cannot be easily shared/analyzed, severely limiting the innovative progress of research, specifically for marginalized/under-represented populations. Existing methods of deidentification are subject to data breaches. The objective of this study was to develop a neural network capable of generating a synthetic version of data for patients with novel postoperative metastatic cancer. We analyzed a metastatic cancer patient cohort of 167,474 patients obtained from the National Surgical Quality Improvement Program. Twenty-seven clinical features were analyzed. We created a volume-matched synthetic cohort of 167,474 patients and a reduced-size synthetic cohort of 5,000 patients. The volume-matched and reduced-size synthetic cohorts were compared against the ground truth data to analyze differences in principal component distribution, underlying statistical properties/associations, intervariable correlations, and machine learning classifier performance when developed on the synthetic data. Among 167,474 patients with metastatic cancer in the original data, 50,669 (30.3%) died within 30 days of their index surgery. Our model was able to accurately capture underlying statistical properties, principal components, and intervariable correlations within the ground truth data, yielding an accuracy of 93.2% with a loss of 0.21%, and develop synthetic data capable of training accurate machine learning classifiers. The reduced-size synthetic data accurately replicated all categorical variables and every continuous variable with statistically similar records (P > .05), with the sole exception of preoperative albumin (P < .05). The volume-matched synthetic data frame was able to accurately replicate all categorical variables (P > .05). This described methodology can be applied to any structured medical data from any setting, significantly expedite scientific analysis/innovation, and be used to develop improved predictive classifiers with boosted tree-based algorithms, serving as the potential new gold standard of medical data sharing and data augmentation.

The impact of preoperative nutritional screening, ERAS protocol, and mini-invasive surgery in surgical oncology: A multi-institutional SEM analysis of patients with digestive cancer.

Mini-invasive surgery (MIS), ERAS, and preoperative nutritional screening are currently used to reduce complications and the length of hospital stay (LOS); however, inter-variable correlations have seldom been explored. This research aimed to define inter-variable correlations in a large series of patients with gastrointestinal cancer and their impact on outcomes. Patients with consecutive cancer who underwent radical gastrointestinal surgery between 2019 and 2020 were analyzed. Age, BMI, comorbidities, ERAS, nutritional screening, and MIS were evaluated to determine their impact on 30-day complications and LOS. Inter-variable correlations were measured, and a latent variable was computed to define the patients' performance status using nutritional screening and comorbidity. Analyses were conducted using structural equation modeling (SEM). Of the 1,968 eligible patients, 1,648 were analyzed. Univariable analyses documented the benefit of nutritional screening for LOS and MIS and ERAS (≥7 items) for LOS and complications; conversely, being male and comorbidities correlated with complications, while increased age and BMI correlated with worse outcomes. SEM analysis revealed that (a) the latent variable is explained by the use of nutritional screening (p0·004); (b) the variables were correlated (age-comorbidity, ERAS-MIS, and ERAS-nutritional screening, p < 0·001); and (c) their impact on the outcomes was based on direct effects (complications: sex, p0·001), indirect effects (LOS: MIS-ERAS-nutritional screening, p < 0·001; complications: MIS-ERAS, p0·001), and regression-based effects (LOS: ERAS, MIS, p < 0·001, nutritional screening, p0·021; complications: ERAS, MIS, p < 0·001, sex, p0·001). Finally, LOS and complications were correlated (p < 0·001). Enhanced recovery after surgery (ERAS), MIS, and nutritional screening are beneficial in surgical oncology; however, the inter-variable correlation is reliable, underlying the importance of the multidisciplinary approach.

