Principal Component Analysis Output Research Articles

A Hybrid Data Preprocessing-Based Hierarchical Attention BiLSTM Network for Remaining Useful Life Prediction of Spacecraft Lithium-Ion Batteries.

As a crucial energy storage for the spacecraft power system, lithium-ion batteries degradation mechanisms are complex and involved with external environmental perturbations. Hence, effective remaining useful life (RUL) prediction and model reliability assessment confronts considerable obstacles. This article develops a new RUL prediction method for spacecraft lithium-ion batteries, where a hybrid data preprocessing-based deep learning model is proposed. First, to improve the correlation between battery capacity and features, the empirically selected high-dimensional features are linearized by using the Box-Cox transformation and then denoised via the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) method. Second, the principal component analysis (PCA) algorithm is employed to perform feature dimensionality reduction, and the output of PCA is further processed by the sliding window technique. Third, a multiscale hierarchical attention bi-directional long short-term memory (MHA-BiLSTM) model is constructed to estimate the capacity in future cycles. Specifically, the MHA-BiLSTM model can predict the RUL of lithium-ion batteries by considering the correlation and significance of each cycle's information during the degradation process on different scales. Finally, the proposed method is validated based on multiple types of experiments under two lithium-ion battery datasets, demonstrating its superior performance in terms of feature extraction and multidimensional time series prediction.

The principal component of body height linear type traits and the relationship level to milk yields as Holstein cattle selection criterion

The essential body height characteristics associated with milk yields need to be better identified because research on that topic is rarely undertaken. On the other hand, body height measurements are widely known to be related to cow live weight. In particular, this study sought to identify the most relevant body height dimension trait as selection criteria for the milk yield increase program. The test animals for the study were 122 heads of Holstein cows, and seven characteristics of body height were recorded for each cow separately. Principal component analysis (PCA), correlation, and regression were used to analyze the data. As an analytical tool, the R program 4.2.1 with RStudio was employed. The discovered primary elements of PCA's output were the wither height (WTH), back height (BCH), rump height (RMH), thurl height (TLH), tail-head height (THH), and pins height (PNH). Afterward, the correlation and regression analysis findings showed that the rear udder height (RUH) had the highest priority in correlating with milk yields, followed by the thurl height (TLH). In conclusion, it is proposed that the RUH be utilized for the cow selection program while the TLH is used for the calf and heifer selection program. However, wither height (WTH) is the fittest metric for beef cattle and barn facility-related research.

The effect of silvicultural systems on soil function depends on bedrock geology and altitude

Soil quality and function in forest environments are influenced by the interaction of soil-forming parameters and silvicultural systems. Hyrcanian forests were recently accepted as a UNESCO World Heritage Site, which extends across an area of approximately 1.8 million hectares and ascend to an elevation of 2800 m above sea level (m.a.s.l). In these woodlands, Oriental Beech (Fagus orientalis Lipsky) is the predominant tree species and could be observed at 700–1500 m.a.s.l., and occur on different parent rocks. Shelterwood and single-tree selection techniques have been the primary management methods for beech forests for the past forty years. Studies investigating the impacts of silvicultural systems have not yet been done on soil and forest floor features on different bedrock geology and altitudes. Therefore, in this study, we examined the influence of single-tree selection and shelterwood methods, 25 years after employing those methods, on soil quality and function compared to control areas (intact forests) in Hyrcanian beech stands. For this purpose, 15 forest floor (30 × 30 cm) and topsoil (0–10 cm depth) samples in each silvicultural systems (i.e., single-tree selection and shelterwood methods and control zones) × 4 regions (including Rasht, Nowshahr, Sari and Gorgan) × 4 altitude levels (with averages of 800, 1000, 1200 and 1400 m.a.s.l.) were considered. According to our findings, the investigated forest regions, forest floor and soil characteristics across various locations spots could be separated by principal component analysis output, and more than 85% of the variance was explained by the first and second axes. The structural equation model showed that the region, altitude and silvicultural systems had an effective role in the changes in soil biological activities by influencing the forest floor, and the soil physicochemical features. Based upon the network model, the C/N ratio, the sand content, the soil aggregate stability, the available K, the fulvic acid, and the Acarina density were found to be prominent factors with regard to soil function. In the control sites, increased soil organic material fractions, microbial/enzyme and biota activities were detected, particularly at the lower altitudes of the Nowshahr site, which had geological formations of dolomite and calcic layers. Taken together, it seems that the single-tree method, commonly referred to as the close-to-nature technique produces more suitable conditions for soil functioning compared to the shelterwood management approach. Silvicultural systems, bedrock geology and altitude can have major detrimental effects on soil function and fertility, over the long-term, impacts may increase with harvest intensity.

