Search

The design of optimal precipitation station configuration (network) is pivotal for obtaining accurate spatiotemporal data in a cost-efficient manner in terms of high operation, management and maintenance costs of stations, and missing data completion. In the present study, historical spatiotemporal precipitation data of 18 stations located in the Upper Euphrates watershed basin are initially exposed to empirical orthogonal function (EOF) analysis to exploit the general intrinsic low dimensionality of the precipitation phenomenon. Along with the basic EOF analysis, robust and mean-centered versions are also developed to improve the prediction accuracy of spatiotemporal precipitation data and optimize the number of stations in the watershed basin. Importantly, for the first time, robust EOF (R-EOF) analysis has been carried out in a hydrological predictive study. The matrix that contains the obtained modes (EOFs) is fed into the QR factorization with a column pivoting algorithm and sparse precipitation gauge locations are identified. The assessment of the model using the Nash–Sutcliffe coefficient of efficiency (CE), root mean square error (RMSE), and mean absolute error (MAE) metrics reveals that the complete dimensional state space can be reconstructed effectively, and its future evolution can be predicted accurately even with a small number of observation stations. Remarkably, the spatiotemporal precipitation data for the entire field can be reconstructed using only four, five, 10, or 12 stations, utilizing robust mean-centered (R-MC-EOF), robust (R-EOF), mean-centered (MC-EOF), and standalone EOF models. These models demonstrate high performance with CE values of 0.96, 0.94, 0.84, and 0.81 and RMSE values of 2.2, 3.8, 5.7, and 6.8 mm, respectively. Notably, both the R-EOF and MC-EOF models outperformed their standalone counterparts in terms of model performance. When a sufficient amount of spatiotemporal data is available, the optimal number and locations of precipitation gauges can be easily determined using the QR with a column pivoting algorithm. This algorithm is user friendly and can be implemented in popular programming environments such as Python, MATLAB, and R. Due to the limited budgets and/or low accessibility conditions, challenging basin topography, and bad weather conditions, not many areas are extensively equipped with instruments to measure the precipitation; thus, high-resolution data is not always available. Acquiring reliable and accurate data is critical for water resources management, flood and drought warning, irrigation networks, hydrological (e.g., watershed, rain-runoff) modeling, and urban and environmental planning. This renders the proposed methodology very crucial in obtaining high-fidelity spatiotemporal data. [ABSTRACT FROM AUTHOR]

In East Asia, where concentrated summer precipitation often leads to climate disasters, understanding the factors that cause such extreme rainfall is crucial for effective forecasting and preparedness. The western North Pacific subtropical high (WNPSH) is a key driver of summer precipitation variability, and therefore, its monitoring is critical to predicting the wet or dry periods during the East Asian summer monsoon. Using the Geo-KOMPSAT 2A (GK2A) satellite cloud amount data and ERA5 reanalysis data during the years 2020–2023, this study identified three leading empirical orthogonal function (EOF) modes and investigated the associated WNPSH variability at synoptic and subseasonal scales. The analysis includes a linear regression of meteorological fields onto the principal component (PC) time series. All three modes play a role in the spatiotemporal variability of the WNPSH, exhibiting lead–lag relationships. In particular, the second mode is responsible for its northwestward shift and intensification. As the WNPSH moves northwestward, the position of the monsoon rain band also shifts, and its intensity is modulated mainly by the moisture transport along the WNPSH boundary. Our results highlight the potential of high-resolution, real-time data from the GK2A satellite to elucidate WNPSH variability and its impact on the East Asian summer monsoon. By addressing the variability of the WNSPH using GK2A data, we pave the way for the development of a real-time monitoring framework with GK2A, which will improve our predictability and readiness for extreme weather events in East Asia. [ABSTRACT FROM AUTHOR]