Spatiotemporal characterization of wave-augmented varicose explosions

The onset of three-dimensional instabilities during “Wave-Augmented Varicose Explosions” (WAVE) liquid disintegration is characterized for an annular flow of a shear-thinning slurry that is fed into a central transonic steam flow. Droplet production inside the nozzle is enhanced by ligaments radially flicking up from the slurry wave into the steam flow with radial:axial velocity ratios exceeding 0.5. The wave also leaves residual ligaments in its wake, which facilitate further disintegration. After birth, a wave spends 80% of the wave cycle period building up to peak height at the nozzle exit. Two effervescent mechanisms are provided as 1) steam penetrates the rising wave and surface deformation allows steam fingers to force through, and 2) the wave collapses on itself, trapping steam. Baroclinic torque drives the development of Rayleigh-Taylor (RT) instabilities and reaches values on the order of 1×1013 1/s2. Both RT and Kelvin-Helmholtz instabilities are self-amplified in a viscosity-shear-temperature instability cycle because the slurry is non-Newtonian. Velocities inside the nozzle (wave formation region) are generally azimuthally similar (two-dimensional), but those outside the nozzle (radial bursting region) are azimuthally uncorrelated (three-dimensional). Inter-variable correlations show significant decoupling of quantities beyond the nozzle exit, and local strain rate fluctuations were found to correlate particularly well with bulk system pulsation. Although adaptive mesh refinement (AMR) can provide computationally efficient resolution of gas–liquid interfaces, this technique produced different results than an equivalent non-dynamic mesh when modeling WAVE. Gradients were particularly affected by AMR, and turbulent kinetic energy showed differences greater than 150% outside the nozzle.

Multiscale Wavelet-Driven Graph Convolutional Network for Blade Icing Detection of Wind Turbines

Blade icing detection is critical to maintaining the health of wind turbines, especially in cold climates. Rapid and accurate icing detection allows proper control of wind turbines, including shutting down and clearing the ice, thus ensuring turbine safety. This article presents a wavelet-driven multiscale graph convolutional network (MWGCN), which is a supervised deep learning model for blade icing detection. The proposed model first uses wavelet decomposition to capture multivariate information in the time and frequency domains, and then employs a temporal graph convolutional network (GCN) to model the intervariable correlations of the decomposed multiscale wavelets and their temporal dynamics. In addition, this article introduces scale attention to the MWGCN for a further improvement of the model and proposes the method to address the class imbalance problem of the training data sets. Finally, the article conducts comprehensive experiments to evaluate the proposed model, and the results demonstrate the effectiveness of the model in blade icing detection and its better performance over eight state-of-the-art algorithms, with 17.2% and 11.3% higher F1 scores over the best state-of-the-art baseline on the labeled datasets.

Changes in temperature–precipitation correlations over Europe: are climate models reliable?

Inter-variable correlations (e.g., between daily temperature and precipitation) are key statistical properties to characterise probabilities of simultaneous climate events and compound events. Their correct simulations from climate models, both in values and in changes over time, is then a prerequisite to investigate their future changes and associated impacts. Therefore, this study first evaluates the capabilities of one 11-single run multi-model ensemble (CMIP6) and one 40-member single model initial-condition large ensemble (CESM) over Europe to reproduce the characteristics of a reanalysis dataset (ERA5) in terms of temperature–precipitation correlations and their historical changes. Next, the ensembles’ correlations for the end of the 21st century are compared. Over the historical period, both CMIP6 and CESM ensembles have season-dependent and spatially structured biases. Moreover, the inter-variable correlations from both ensembles mostly appear stationary. Thus, although reanalysis displays significant correlation changes, none of the ensembles can reproduce them, with internal variability representing only 30% on the inter-model variability. However, future correlations show significant changes over large spatial patterns. Yet, those patterns are rather different for CMIP6 and CESM, reflecting a large uncertainty in changes. In addition, for historical and future projections, an analysis conditional on atmospheric circulation regimes is performed. The conditional correlations given the regimes are found to be the main contributor to the biases in correlation over the historical period, and to the past and future changes of correlation. These results highlight the importance of the large-scale circulation regimes and the need to understand their physical relationships with local-scale phenomena associated to specific inter-variable correlations.

The effect of courage on stress: The mediating mechanism of behavioral inhibition and behavioral activation in high-risk occupations.