Privacy preserving identification of population stratification for collaborative genomic research.

The rapid improvements in genomic sequencing technology have led to the proliferation of locally collected genomic datasets. Given the sensitivity of genomic data, it is crucial to conduct collaborative studies while preserving the privacy of the individuals. However, before starting any collaborative research effort, the quality of the data needs to be assessed. One of the essential steps of the quality control process is population stratification: identifying the presence of genetic difference in individuals due to subpopulations. One of the common methods used to group genomes of individuals based on ancestry is principal component analysis (PCA). In this article, we propose a privacy-preserving framework which utilizes PCA to assign individuals to populations across multiple collaborators as part of the population stratification step. In our proposed client-server-based scheme, we initially let the server train a global PCA model on a publicly available genomic dataset which contains individuals from multiple populations. The global PCA model is later used to reduce the dimensionality of the local data by each collaborator (client). After adding noise to achieve local differential privacy (LDP), the collaborators send metadata (in the form of their local PCA outputs) about their research datasets to the server, which then aligns the local PCA results to identify the genetic differences among collaborators' datasets. Our results on real genomic data show that the proposed framework can perform population stratification analysis with high accuracy while preserving the privacy of the research participants.

Mapping of mafic-ultramafic rocks in SMUC-SGT, India using ASTER & Sentinel-2A satellite images

The Salem Mafic–Ultramafic Complex (SMUC) is a major complex in southern Granulite Terrane (SGT), in the northern part of the Cauvery Suture Zone (CSZ), India, also known as the Chalk Hills of Salem. The present study aims to review the efficiency of satellite image processing and related techniques to map the Mafic-Ultramafic rocks in Salem Mafic-Ultramafic Complex (SMUC). This mapping is done using atmospherically corrected Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and Sentinel-2A satellite image data. The processes employed were image enhancement techniques like Band Combinations (BC), Principal Component Analysis (PCA), and Minimum Noise Fraction (MNF) to map the various rock types. Band combinations constructed from the output of PCA and MNF in RGB format showed promising results in discriminating various rock units. The Spectral Angle Mapper (SAM) and Support Vector Machine (SVM) classifications techniques were used to classify ASTER and Sentinel-2A satellite images for the study area. The classified lithologies showed strong match with the lithology map published by the Geological Survey of India (GSI), other researches, and field investigation. Particularly, SVM classification of Sentinel-2A MNF + DEM showed 90% overall accuracy when compared with field investigation. The study proved that the processing of VNIR and SWIR bands of ASTER and Sentinel-2A Satellite images coupled with DEM and ground investigation has excellent potential in mapping Mafic-Ultramafic rocks in SMUC SGT.

Evaluation of in vivo staging of amyloid deposition in cognitively unimpaired elderly aged 78–94

Amyloid-beta (Aβ) deposition is common in cognitively unimpaired (CU) elderly >85 years. This study investigated amyloid distribution and evaluated three published in vivo amyloid-PET staging schemes from a cognitively unimpaired (CU) cohort aged 84.9 ± 4.3 years (n = 75). SUV-based principal component analysis (PCA) was applied to 18F-flutemetamol PET data to determine an unbiased regional covariance pattern of tracer uptake across grey matter regions. PET staging schemes were applied to the data and compared to the PCA output. Concentration of p-tau181 was measured in blood plasma. The PCA revealed three distinct components accounting for 91.2% of total SUV variance. PC1 driven by the large common variance of uptake in neocortical and striatal regions was significantly positively correlated with global SUVRs, APOE4 status and p-tau181 concentration. PC2 represented mainly non-specific uptake in typical amyloid-PET reference regions, and PC3 the occipital lobe. Application of the staging schemes demonstrated that the majority of the CU cohort (up to 93%) were classified as having pathological amount and distribution of Aβ. Good correspondence existed between binary (+/−) classification and later amyloid stages, however, substantial differences existed between schemes for low stages with 8–17% of individuals being unstageable, i.e., not following the sequential progression of Aβ deposition. In spite of the difference in staging outcomes there was broad spatial overlap between earlier stages and PC1, most prominently in default mode network regions. This study critically evaluated the utility of in vivo amyloid staging from a single PET scan in CU elderly and found that early amyloid stages could not be consistently classified. The majority of the cohort had pathological Aβ, thus, it remains an open topic what constitutes abnormal brain Aβ in the oldest-old and what is the best method to determine that.