Underwater sound speed is one of the most significant factors that affects high-accuracy underwater acoustic positioning and navigation. Due to its complex temporal variation, the forecasting of the underwater sound speed field (SSF) becomes a challenging task. Taking advantage of machine learning methods, we propose a new method for SSF forecasting based on the least square support vector machine (LSSVM) and a multi-parameter model, aiming to enhance the forecasting accuracy of underwater SSF with hourly resolution. We first use a matching extension method to standardize profile data and train the LSSVM with the parameters of observation time, temperature, salinity, and depth. We then employ radial basis function kernels to construct the forecasting model of SSF. We validate the feasibility and effectiveness of the LSSVM model by comparing it with the polynomial fitting (PF) and back propagation neural network (BPNN) methods, using hourly data obtained from the measured data and open data. The results show that the means of the root mean square for the LSSVM based on the observation time parameter and the LSSVM based on the multi-parameter model achieve 0.51 m/s and 0.45 m/s, respectively, presenting a significant improvement compared with the PF (0.82 m/s) and BPNN (0.76 m/s) methods. [ABSTRACT FROM AUTHOR]

Subsidence occurs from excessive groundwater drawdown, but varies in response to underlying hydrogeologic conditions, land use factors, and variations of pumping rates. For subsidence feature decomposition, the empirical orthogonal function (EOF) is used to identify to extract the main components of the land subsidence data, such as continuous trend of subsidence and seasonal subsidence from various regions. Result shows that the major subsidence feature components contain the long-term, periodic (seasonal), and intra-seasonal ones which are related to human activities and hydrogeology from the inland, distal-fan area and coastal area in west-central Taiwan. The subsidence trend and seasonal variation at the observations can be separated from empirical mode decomposition (EMD) for validation. Moreover, subsidence and groundwater monitoring data are used to generate the stress–strain relations at the major EOFs locations. The outcome implies a strongly elastic nature, yet reveals a diverse correlation between stress and strain within the subsidence region. The decomposition and identification of subsidence features offer valuable applications for the effective management of land subsidence and groundwater resources. [ABSTRACT FROM AUTHOR]

Many mobile marine taxa are changing their distributions in response to climate change. Such movements pose a challenge to fisheries monitoring and management, particularly in systems where climate‐adaptive and ecosystem‐based management objectives are emphasized. While shifts in species distributions can be discerned from long‐term fisheries‐independent monitoring data, distilling coherent patterns across space and time from such datasets can be challenging, particularly for transboundary stocks. One approach for identifying dominant patterns of spatiotemporal variation that has been widely used in physical atmospheric and oceanographic studies is empirical orthogonal function (EOF) analysis, wherein spatiotemporal variation is separated into time‐series of annual factor loadings and spatial response maps. Here, we apply an extension of EOF analysis that has been modified for compatibility with biological sampling data to a combined US–Russian fisheries‐independent survey dataset that spans the eastern (United States) and western (Russia) Bering Sea shelf to estimate dominant patterns of spatiotemporal variation for 10 groundfish species at a shelf‐wide scale. EOF identified one axis of variability that was coherent with the extent of cold (≤0°C) near‐bottom waters (the cold pool) previously shown to be a key influence on species distributions and ecosystem structure for the Bering Sea. However, the leading axis of variability identified by our EOF analysis was characterized by low frequency changes in the distributions of several species over longer time scales. Our analysis has important implications for predicting variation in species distributions over time and demonstrates a widely applicable method for leveraging combined fisheries‐independent survey datasets to characterize community‐level responses to ecosystem change at basin‐wide scales. [ABSTRACT FROM AUTHOR]

The increasing global climate change has resulted in more frequent drought disasters, and using the drought index to assess spatial and temporal changes accurately is of practical importance. In this study, the Spatio-temporal evolution characteristics of different drought indicators at annual and seasonal scales were analyzed by combining spatial interpolation, correlation analysis, trend tests, and the empirical orthogonal function based on the meteorological drought composite index (MCI) of Penman–Monteith. The study focused on drought research in the Pearl River basin from 1961 to 2020. The results indicated that (1) on the annual scale, drought exhibited a high spatial distribution in the west and low in the east, as well as high in the south and low in the north. The frequency, intensity, and extent displayed a non-significant decreasing trend, with the drought degree being mild in the past ten years. (2) On the seasonal scale, drought exhibited seasonal variability, particularly in spring and autumn, which were inversely distributed. Summer drought was the least severe, with the drought range mostly below 10%, while winter drought was the severest, exceeding 50%, indicating an areawide drought. (3) The consistent change of drought intensity across the basin is the primary mode, and the reverse distribution of east–west and north–south is the secondary mode. The overall higher year of drought intensity is close to 24 years, and the overall lower year is close to 28 years. The study based on the MCI can provide a reference for drought research and management in different regions. [ABSTRACT FROM AUTHOR]

