Empirical Orthogonal Function Coefficients Research Articles

Modeling TEC Irregularities in the Arctic Ionosphere Using Empirical Orthogonal Function Method

AbstractWe develop a climatological model for the Arctic ionosphere based on more than one solar cycle of data (2010–2021) of the rate of change of the total electron content index (ROTI) maps from the International GNSS Service (IGS). The IGS ROTI maps are daily averaged in magnetic latitude and local time coordinates. To develop a climatological model, the ROTI maps are decomposed into base functions and coefficients using the empirical orthogonal function (EOF) method. The EOF method converges very quickly, and the first four EOFs reflect the majority (96%) of the total data variability. Furthermore, different EOF components can reflect different drivers of ionospheric irregularities. The first EOF reflects the averaged ROTI activity and the impact of the solar radiation and geomagnetic activity; the second EOF reflects the impact of the interplanetary magnetic field (IMF) Bz and electric field; the third and fourth EOFs reflect the dawn‐dusk asymmetry around the auroral oval and polar cap, and they can be related to the IMF By. To build an empirical model, we fit the EOF coefficients using helio‐geophysical indices from four different categories (solar activity; geomagnetic indices; IMF; the solar wind coupling function). The final EOF model is dependent on seven selected indices (F10.7P, Kp, Dst, Bt, By, Bz, and EKL). The statistical data‐model comparisons show satisfactory results with a good correlation coefficient. However, the model cannot capture the significant expansion of the dayside ROTI activity during strong geomagnetic storms. Future effort is needed to provide corrections to the model during severe storms.

Climatology and modeling of ionospheric irregularities over Greenland based on empirical orthogonal function method

This paper addresses the long-term climatology (over two solar cycles) of total electron content (TEC) irregularities from a polar cap station (Thule) using the rate of change of the TEC index (ROTI). The climatology reveals variabilities over different time scales, i.e., solar cycle, seasonal, and diurnal variations. These variations in different time scales can be explained by different drivers/contributors. The solar activity (represented by the solar radiation index F10.7P) dominates the longest time scale variations. The seasonal variations are controlled by the interplay of the energy input into the polar cap ionosphere and the solar illumination that damps the amplitude of ionospheric irregularities. The diurnal variations (with respect to local time) are controlled by the relative location of the station with respect to the auroral oval. We further decompose the climatology of ionospheric irregularities using the empirical orthogonal function (EOF) method. The first four EOFs could reflect the majority (99.49%) of the total data variability. A climatological model of ionospheric irregularities is developed by fitting the EOF coefficients using three geophysical proxies (namely, F10.7P, Bt, and Dst). The data-model comparison shows satisfactory results with a high Pearson correlation coefficient and adequate errors. Additionally, we modeled the historical ROTI during the modern grand maximum dating back to 1965 and made the prediction during solar cycle 25. In such a way, we can directly compare the climatic variations of the ROTI activity across six solar cycles.

New England shelf break acoustic experiment (NESBA): Assimilation of acoustic pressure observations

The NESBA Signals and Noise experiment was conducted in April–May 2021. The goal of the experiment is to assess the potential for sonar prediction effectiveness gains given improved environmental awareness. This talk addresses the NESBA sub-goal of demonstrating that acoustic pressure observations could be assimilated into high-resolution ocean models (i.e., Regional NCOM) resulting in lower analysis and forecast errors for both the ocean environment variables and acoustic predictions. The steps involved in this experimental analysis include: (1) measurement of broadband signals received by drifting vertical arrays; (2) generation of empirical orthogonal function (EOF) representations of the local sound speed environment based on NOAA Regional NCOM forecasts; (3) use of acoustic simulation to generate simulated signals for general elements of the EOF state space; and (4) the use of optimization techniques to determine the EOF coefficients that best match each set of acoustic measurements. Multiple optimizations are to be conducted, each resulting in a synthetic CTD measurement at a specific location and time. These synthetic CTDs will be assimilated into Regional NCOM and comparisons will be made between baseline NCOM forecasts, NCOM forecasts including the synthetic CTD data, and CTD measurement data. [Work supported by the Office of Naval Research.]