Employees in high-risk occupations are exposed to tremendous work acute stress or prolonged stress disorders that are likely to undermine the health and organizational effectiveness. Based on positive psychology, courage which refers to behavioral approach despite the experience of fear could buffer the negative effects on stress. However, there is little known about the mechanisms by which courage decreases the risk of stress. Motivational systems may play an underlying role in this process, as behavioral inhibition system (BIS) is inhibited and behavioral activation system (BAS) is evoked by risk or threat. The current study aimed to examine the mediating effects of behavioral inhibition and activation on the relationship between courage and stress in the high-risk occupations. This study recruited 1,761 high-risk employees aged from 18 to 27 (M = 19.32; SD = 4.14) with a cluster sampling method who completed Courage Measure (CM), the BIS/BAS Scales and the Psychological Stress Evaluation Test (PSET). The correlation and mediation analyses examined the inter-variable correlations as well as the underlying mechanism between courage and stress. The results support the hypothesis and reveal that the behavioral inhibition mediates the association between courage and stress (B indirect = -0.02, p < 0. 01, 95%CI = -0.03 to -0.003). The behavioral activation of fun seeking mediates the association between courage and stress as well (B indirect = -0.04, p < 0. 01, 95%CI = -0.058 to -0.029). These findings suggest that behavioral inhibition and activation of fun seeking play imperative mechanism underpinning the buffering effect of courage on stress. Other theoretical and applied implications for desensitizing stress in the high-risk occupations are discussed.

Sex Differences in Depression and Anxiety Symptoms: Measurement Invariance, Prevalence, and Symptom Heterogeneity Among University Students in South Africa.

Adequate measurement is an essential component of the assessment of mental health disorders and symptoms such as depression and anxiety. The present study investigated sex-specific differences in the Patient Health Questionnaire-9 (PHQ-9) and Generalized Anxiety Disorder-7 (GAD-7). This comprehensive cross-sectional design study pursued four objectives: measurement invariance of PHQ-9 and GAD-7 between male and female; depression and anxiety prevalence differences; cross-sex differences in the relationship between depression and anxiety; and a comparison of symptom heterogeneity. A sample of 1966 (male = 592; female = 1374; mean age = 21 years) students from South Africa completed the PHQ-9 and the GAD-7. Data analyses for measurement invariance, latent class analysis, inter-variable correlations and group comparisons were conducted in Mplus. The two-dimensional PHQ-9 achieved scalar invariance, while the GAD-7 yielded metric invariance. The somatic and non-somatic latent dimensions of depression were compared and showed no significant difference between male and female groups. The positive relationship between depression and anxiety was also not significantly different between the two groups. While the PHQ-9 symptoms formed three classes in the male group, and four classes in the female group, the GAD-7 had the same number of classes (three) and a similar pattern between the two groups. These findings hold implications for the measurement, assessment and understanding of symptom manifestation and distribution, as well as the treatment of depression and anxiety in South Africa.

Hedonic Happiness of Foreign Tourists in Bali

There are many reasons why tourists travel or visit certain places, and chief among them is to seek happiness. For example, they feel happy when they get the hedonic experience. Unfortunately, the effect of hedonic values and tourism experiences on tourists' happiness has thus far remained unclear. This study was aimed to analyze the effect of hedonic value on the tourist experience, prove the effect of hedonic value on tourists' hedonic happiness, and examine the effect of tourist experience on tourists' hedonic happiness. This research took place in Bali, and the research subjects were foreign tourists who visited Bali. The total number of respondents was 171. All structured interviews and questionnaires were performed and distributed online. Data analysis was conducted using SEM (Structural Equation Model). The research findings show the positive and significant effect of hedonic values on tourist experience. There is a positive and significant effect of tourist experience on tourist hedonic happiness. It was found that there was no direct effect of hedonic value on tourist hedonic happiness. Out of the demonstrated intervariable correlations, we managed to create the Model of Hedonic Tourist Happiness (MHTH). This study concludes that the MHTH consists of hedonic value, tourist experiences, and tourist hedonic happiness variables. This MHTH can be applied to improve tourism services in Bali, and particularly to boost the tourist experience and tourist happiness. Keywords: Bali, Hedonic Happiness, Tourists Experiences.