Community-based tourism is a sustainable form of tourism development where tourists visit residential communities to interact with local lives and cultures for an enhanced travel experience. Identifying and tracking tourist activities in community-based tourism is particularly challenging, as tourists have shared activity spaces with residents. The paper proposes a new method to study the space–time patterns of the tourist flow using Wi-Fi data. Specifically, we have tracked Wi-Fi probe requests over six months in the Shichahai scenic area, a famous community-based tourist attraction in Beijing, China. After deriving the tourist flow from the Wi-Fi data, we have applied the empirical orthogonal function (EOF) method to the identification of the spatial aggregation pattern and the temporality of the tourist flow. A follow-up explanatory analysis examines the environmental impacts, such as weather conditions, air quality, and travel days, on the space–time patterns. The study is among the first to employ Wi-Fi data to study travel behaviours in community-based tourism. The proposed method can shed insights into a better understanding of tourist behaviours in open-space, tourism-oriented urban communities. [ABSTRACT FROM AUTHOR]

Winter (December to March) precipitation is the major source of rainfed agriculture, storage, and perennial water flow in the western river system of Pakistan. Hence, this study uses precipitation data and variables of land–ocean and atmosphere from the Pakistan Meteorological Department and European Centre for Medium-Range Weather Forecasts (ECMWF) and fifth-generation reanalysis data (ERA5), respectively, to investigate the changes in winter precipitation and its sensitivity to different land–ocean and atmosphere variables, which are rarely investigated in Pakistan. Non-parametric techniques, such as the modified Mann–Kendal, Sen slope, kernel density-based probability function (PDF), empirical orthogonal function (EOF), and correlation analysis, were used to assess the changes and modes of variability in winter precipitation. The overall seasonal precipitation showed a significant decreasing trend with a (−0.1 mm d−1 yr−1) in the seasonal mean and monthly precipitation, except in February which showed a significant increase (>0.11 mm d−1 yr−1). The highest decrease in daily precipitation (<−0.1 mm d−1 yr−1) was in the north, with a moderate decrease in the southeast. The extreme precipitation indices exhibited an erratic decreasing tendency, but the maximum daily precipitation index increased; post-2000 precipitation extremes displayed an increase, and the seasonal and monthly precipitation exhibited the highest deviations during the drought period (1995–2000). The leading precipitation mode (EOF1) is sensitive to the local land surface processes and significantly correlated (>0.60) with the central Pacific and Indian Ocean's basin-wide sea surface temperature, corroborating the influence of ENSO-induced meridional/zonal deviation of Hadley–Walker circulations. The Hadley and Walker cells affect the south-westerlies' jet stream strength, impacting the water vapor transport and precipitation over Pakistan. These changes in the precipitation magnitude will affect rain-fed agriculture, especially the Rabi cropping pattern and perennial river flow. [ABSTRACT FROM AUTHOR]

Underwater sound speed is one of the most significant factors that affects high-accuracy underwater acoustic positioning and navigation. Due to its complex temporal variation, the forecasting of the underwater sound speed field (SSF) becomes a challenging task. Taking advantage of machine learning methods, we propose a new method for SSF forecasting based on the least square support vector machine (LSSVM) and a multi-parameter model, aiming to enhance the forecasting accuracy of underwater SSF with hourly resolution. We first use a matching extension method to standardize profile data and train the LSSVM with the parameters of observation time, temperature, salinity, and depth. We then employ radial basis function kernels to construct the forecasting model of SSF. We validate the feasibility and effectiveness of the LSSVM model by comparing it with the polynomial fitting (PF) and back propagation neural network (BPNN) methods, using hourly data obtained from the measured data and open data. The results show that the means of the root mean square for the LSSVM based on the observation time parameter and the LSSVM based on the multi-parameter model achieve 0.51 m/s and 0.45 m/s, respectively, presenting a significant improvement compared with the PF (0.82 m/s) and BPNN (0.76 m/s) methods.