Assessing Image Classification Accuracy with Principal Component Analysis Algorithm Case Study: Odeda LGA of Ogun State, Southwest Nigeria

The aim of this study is to assess image classification accuracy using the instrumentality of Principal Component Analysis (PCA). It is focused on evaluating the accruable benefits of Principal Component Analysis as part of an image preprocessing procedure for image classification. Land use land cover (LULC) and accuracy assessment datasets were obtained with remote sensing and geographic information system’s software. The principal component analysis was statistically used to assess the level of correlation amongst bands in Landsat 8. The image classification was premised on the Maximum Likelihood classifier for land use land cover analysis. To ascertain the accuracy of the classified images, the Producer’s accuracy, User’s accuracy and Kappa coefficient derivatives of accuracy assessment was calculated. The results revealed that the first three PCs of the raw Landsat data accounted for 99.37 % variance of the original Landsat data, while the last three PCs represented only 0.63% of the original data. The results of land use land cover based on raw bands composite were Forest (41%), Shrubs (33%) and Built-up (26%) respectively. On the other hand, land use land cover based on Principal Component Analysis showed Forest (39%), Shrubs (39%) and Built-up (22%) respectively. Comparing the results of Kappa coefficients of both LULC of raw bands’ composite was 0.88 while that of PCA was 0.91. Conclusively, there is a significant level of difference in the classification outputs of PCA derived classification and that of raw Landsat bands’ composite.

Age-related changes in three-dimensional foot motion during barefoot walking in children aged between 7 and 11 years old

BackgroundThe biomechanical complexity of children’s feet changes throughout childhood, yet kinematic development of the feet is poorly understood. Further work exploring the kinematic profile of children’s feet would be beneficial to help inform our understanding of the typical development of children’s feet. Research QuestionDo three-dimensional segmental kinematics of the feet during gait relate to age in a sample of children age 7–11 years? MethodsThis study was a secondary analysis of an existing database representing one hundred and twenty-one children age 7 – 11 years (90 male, 31 female; mean ± SD: age 9.57 ± , 1.17 years, height 1.37 ± 0.08 m, body mass 35.61 ± 9.33 kg). Fifteen, 9 mm retroreflective markers were attached to the right shank and foot of each participant in, line with the 3DFoot model. Multi-segmental joint kinematics were collected during barefoot walking. Sagittal, frontal, and transverse planar motion was described for the shank-calcaneus, calcaneus-midfoot, and midfoot-metatarsals segment of the right foot. Principal component analysis (PCA) was used to reduce the major modes of variation in the data to fully explore foot segment motion over the entire gait cycle. Correlations and multiple regression between PCA outputs with age, and potential confounding factors are presented. ResultsSignificant positive correlations were found between age and greater calcaneus, dorsiflexion, midfoot inversion and adduction, and metatarsal dorsiflexion, plantarflexion and abduction. There were no significant confounding effects of height, body mass, walking speed or gender on the relationships between age and PCA outputs. SignificanceThe findings from this study demonstrated a relationship between foot kinematics and age suggesting that the development of foot kinematics is ongoing until at least the age of 11 years. This work offers a comprehensive data set of inter-segmental kinematics which helps to advance understanding of the development of the pediatric foot.

Comparing Four Indexing Approaches to Define Soil Quality in an Intensively Cropped Region of Northern India

The usefulness of the soil quality index (SQI) as a tool to evaluate management options has mostly been studied within the boundaries of a crop or experimental field, calling for the need to enhance its utility in regional-scale soil health assessment. Thus, four quantitative approaches for computing the SQI were evaluated with samples collected from 0 to 15 and 15 to 30 cm depths at 156 points from the Trans-Gangetic Plains of North India. Principal component analysis (PCA) and soil function (SF)-based approaches were used to select the minimum dataset from 18 soil parameters and assign weights to key indicators. In both approaches, two different data transformation methods were followed: 1) routine method with maximum or minimum values of indicator parameters and 2) percentile method with the 90th or 10th percentile value as the denominator or numerator for “more is better” and “less is better” scoring functions, respectively. The PCA output with factor loadings from the varimax rotation showed six principal components accounting for 75% of the total variance, with PC1 explaining the highest variance (26.8%) followed by PC2 (16%). The SF-based approach was better than PCA in terms of a higher correlation of SQI with rice and wheat yields. The percentile method showed a higher correlation in both PCA and SF methods. The SQI computed from 0 to 30 cm soil data did not show any superiority over that from 0 to 15 cm soil. Thus, the soil function–based approach with the percentile method of data transformation proved better to compute the SQI and establish a relationship with production function.