Modeling the global ionospheric electron densities based on the EOF decomposition of the ionospheric radio occultation observation

An attempt is made to model the three-dimensional global structures and temporal variations of the ionospheric electron density (Ne) using the radio occultation (RO) Ne profile data obtained by the COSMIC/FORMOSAT-3, CHAMP and GRACE missions during the period from July 2006 to June 2017. The modeling technique we adopted is based on the empirical orthogonal function (EOF) analysis of the RO Ne dataset which is binned with grids of 2.5° in geomagnetic latitudes and 1/3 hr in geomagnetic local time (equivalent to 5° in geomagnetic longitudes) and 10 km in height. The EOF analysis decomposed the binned (gridded) Ne dataset into a series of eigen modes or basis functions (E i ) representing the variations with geomagnetic latitude, geomagnetic local time as well as height and the associated EOF amplitude coefficients (A i ) representing the variations with seasons as well as solar cycle activity. Our results showed that the EOF components (E i , A i ) obtained by the EOF decomposition have different three-dimensional spatial structures with distinct features attributable to different factors or processes controlling the variations of the ionosphere: the first EOF component represents mainly the global mean structure of Ne and its temporal variation with seasons and solar cycle activity; the second EOF component represents mainly the seasonal control of the solar zenith angle on the ionosphere; the third EOF component is representative of the uplifting or lowering of the peak height where Ne reaches its maximum value; the fourth and fifth EOF components represent respectively the variations of the thickness of the ionosphere in the southern and northern hemispheres. It is found that the obtained eigen series converges quickly, with the first five EOF components contributing as high as 98% of the variances of the Ne dataset. We then modeled the EOF coefficients A i obtained by the EOF decomposition using the harmonic functions representing the annual and semi-annual variations, with the solar cycle dependences being taken into account by including the changes of the harmonic amplitudes with the solar irradiance flux index F10.7. The Ne model is constructed using the A i thus modeled and E i obtained by the EOF decomposition. Comparison between the Ne model output results and the COSMIC-2 observational data showed that the modeled results capture well the global structures of the observational data and the model output results have very high linear correlation coefficients with the observational ones (R > 0.9), justifying the modeling technique used in our present study.

Flow structures over mesophotic coral ecosystems in the eastern Gulf of Mexico

Simultaneous time series of current velocity profiles are used to characterize flow structures over intermediate-depth coral ecosystems in the eastern Gulf of Mexico. Understanding of temporal variability and spatial coherence in flow is necessary to establish connectivity among these ecosystems. Time series were collected at Pulley Ridge (the westernmost site), Northern Dry Tortugas, and Southern Dry Tortugas. Overlapping data spanned the period from March 22, 2013 to June 20, 2015. The strongest currents were approximately 1 m s-1 southeastward at Pulley Ridge. Subtidal velocities from the three sites were decomposed into real-vector, concatenated empirical orthogonal functions (EOFs). Results from EOFs indicated that Mode 1, which explained 63% of the subtidal variance, was roughly in the same direction at each of the three sites. Mode 1 directionality indicated potential interconnectivity between Pulley Ridge and Southern Dry Tortugas, and between Northern Dry Tortugas and Pulley Ridge. Mode 1 also suggested limited to no connectivity between the two Dry Tortugas sites as the flows over the two sites were parallel. Mode 2 explained close to 24% of the variance and showed incoherence among the three sites. Wavelet analysis of EOF coefficients indicated dominance of >1 week variability in this region. Flow variability may be associated with wind forcing and Loop Current variability as confirmed by satellite altimetry. Wind forcing caused part of the intra-monthly (<1 month) periodicity in flows. Sea level in the area of the Loop Current, as derived from EOF application on altimetry data during the period of velocity measurements, was related to Mode 1 of the currents at sub-monthly (>1 month) periods. The relationship was more robust, but inverse, when comparing sea level off the northwestern coast of Cuba to the Mode 1 of the currents. These results characterize physical connectivity among South Florida coral ecosystems and have biophysical implications for coral fish populations.