The development of global navigation satellite system and low Earth orbit (LEO) satellite technology provides favorable conditions and opportunities for studying the temporal and spatial variation of plasmaspheric electron content (PEC). This paper carried out PEC detection and modeling using the MetOp series of satellites. The global PEC time series of two local times (09:30LT and 21:00LT) from October 1, 2019 to September 30, 2020 were constructed, and the ionospheric total electron content (TEC) values at the corresponding local times were extracted from IGS global ionosphere maps. PEC and TEC are higher during daytime than nighttime, but the diurnal difference of PEC is considerably smaller than that of TEC. The difference in PEC between the winter and summer hemispheres is not as evident as that of TEC. The results of empirical orthogonal function method show that the main spatial distribution pattern of PEC is the variation with the latitude. While, the latitude variation and interhemispheric asymmetry are the main spatial distribution modes of TEC. Unlike the semi-annual variation cycle of the TEC latitude distribution, the latitude variation of PEC has an annual cycle. In addition, spatiotemporal patterns of PEC at 09:30LT and 21:00LT are markedly consistent, and there is only minimal difference between the two moments in the longitude distribution difference component. The EOF decomposition results provide a further explanation of the differences between PEC and TEC. [ABSTRACT FROM AUTHOR]

The station observations and reanalysis dataset are utilized to explore the effects of Madden-Julian Oscillation (MJO) on the cold extremes over the whole country of China, and the possible mechanisms from the perspective of the thermodynamics. Here, we focus on the principal modes of different phases of MJO in winter and their influences on the large scale cold events (LSCEs) that are identified firstly. The time evolution and the spatial features of LSCEs are checked for a basic insight of the LSCEs in China, of which the annual variability and regional differences are obvious among the chosen LSECs. In addition, the first (second) empirical orthogonal decomposition mode of MJO shows an opposite feature that positive values in Phases 1–3 (Phase 4–5) and negative values in Phases 5–7 (Phase 7–8), with the explanation of variance at 30.6% (26%). Furthermore, according to the threshold of ±1.5 in standardized time series of two principal components (PC1 and PC2), the events are chosen and clarified into four cases (+PC1, –PC1, +PC2 and –PC2). All the MJO-related cases present the increases of LSCEs but with regional and intensity differences. For the case of +PC1, the cold advection from higher latitudes transport to eastern Asia inducing negative temperature anomalies thereof. For the case of –PC1, besides the eastern Asian region, there still the cold advections across the Inner Mongolia regions, leaving negative anomalies over the region either. For the case of +PC2, the southward wind and the accompanied cold advection are stronger than the others affecting mostly regions in China, which leads to the more decreases of temperature. For the case of –PC2, the cold advections are weaker, resulting in the less temperature decreases over the southeastern China. Meanwhile, the tropospheric cyclonic and anti-cyclonic circulation anomalies are beneficial to the persistence of local extremes.

It is of utmost necessity to understand the dynamics of regional active accumulated temperature (AAT) to cope with the negative impacts of global warming on agroforestry development and food security and to provide a real-time and effective reference basis for regional agroforestry planning. The daily temperature data from 30 meteorological stations in Sichuan Province from 1970 to 2020, and sea surface temperature (SST) index data from the Atlantic Multiphase Oscillation (AMO) and Pacific Decadal Oscillation (PDO) were used for the study. Sichuan Province was divided into the western region (WS) and the eastern region (ES), considering 1000 m above sea level as the boundary. The spatiotemporal characteristics of ≥0°C and ≥10°C active accumulated temperature (AAT0, AAT10) in WS and ES were analyzed comprehensively using 5-day average sliding, empirical orthogonal function (EOF), ensemble empirical mode decomposition (EEMD), and multiple mutation tests. The results show that (1) AAT0 and AAT10 of WS ranged from 3034°C to 3586°C and 1971°C to 2636°C, respectively, while the AAT0 and AAT10 of ES ranged from 5863°C to 6513°C and 4847°C to 5875°C, respectively. The period around 1997 was a significant abrupt change, and the AAT in the province generally increased during the subsequent time period (2) AAT in the study area is mainly driven by the fluctuations of AMO, as reflected by the low-to-high variation of AAT coinciding with the jump of the cold-to-warm phase of AMO. Considering different time scale fluctuations in the past 51 years, the major cycle for both AAT0 and AAT10 in WS is 3.40 a, while the major cycles in ES are 3.64 a and 3.19 a, respectively with a sub-cycle of 7.29 a. AAT fluctuation has an insignificant periodic characteristic of 25.50 a on the interdecadal scale (3) The spatial heterogeneity of AAT in WS is prominent and is mainly reflected by the significantly warm conditions in the south of the WS region and relatively slight warm conditions in the north, as well as by the isolated cooling area in the form of "freezing point", i.e., Xiaojin county. In contrast, the spatial variability of AAT in ES is more or less consistent, with the warming areas concentrated in the foothills of the western edge of the basin and a slight increase in AAT observed in the central part of the basin. [ABSTRACT FROM AUTHOR]