Spatial variability of forward modelled attenuated backscatter in clear-sky conditions over a megacity: Implications for observation network design.

Sensors that measure the attenuated backscatter coefficient (e.g., automatic lidars and ceilometers [ALCs]) provide information on aerosols that can impact urban climate and human health. To design an observational network of ALC sensors for supporting data assimilation and to improve prediction of urban weather and air quality, a methodology is needed. In this study, spatio‐temporal patterns of aerosol‐attenuated backscatter coefficient are modelled using Met Office numerical weather prediction (NWP) models at two resolutions, 1.5 km (UKV) and 300 m (London Model [LM]), for 28 clear‐sky days and nights. Initially, attenuated backscatter coefficient data are analysed using S‐mode principal component analysis (PCA) with varimax rotation. Four to seven empirical orthogonal functions (EOFs) are produced for each model level, with common EOFs found across different heights (day and night) for both NWP models. EOFs relate strongly to orography, wind, and emissions source location, highlighting these as critical controls of attenuated backscatter coefficient spatial variability across the megacity. Urban–rural differences are largest when wind speeds are low and vertical boundary‐layer dynamics can more effectively distribute near‐surface aerosol emissions vertically. In several night‐time EOFs, gravity‐wave features are found for both NWP models. Increasing the horizontal resolution of native ancillaries (model input parameters) and improving the urban surface scheme in the LM may enhance the urban signal in the EOFs. PCA output, with agglomerative Ward cluster analysis (CA), minimises intra‐group variance. The UKV and LM CA shape and size results are similar and strongly related to orography. PCA‐CA is a simple, but adaptable methodology, allowing close alignment with observation network design goals. Here, CA is used with wind roses to suggest the optimised ALC deployment is one in the city to observe the urban plume and others surrounding the city, with priority given to cluster size and frequency of upwind advection.

Meta-Learner Hybrid Models to Classify Hyperspectral Images

Hyperspectral (HS) images are adjacent band images that are generally used in remote-sensing applications. They have numerous spatial and spectral information bands that are extremely useful for material detection in various fields. However, their high dimensionality is a big challenge that affects their overall performance. A new data normalization method was developed to enhance the variations and data distribution using the output of principal component analysis (PCA) and quantile transformation, called QPCA. This paper also proposes a novel HS images classification framework using the meta-learner technique to train multi-class and multi-size datasets by concatenating and training the hybrid and multi-size kernel of convolutional neural networks (CNN). The high-level model works to combine the output of the lower-level models and train them with the new input data, called meta-learner hybrid models (MLHM). The proposed MLHM framework with our external normalization (QPCA) improves the accuracy and outperforms other approaches using three well-known benchmark datasets. Moreover, the evaluation outcomes showed that the QPCA enhanced the framework accuracy by 13% for most models and datasets and others by more than 25%, and MLHM provided the best performance.

The influences of covid-19 pandemic on macroeconomic indexes for European countries

ABSTRACT The Covid-19 pandemic had a major impact in the entire world, both sanitary and economic. The purpose of this study is to analyse the European countries from this point of view. The main economic indicators related to import, wages, labour productivity, consumption, taxes, gross domestic product (GDP), as well as variables related to Covid-19 pandemic measures were considered in order to estimate the impact that pandemic measures have in economy. Principal component analysis was performed in order to reduce the dataset dimensionality and to create new indicators. Selected European countries are characterized and compared using principal components analysis outputs while regression is used to identify an influence of independent variables on GDP. The main results show that, comparing European countries economic situation in 2020 with the previous 2 years and considering variables that show the adopted closure measures, the Covid-19 pandemic have influenced the GDP from European countries.

Paper Insulation Aging Estimation Using Swept-Source Optical Coherence Tomography

In this article, an algorithm that uses subsurface structural images, obtained by swept-source optical coherence tomography (OCT), is developed to estimate the aging of insulation paper. Swept-source OCT, a novel and noninvasive diagnostic technique, is used to study the subsurface morphological changes in thermally stressed paper insulation. Obtaining subsurface structural images of the insulation paper enables an accurate estimation of the age of thermally deteriorated insulation paper. Accelerated thermal aging is employed to generate samples of oil-impregnated insulation paper with different levels of deterioration. Textural features, based on a spatial grayscale-level dependence matrix (SGLDM) obtained from the OCT images, are used to estimate the aging time of insulation paper based on changes in the subsurface morphology of the paper. Principal component analysis (PCA) is applied to the feature sets to reduce its dimension. The PCA output is then used to construct the age estimation model, and the leave-one-out-cross-validation (LOOCV) method is used to assess the estimation performance of the fitted model. The results of this work demonstrate that the development of OCT-equipped scanning probes to assess the condition of paper insulation is promising.