Global total electron content prediction performance assessment of the IRI-2016 model based on empirical orthogonal function decomposition

Abstract. In this study, the empirical orthogonal function (EOF) decomposition technique was utilized to analyze the similarities and differences of the spatiotemporal characteristics between the total electron content (TEC) of the International GNSS Service global ionospheric map (GIM) and that derived from the International Reference Ionosphere 2016 (IRI-2016) model in 2013. Results showed that the main spatial patterns and time-varying features of the data set have good consistency. The following four main spatiotemporal variation features can be extracted from both data sets through EOF decomposition: the variation with the geomagnetic latitude reflecting the daily averaged solar forcing, the diurnal and semidiurnal periodic changes with longitude due to local time, and the interhemispheric asymmetry caused by the annual variation of the inclination angle of the Earth's orbit. The differences between the spatial patterns represented by the EOF base functions of IRI-2016 and GIM TECs were analyzed by extracting the same time-varying coefficients. The deviations of the interhemispheric asymmetry component between the two data sets showed roughly equal values throughout the Southern or Northern Hemisphere, whereas those of the other spatial modes were mainly concentrated on the equatorial region. The differences of the time-varying characteristics between the IRI-2016 and GIM TECs were also compared by extracting the same EOF base functions. Although the EOF coefficients of the two data sets presented consistent seasonal variations, the magnitude of IRI-2016 TEC changes over time was less than that of GIM TEC. The diurnal variation of the daily averaged solar forcing component and the annual variation of the interhemispheric asymmetry component exhibited relatively large deviations between the two data sets. Considering the variance contribution of the different EOF components and their average relative deviations, both analyses showed that the daily averaged solar forcing and interhemispheric asymmetry components were the main factors for the deviation between the IRI-2016 and GIM TECs.

An Improved Particle Filtering Technique for Source Localization and Sound Speed Field Inversion in Shallow Water

Both source localization and environmental inversions are practical problems for long-standing applications in underwater acoustics. This paper presents an approach of the moving source localization and sound speed field (SSF) inversion in shallow water. The approach is formulated in a state-space model with a state equation for both the source parameters (e.g., source depth, range, and speed) and SSF parameters (first three empirical orthogonal function coefficients, EOFs) and a measurement equation that incorporates underwater acoustic information via a vertical line array (VLA). As a sequential processing algorithm that operates on nonlinear systems with non-Gaussian probability densities, an improved sequential importance resampling type particle filtering (SIR PF) is proposed to counter degeneracy. The improved PF performs tracking of source and SSF parameters simultaneously, and evaluates their uncertainties in the form of time-evolving posterior probability densities (PPDs). The performance of improved PF is illustrated with well-tracked simulations of real-time source localization and time-varying SSF inversion. Moreover, the influence of different particle numbers on PF tracking accuracy and computational cost is also demonstrated. Simulation results show that the high-particle-number PF has an outperform performance. For a given hardware system, the reasonable compromise between accuracy and computational cost is a matter of tradeoff.

Joint Inversion for Sound Speed Field and Moving Source Localization in Shallow Water

This paper develops a joint approach for time-evolving sound speed field (SSF) inversion and moving source localization in shallow water environment. The SSF is parameterized in terms of the first three empirical orthogonal function (EOF) coefficients. The approach treats both first three EOF coefficients and source parameters (e.g., source depth, range and speed) as state vectors of evolving with time, and a measurement vector that incorporates acoustic information via a vertical line array (VLA), and then the inversion problem is formulated in a state-space model. The processors of the extended Kalman filter (EKF) and ensemble Kalman filter (EnKF) are used to estimate the evolution of those six parameters. Simulation results verify the proposed approach, which enable it to invert the SSF and locate the moving source simultaneously. The root-mean-square-error (RMSE) is employed to evaluate the effectiveness of this proposed approach. The interfile comparison shows that the EnKF outperform the EKF. For the EnKF, the robustness of the approach under the sparse vertical array configuration is verified. Moreover, the impact of the source-VLA deployment on the estimation is also concerned.