A new multi-site multi-variable stochastic model with inter-site and inter-variable correlations, low frequency attributes and stochasticity: A case study in the lower Yellow River basin

• A stochastic model that can reproduce spatial and temporal correlation, as well as stochasticity is proposed. • A new indicator is proposed to measure the reproduction of natural stochasticity. • The new stochastic model is evaluated and compared to two other stochastic models.

Normative Values of Transcranial Magnetic Stimulation-Evoked Parameters for Healthy Developing Children and Adolescents

Context: Normative data-containing transcranial magnetic stimulation (TMS) evoked parameters are essential for correctly interpreting healthy development and assessing neuroplastic changes in certain neurologic disorders. Aims: The aim is to investigate corticospinal pathways by applying TMS to healthy developing children and adolescents. Settings and Design: In this cross-sectional study, we measured TMS evoked parameters associated with cortical and spinal stimulation obtained from the four extremities of 46 healthy children and adolescents (21 boys and 25 girls; mean ± standard deviation age: 6.4 ± 1.2 years; range: 3.0–20.5 years). Statistical Analysis: Spearman's correlation coefficients were calculated for each variable (weight and height) as a function of motor evoked potential (MEP) response latency and central motor conduction time (CMCT). Pearson's Chi-square test was used to determine the inter-variable correlations. Results: Latencies of MEPs were correlated with age (P &lt; 0.001, r = 0.6948) and height (P &lt; 0.006, r = 0.7994). Amplitudes of active-state MEPs were significantly higher than those of resting-state MEPs associated with the upper and lower extremities. The mean values for active-state MEP latencies were lower than those for resting-state MEPs. The CMCT and magnitudes of latency jumps were calculated using reliable MEP data for children and adolescents. Additionally, the unresponsiveness rates were significantly higher for children aged below 7 years. Conclusion: The TMS evoked parameters investigated in this study are necessary to accurately assess corticospinal pathway development in healthy children and adolescents.

Relative impact of red blood cell transfusion and anaemia on 5-year mortality in cardiac surgery.

The aim was to compare the relative effects of red blood cell (RBC) transfusion and preoperative anaemia on 5-year mortality following open-heart cardiac surgery using structural equation modelling. We hypothesized that patient risk factors associated with RBC transfusion are of larger importance than transfusion itself. This prospective cohort study, part of the Cardiac Surgery Outcome Study at St. Olavs University Hospital, Trondheim, Norway, included open-heart on-pump cardiac surgery patients operated on from 2000 through 2017 (n = 9315). Structural equation modelling, which allows for intervariable correlations, was used to analyse pathway diagrams between known risk factors and observed mortality between 30 days and 5 years postoperatively. Observation times between 30 days and 1 year, and 1-5 years postoperatively were also compared with the main analysis. In a simplified model, preoperative anaemia had a larger effect on 5-year mortality than RBC transfusion (standardized coefficients: 0.17 vs 0.09). The complete model including multiple risk factors showed that patient risk factors such as age (0.15), anaemia (0.10), pulmonary disease (0.11) and higher creatinine level (0.12) had larger effects than transfusion (0.03). Results from several sensitivity analyses supported the main findings. The models showed good fit. Preoperative anaemia had a larger impact on 5-year mortality than RBC transfusion. Differences in 5-year mortality were mainly associated with patient risk factors.