Principal component analysis and K-means clustering as tools during exploration for Zn skarn deposits and industrial carbonates, Sala area, Sweden

This contribution presents an application of principal component analysis (PCA) and K-means clustering as tools for data dimension reduction and grouping of multivariate, whole-rock lithogeochemical data. The study dataset consists of 64 geochemical variables and measurements of spectrophotometric brightness determined from 181 dolomite marble samples, collected at various distance from two contrasting types of mineral deposits, 1) stratabound, dolomite marble- and skarn-hosted Zn-Pb-Ag sulphide deposits and 2) industrial dolomite deposits. Clustering and PCA outputs are assessed based on spatial distribution relative to known mineral deposits and interpretability using geological domain knowledge, to test if the methods can provide a non-biased classification of dolomite samples which is useful for exploration vectoring. The PCA illustrate that three principle components derived from centered log-ratio transformed data can account for 79.69% of the dataset variance. K-means clustering provide unsupervised division of samples into different groups reflecting relative contents of detrital (siliciclastic-volcaniclastic), biogenic and hydrothermal components in the marble protoliths. Spatial analysis of principal components and K-means clusters reveal systematic distribution patterns relative to known deposits, thus providing an exploration guide. The samples most prospective for Zn-Pb-Ag deposits are divided into groups of ‘halo dolomite’ exhibiting elevated Fe and Mn, and an ‘ore dolomite’ also showing elevated Zn, Pb, Ag, Sb, Hg. This can be reconciled with magnetite and Mn-bearing Mg-silicates and carbonates in hydrothermal alteration haloes, and proximal enrichment in hydrothermal sulphides (galena, pyrrhotite, pyrite, sphalerite). Samples in these groups returned low spectrophotometric brightness, resulting from sulphides and Fe oxides grinding to dark powders during sample preparation, significantly lowering the brightness of powdered dolomite marble, even when occurring in low concentrations. Conversely, a ‘clean dolomite’ group is characterized by low contents of the elements above, high contents of Ca, Mg, Sr and total carbon, low magnetic susceptibility and high spectrophotometric brightness, and spatially coincide with known industrial dolomite deposits. An additional group of ‘detrital-rich dolomite’ is distinct from the other groups in an elevated content of high field strength elements and Al, and intermediate spectrophotometric brightness. This variety represent samples containing a higher content of co-settled volcaniclastic-siliciclastic material in the marble precursor. Assessment of the clustered data in relation to magnetic susceptibility measurements from the same samples show that Halo and Ore dolomite can be differentiated from other dolomite types by geomagnetic methods, hence providing a proxy for their indirect detection during geophysical surveys.

Fast Grain Mapping with Sub-Nanometer Resolution Using 4D-STEM with Grain Classification by Principal Component Analysis and Non-Negative Matrix Factorization.

High-throughput grain mapping with sub-nanometer spatial resolution is demonstrated using scanning nanobeam electron diffraction (also known as 4D scanning transmission electron microscopy, or 4D-STEM) combined with high-speed direct-electron detection. An electron probe size down to 0.5 nm in diameter is used and the sample investigated is a gold–palladium nanoparticle catalyst. Computational analysis of the 4D-STEM data sets is performed using a disk registration algorithm to identify the diffraction peaks followed by feature learning to map the individual grains. Two unsupervised feature learning techniques are compared: principal component analysis (PCA) and non-negative matrix factorization (NNMF). The characteristics of the PCA versus NNMF output are compared and the potential of the 4D-STEM approach for statistical analysis of grain orientations at high spatial resolution is discussed.