The Linkage of the Precipitation in the Selenga River Basin to Midsummer Atmospheric Blocking

The linkage between atmospheric blocking (blocking frequency, BF) and total monthly July precipitation in the Selenga River Basin, the main tributary of Lake Baikal, for the period 1979–2016 was investigated. Based on empirical orthogonal functions (EOF) analysis, two dominant modes of precipitation over the Selenga River Basin were extracted. The first EOF mode (EOF 1) is related to precipitation fluctuations mainly in the Mongolian part of Selenga; the second EOF mode (EOF 2)—in the Russian part of Selenga. Based on two different modes obtained, the total amount of precipitation individually for the Russian and Mongolian part of Selenga was calculated. Correlation analysis has demonstrated that precipitation over the Mongolian part of the Selenga Basin is positively correlated to blocking over Eastern Siberia/Mongolia (80–120° E, ESM-BF). Precipitation over the Russian part of the Selenga Basin is positively correlated to blocking over the Urals-Western Siberia (50–80° E, UWS-BF) and European blocking (0–50° E, E-BF). However, the linkage is not stable, and after the mid-1990s, the obtained positive correlation became insignificant. The analysis has shown that the dominance of E or ESM-blocking in July was the primary driver of the existence of two precipitation modes over the Selenga River Basin. During 1996–2016, the negative trend of time coefficients of EOF 1 and 2 for precipitation in Selenga had been observed, which was characterized by displacement of positive precipitation anomalies outside the basin. At the same time, there was a weakening of the linkage between precipitation in the Selenga Basin and blocking frequency. We have revealed two wave-like modes over Northern Eurasia and the subtropical part of Eurasia corresponding to E, ESM-blocks in 1979–1995 and 1996–2016. The change of the Northern and subtropical wave modes is one of the causes for the weakening of the linkage between atmospheric blocking and precipitation in the Selenga Basin and as a consequence decreased precipitation in the Russian and Mongolian part of Selenga during 1979–2016.

An Ionosphere Specification Technique Based on Data Ingestion Algorithm and Empirical Orthogonal Function Analysis Method

AbstractA data ingestion method in reproducing ionospheric electron density and total electron content (TEC) was developed to incorporate TEC products from the Madrigal Database into the NeQuick 2 model. The method is based on retrieving an appropriate global distribution of effective ionization parameter (Az) to drive the NeQuick 2 model, which can be implemented through minimizing the difference between the measured and modeled TEC at each grid in the local time‐modified dip latitude coordinates. The performance of this Madrigal TEC‐driven‐NeQuick 2 result is validated through the comparison with various International Global Navigation Satellite Systems Services global ionospheric maps and ionosonde data. The validation results show that a general accuracy improvement of 30–50% can be achieved after data ingestion. In addition, the empirical orthogonal function (EOF) analysis technique is used to construct a parameterized time‐varying global Az model. The quick convergence of EOF decomposition makes it possible to use the first six EOF series to represent over 90% of the total variances. The intrinsic diurnal variation and spatial distribution in the original data set can be well reflected by the constructed EOF base functions. The associated EOF coefficients can be expressed as a set of linear functions of F10.7 and Ap indices, combined with a series of trigonometric functions with annual/seasonal variation components. The NeQuick TEC driven by EOF‐modeled Az shows 10–15% improvement in accuracy over the standard ionosphere correction algorithm in the Galileo navigation system. These preliminary results demonstrate the effectiveness of the combined data ingestion and EOF modeling technique in improving the specifications of ionospheric density variations.