R&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt;D&amp;lt;sup&amp;gt;2&amp;lt;/sup&amp;gt; v2.0: accounting for temporal dependences in multivariate bias correction via analogue rank resampling

Abstract. Over the last few years, multivariate bias correction methods have been developed to adjust spatial and/or inter-variable dependence properties of climate simulations. Most of them do not correct – and sometimes even degrade – the associated temporal features. Here, we propose a multivariate method to adjust the spatial and/or inter-variable properties while also accounting for the temporal dependence, such as autocorrelations. Our method consists of an extension of a previously developed approach that relies on an analogue-based method applied to the ranks of the time series to be corrected rather than to their “raw” values. Several configurations are tested and compared on daily temperature and precipitation simulations over Europe from one Earth system model. Those differ by the conditioning information used to compute the analogues and can include multiple variables at each given time, a univariate variable lagged over several time steps or both – multiple variables lagged over time steps. Compared to the initial approach, results of the multivariate corrections show that, while the spatial and inter-variable correlations are still satisfactorily corrected even when increasing the dimension of the conditioning, the temporal autocorrelations are improved with some of the tested configurations of this extension. A major result is also that the choice of the information to condition the analogues is key since it partially drives the capability of the proposed method to reconstruct proper multivariate dependences.

A Statistical–Dynamical Methodology to Downscale Regional Climate Projections to Urban Scale

AbstractThe demand of city planners for quantitative information on the impact of climate change on the urban environment is increasing. However, such information is usually extracted from decadelong climate projections generated with global or regional climate models (RCMs). Because of their coarse resolution and unsuitable physical parameterization, however, their model output is not adequate to be used at city scale. A full dynamical downscaling to city level, on the other hand, is computationally too expensive for climatological time scales. A statistical–dynamical computationally inexpensive method is therefore proposed that approximates well the behavior of the full dynamical downscaling approach. The approach downscales RCM simulations using the combination of an RCM at high resolution (H-RES) and a land surface model (LSM). The method involves the setup of a database of urban signatures by running an H-RES RCM with and without urban parameterization for a relatively short period. Using an analog approach, these signatures are first selectively added to the long-term RCM data, which are then used as forcing for an LSM using an urban parameterization in a stand-alone mode. A comparison with a full dynamical downscaling approach is presented for the city of Brussels, Belgium, for 30 summers with the combined ALADIN–AROME model (ALARO-0) coupled to the Surface Externalisée model (SURFEX) as H-RES RCM and SURFEX as LSM. The average bias of the nocturnal urban heat island during heat waves is vanishingly small, and the RMSE is strongly reduced. Not only is the statistical–dynamical approach able to correct the heat-wave number and intensities, it can also improve intervariable correlations and multivariate and temporally correlated indices, such as Humidex.

Impacts of Using State‐of‐the‐Art Multivariate Bias Correction Methods on Hydrological Modeling Over North America

AbstractBias correction techniques are widely used to bridge the gap between climate model outputs and input requirements of hydrological models to assess the climate change impacts on hydrology. In addition to univariate bias correction methods, several multivariate bias correction methods were proposed recently, which can not only correct the biases in marginal distributions of individual climate variables but also properly adjust the biased intervariable correlations simulated by climate models. Due to the diversities of climate regime and climate model bias, hydrological simulation for watersheds under different climate conditions may show various sensitivities to the correction of intervariable correlations. Therefore, it is of great importance to investigate (1) whether the correction of intervariable correlations has impacts on the hydrological modeling and (2) how these impacts vary with watersheds under different climate conditions. To achieve these goals, this study evaluates behaviors and their spatial variability of multiple state‐of‐the‐art multivariate bias correction methods in hydrological modeling over 2,840 watersheds distributed in different climate regimes in North America. The results show that, compared to using a quantile mapping univariate bias correction method, applying multivariate methods can improve the simulation of snow proportion, snowmelt, evaporation, and several streamflow variables. In addition, this improvement is more clear for watersheds with arid and warm temperate climates in southern regions, while it is limited for northern snow‐characterized watersheds. Overall, this study demonstrates the importance of using multivariate bias correction methods instead of univariate methods in hydrological climate change impact studies, especially for watersheds with arid and warm temperate climates.