Noise reduction of shot-noise-dominated hyperspectral imagery by combining PCA with existing denoising methods

In this paper, we revisit the effects of principal component analysis (PCA) on hyperspectral imagery denoising. Our previous work combined PCA with wavelet shrinkage and particularly good denoising results has been achieved. We debate that any denoising methods can be used to replace wavelet shrinkage in our PCA+wavelet shrinkage algorithm. The major difference between this work and our previous PCA-based denoising method is that we consider a mixture of Gaussian and shot noise in this work whereas our previous methods studied Gaussian white noise alone. In addition, we retain [Formula: see text] [Formula: see text] PCA output components in our forward PCA transform in this paper whereas we keep all PCA output components [Formula: see text] in our previous works. The [Formula: see text] above is the number of spectral bands in the original hyperspectral imagery data cube. In addition, PCA is much better than nonlinear PCA for hyperspectral imagery denoising when Gaussian white noise and shot noise are introduced as demonstrated in this paper. Extensive experiments demonstrate that the method proposed in this paper outperforms the existing methods significantly in terms of signal-to-noise ratio for two testing hyperspectral imagery data cubes.

Water quality prediction and classification based on principal component regression and gradient boosting classifier approach

Estimating water quality has been one of the significant challenges faced by the world in recent decades. This paper presents a water quality prediction model utilizing the principal component regression technique. Firstly, the water quality index (WQI) is calculated using the weighted arithmetic index method. Secondly, the principal component analysis (PCA) is applied to the dataset, and the most dominant WQI parameters have been extracted. Thirdly, to predict the WQI, different regression algorithms are used to the PCA output. Finally, the Gradient Boosting Classifier is utilized to classify the water quality status. The proposed system is experimentally evaluated on a Gulshan Lake-related dataset. The results demonstrate 95% prediction accuracy for the principal component regression method and 100% classification accuracy for the Gradient Boosting Classifier method, which show credible performance compared with the state-of-art models.

Hyperspectral image classification via principal component analysis, 2D spatial convolution, and support vector machines

Hyperspectral image (HSI) classification has many applications in different diverse research fields. We propose a method for HSI classification using principal component analysis (PCA), 2D spatial convolution, and support vector machine (SVM). Our method takes advantage of correlation in both spatial and spectral domains in an HSI data cube at the same time. We use PCA to reduce the dimensionality of an HSI data cube. We then perform spatial convolution to the dimension-reduced data cube once and then to the convolved data cube for the second time. As a result, we have generated two convolved PCA output data cubes in a multiresolution way. We feed the two convolved data cubes to SVM to classify each pixel to one of the known classes. Experiments on three widely used hyperspectral data cubes (i.e., Indian Pines, Pavia University, and Salinas) demonstrate that our method can improve the classification accuracy significantly when compared to a few existing methods. Our method is relatively fast in terms of central processing unit computational time as well.

Employing principal component analysis for assessment of damage in GFRP composites using electromechanical impedance

In the presented research authors investigated the electromechanical impedance (EMI) technique in the context of damage detection. The EMI method has been studied for considerable amount of time for detecting damage in metallic, composite, concrete, and other parts. Here, the focus is put on the GFRP (glass fibre reinforced polymer) composites. Both a beam and plate was investigated. The samples were prepared with simulated damage. Impact damage was introduced as well as delamination. Authors have developed a data processing tool based on principal component analysis (PCA). The PCA output served as a damage index. It allowed for sensor network data handling, differentiating the damage cases, and damage detection with missing some of the data. The obtained results are promising and show advantages over traditional EMI damage index – the RMSD. The proposed solution can be employed when only one sensor is used as well as when a sensor network is used. Both cases were reported in this work.

How to Recognize a True Mode of Atmospheric Circulation Variability

AbstractIt has been demonstrated several times that when principal component analysis (PCA) is used for detection of modes of atmospheric circulation variability (teleconnections), principal components must be rotated. Despite it, unrotated PCA is still often used. Here we demonstrate on the examples of North Atlantic Oscillation (NAO), Arctic Oscillation (AO), Barents Oscillation (BO), and the summer East Atlantic (SEA) pattern that unrotated PCA results in patterns that are artifacts of the analysis method rather than true modes of variability. This claim is based on the comparison of the spatial patterns of the modes with spatial autocorrelations, on the sensitivity of the patterns to spatial and temporal subsampling, and, for the SEA pattern, on correlations with tropical sea surface temperature. Unlike NAO, which is defined by rotated PCA, the other modes, that is, AO, BO, and SEA pattern, defined by unrotated PCA, do not correspond well to underlying autocorrelation structures and are more sensitive to choices of spatial domain and time interval over which they are defined. We reiterate that a great care must be taken when interpreting outputs of PCA when applied to the detection of modes of circulation variability: a comparison with spatial autocorrelations and check for their spatial and temporal stability are necessary to distinguish true modes from statistical artifacts, which we call “ghost patterns.”