Modeling and analysis of GPS-TEC low latitude climatology during the 24th solar cycle using empirical orthogonal functions

The Total Electron Content (TEC) is an essential component describing the temporal and spatial characteristics of the ionosphere. In this paper, an empirical orthogonal function (EOF) model is constructed by using ground based Global Navigational Satellite System (GNSS) TEC observation data at the Bangalore International GNSS Service (IGS) station (geographic – 13.02° N, 77.57° E; geomagnetic latitude 4.4° N) during an extended period (2009–2016) in the 24th solar cycle. EOF model can be decomposed into base functions and its corresponding coefficients. These decomposed modes well represented the influence of solar and geomagnetic activity towards TEC. The first three EOFs modes constitute about 98% of the total variance of the observed data sets. The Fourier Series Analysis (FSA) is carried out to characterize the solar-cycle, annual and semi-annual dependences by modulating the first three EOF coefficients with solar (F10.7) and geomagnetic (Ap and Dst) indices. The TEC model is validated during daytime and nighttime conditions as well as under different solar activity and geomagnetic conditions. A positive correlation (0.85) of averaged daily GPS-TEC with averaged daily F10.7 strongly supports those time-varying characteristics of the ionosphere features depends on the solar activity. Further, the validity and reliability of EOF model is verified by comparing with the GPS-TEC data, and standard global ionospheric models (International Reference Ionosphere, IRI2016 and Standard Plasmasphere-Ionosphere Model, SPIM). The performances of the standard ionosphericmodelsare marked to be relatively better during High Solar Activity (HSA) periods as compared to the Low Solar Activity (LSA) periods.

Sequential Inversion for Geoacoustic Parameters in the South China Sea Using Modal Dispersion Curves

This paper presents an inversion scheme that uses two wideband explosive sources based on modal dispersion characteristics to synchronously obtain sound speed profile and bottom parameters. The signal received on a single hydrophone in a vertical line array is decomposed into a series of propagation modes within the framework of normal mode theory, and the dispersive characteristics of the modes are analyzed using time–frequency analysis. Time-warping transform is applied to resolve the propagation modes of explosive sources from the South China Sea in 2007. The relative travel time differences of the propagation modes are used to invert the environmental parameters, including empirical orthogonal function coefficients, sediment thickness, sound speed, density, basement sound speed and density. Sediment density and basement density are constrained by Hamilton’s empirical relationship, and relative mode energy ratio is used to estimate the bottom attenuation coefficient. An adaptive simplex simulated annealing algorithm is used for the inversion. The reliability of the inversion results is verified via uncertainty analysis. The transmission loss calculated using the deduced parameters matches the experimental data well.

Spatial and temporal variation of total electron content as revealed by principal component analysis

Abstract. Eleven years of global total electron content (TEC) data derived from the assimilated thermosphere–ionosphere electrodynamics general circulation model are analyzed using empirical orthogonal function (EOF) decomposition and the corresponding principal component analysis (PCA) technique. For the daily averaged TEC field, the first EOF explains more than 89 % and the first four EOFs explain more than 98 % of the total variance of the TEC field, indicating an effective data compression and clear separation of different physical processes. The effectiveness of the PCA technique for TEC is nearly insensitive to the horizontal resolution and the length of the data records. When the PCA is applied to global TEC including local-time variations, the rich spatial and temporal variations of field can be represented by the first three EOFs that explain 88 % of the total variance. The spectral analysis of the time series of the EOF coefficients reveals how different mechanisms such as solar flux variation, change in the orbital declination, nonlinear mode coupling and geomagnetic activity are separated and expressed in different EOFs. This work demonstrates the usefulness of using the PCA technique to assimilate and monitor the global TEC field.

Sound speed estimation and source localization with linearization and particle filtering

A method is developed for the estimation of source location and sound speed in the water column relying on linearization. The Jacobian matrix, necessary for the proposed linearization approach, includes derivatives with respect to empirical orthogonal function coefficients instead of sound speed directly. First, the inversion technique is tested on synthetic arrival times, using Gaussian distributions for the errors in the considered arrival times. The approach is efficient, requiring a few iterations, and produces accurate results. Probability densities of the estimates are calculated for different levels of noise in the arrival times. Subsequently, particle filtering is employed for the estimation of arrival times from signals recorded during the Shallow Water 06 experiment. It has been shown in the past that particle filtering can be employed for the successful estimation of multipath arrival times from short-range data and, consequently, in geometry, bathymetry, and sound speed inversion. Here probability density functions of arrival times computed via particle filtering are propagated backward through the proposed inversion process. Inversion estimates are consistent with values reported in the literature for the same quantities. Last it is shown that results are consistent with estimates resulting from fast simulated annealing applied to the same data.

Sound speed estimation and source localization with particle filtering and a linearization approach

In previous work, a particle filtering method was developed that provided estimates of multipath arrival times from short-range data and, subsequently, employed them in geometry, bathymetry, and sound speed inversion. The particle filter provided probability density functions of arrival times, that were then “propagated” backwards through a sound propagation model for inversion. That implies that every particle from the probability density is employed in the inversion scheme, creating a potentially computationally cumbersome process. In this work, we develop a new method for such parameter estimation which relies on linearization. The novel aspect is that the Jacobian matrix now includes derivatives with respect to Empirical Orthogonal Function coefficients. The approach, requiring only a few iterations to converge, is particularly efficient. Results from the application of this technique to synthetic and real (SW06) data are presented and compared to full-field inversion estimates. [Work supported by ONR and the NSF CSUMS program.]

Dual-Regression Retrieval Algorithm for Real-Time Processing of Satellite Ultraspectral Radiances

AbstractA fast physically based dual-regression (DR) method is developed to produce, in real time, accurate profile and surface- and cloud-property retrievals from satellite ultraspectral radiances observed for both clear- and cloudy-sky conditions. The DR relies on using empirical orthogonal function (EOF) regression “clear trained” and “cloud trained” retrievals of surface skin temperature, surface-emissivity EOF coefficients, carbon dioxide concentration, cloud-top altitude, effective cloud optical depth, and atmospheric temperature, moisture, and ozone profiles above the cloud and below thin or broken cloud. The cloud-trained retrieval is obtained using cloud-height-classified statistical datasets. The result is a retrieval with an accuracy that is much higher than that associated with the retrieval produced by the unclassified regression method currently used in the International Moderate Resolution Imaging Spectroradiometer/Atmospheric Infrared Sounder (MODIS/AIRS) Processing Package (IMAPP) retrieval system. The improvement results from the fact that the nonlinear dependence of spectral radiance on the atmospheric variables, which is due to cloud altitude and associated atmospheric moisture concentration variations, is minimized as a result of the cloud-height-classification process. The detailed method and results from example applications of the DR retrieval algorithm are presented. The new DR method will be used to retrieve atmospheric profiles from Aqua AIRS, MetOp Infrared Atmospheric Sounding Interferometer, and the forthcoming Joint Polar Satellite System ultraspectral radiance data.

Joint projections of North Pacific sea surface temperature from different global climate models

The goal of this work is to develop a general methodology to obtain joint projections of climate indexes, based on ensembles of global climate model (GCM) output and historical records. As a case study, we consider sea surface temperature (SST) in the North Pacific Ocean. We use two ensembles of 17 different GCM results, made available in the fourth assessment report of the Intergovernmental Panel on Climate Change: one corresponds to 20th‐century forcing conditions, and the other corresponds to the A1B emission scenario for the 21st century. Given a representation of the SST spatio‐temporal fields based on a common set of empirical orthogonal functions, we use a hierarchical Bayesian model for the empirical orthogonal function coefficients to estimate a baseline and a set of model discrepancies. These components are all time varying. The model enables us to extract relevant temporal patterns of variability from both the observations and the simulations and obtain common patterns from all 18 series. This is used to obtain unified 21st‐century forecasts of relevant oceanic indexes as well as whole fields of North Pacific SST forecast. We compare the forecast index for different timescales and compare the SST reconstructions with the GCMs for the 21st century. Although the coarser time resolution produces clearer and faster results, we show that finer timescales produce results with structures that are similar to ones obtained at coarser scales. Copyright © 2012 John Wiley &amp; Sons, Ltd.

Improved statistical prediction of surface currents based on historic HF-radar observations

Accurate short-term prediction of surface currents can improve the efficiency of search-and-rescue operations, oil-spill response, and marine operations. We developed a linear statistical model for predicting surface currents (up to 48 h in the future) based on a short time history of past HF-radar observations (past 48 h) and an optional forecast of surface winds. Our model used empirical orthogonal functions (EOFs) to capture spatial correlations in the HF-radar data and used a linear autoregression model to predict the temporal dynamics of the EOF coefficients. We tested the developed statistical model using historical observations of surface currents in Monterey Bay, California. The predicted particle trajectories separated from particles advected with HF-radar data at a rate of 4.4 km/day. The developed model was more accurate than an existing statistical model (drifter separation of 5.5 km/day) and a circulation model (drifter separation of 8.9 km/day). When the wind forecast was not available, the accuracy of our model degraded slightly (drifter separation of 4.9 km/day), but was still better than existing models. We found that the minimal length of the HF-radar data required to train an accurate statistical model was between 1 and 2 years, depending on the accuracy desired. Our evaluation showed that the developed model is accurate, is easier to implement and maintain than existing statistical and circulation models, and can be relocated to other coastal systems of similar complexity that have a sufficient history of HF-radar observations.

Variations and prediction of the annual number of tropical cyclones affecting Korea and Japan

AbstractThis paper investigates the factors affecting the annual number of tropical cyclones (TCs) passing within 100 km of the coast of Korea and Japan (KJ). Using a training set consisting of the coefficients of empirical orthogonal functions of these factors between 1965 and 2005, equations are derived to predict the annual number of these TCs over the whole season in April, and to update this prediction for the July‐November (JN) season in June. Results show that the El Niño‐Southern Oscillation (ENSO) plays a key role in determining the behaviour of TCs affecting KJ, with more TCs affecting the region during El Niño and less during La Niña. This difference can be attributed to ENSO modifying the flow patterns, which in turn affects TC behaviour. The prediction equations suggest that the 500‐hPa geopotential height is more important in determining the number of TCs affecting KJ in the whole season, while both the 850‐hPa geopotential height and 850‐hPa vorticity play a role in the JN season. Both prediction schemes are able to produce acceptable results, with forecast skills of 42.3 and 33.3% over climatology for the whole, and JN seasons, respectively. The predicted TC number for 2006, 2007, and 2008 are also mostly accurate to within one standard deviation of the observed number. Copyright © 2010 Royal Meteorological Society

Decadal variability and scales of the sea surface structure in the northern Ionian

The variability and scales of the sea surface structure of the northern Ionian Sea from January 1993 to December 2007 were studied by means of altimeter remotely-sensed weekly Sea Level Anomaly (SLA) objective maps. Variability in the sea surface structure was addressed by means of empirical orthogonal function (EOF) analysis and, assuming an exponential correlation model, scales of the SLA field were quantified as e-folding distances of the SLA autocorrelation function. The variability in the sea surface structure, described by the first three EOFs, which cumulatively explain 60.3% of the data set variance, is characterized by a large-scale structure with variability on a time scale of ∼10–13 years and, on shorter scales, an eddy system with variability on an annual scale. The variability in the large-scale structure describes an overturning of the SLA field, which took place in 1997, and determines a reversal of the geostrophic upper-layer circulation. As the large-scale circulation transition takes place, time-dependent spectral analysis of EOF coefficients shows a redistribution of the spectral energy from inter-annual to semi-annual and monthly components. Spatial scales display variability on an annual and inter-annual time scale. On the annual time scale, variability in spatial scales is characterized by longer values in summer–fall and shorter in winter–spring. Inter-annual variability in spatial scales is demonstrated by a remarkable drop in the values during fall in the period 1998–2000. We propose an explanation of the variability in horizontal scales in terms of the redistribution of water masses and related modifications of the vertical structure of the water column associated with different regimes of the basin-scale circulation.