To better understand the impact of climate forcing on Greenland ice sheet (GrIS), we study the GrIS mass variations on the interannual timescale as observed by the satellite mission of Gravity Recovery and Climate Experiment (GRACE). By employing the method of empirical orthogonal function) analysis, we find: (a) That the GrIS interannual mass variations are significantly correlated with the Pacific Decadal Oscillation, suggesting a connection to the changes of the Icelandic Low (a permanent low‐pressure system) related to PDO; (b) An East‐West Costal Dipole related to precipitation anomaly subjects to the North Atlantic Oscillation; (c) Certain contribution of the Atlantic Multidecadal Oscillation to GrIS mass variations in the form of temperature and runoff anomalies. Plain Language Summary: For the past few decades, Greenland ice sheet (GrIS) has undergone accelerated melting under impacts of the global climate changes, leading to significant change in the sea level. GrIS has been well monitored by the Gravity Recovery and Climate Experiment (GRACE) satellite mission that observes Earth's gravity change due to mass transports since its launch in 2002. Here we use GRACE data to understand the responses of GrIS to climatic forcing, which can reveal certain phenomena that are hitherto unreported: (a) We show that the Pacific Decadal Oscillation plays a significant role on GrIS variations; (b) We also report a "seesaw" pattern of an East‐West Costal Dipole related to the North Atlantic Oscillation; (c) The Atlantic Multidecadal Oscillation also contributes to GrIS mass variations. Key Points: Greenland interannual mass variations are correlated with the Pacific Decadal OscillationWe report an East‐West Coastal Dipole related to the North Atlantic OscillationThe contribution of the Atlantic Multidecadal Oscillation is also captured [ABSTRACT FROM AUTHOR]

The MODIS land surface temperature (LST) products contain large areas of missing data due to cloud contamination. Interpolating clear-sky equivalent LSTs on those areas is a first step in a stepwise approach toward fully recovering missing data. A previous study (viz. the Yu method) has implemented an effective clear-sky interpolation method, especially targeting large-area missing data. The Yu method postulates several global reference LST images that contain over 90% of valid pixels and that are assumed to have a close statistical relationship to the interpolated images. However, in practice, such reference images are rarely available throughout a one-year cycle, and the time gaps between the available reference images and the interpolated images are often huge, resulting in compromised interpolation accuracy. In this study, we intended to address those weaknesses and propose a novel clear-sky interpolation approach. The proposed approach uses multiple temporally proximate images as reference images, with which multiple initial estimates are made by an empirically orthogonal function method and then fused by a Bayesian approach to achieve a best estimate. The proposed approach was compared through two experiments to the Yu method and two other widely used methods, i.e., harmonic analysis of time series and co-kriging. Both experiments demonstrate the superiority of the proposed approach over those established methods, as evidenced by higher spatial correlation coefficients (0.90-0.94) and lower root-mean-square errors (1.19-3.64 °C) it achieved when measured against the original data that were intentionally removed.

This article proposes a new sea surface wind speed (SSWS) retrieval modeling algorithm based on the empirical orthogonal function (EOF) analysis for observations acquired by the global navigation satellite system reflectometry (GNSS-R). As a nonparametric modeling algorithm, it is simpler compared with the nonlinear methods. The influence of wind speed and incident angle on the modeling error is analyzed for the first time using a spectrum analysis. Three types of data from 80% CYGNSS 2019–2020 observations [delay Doppler map average (DDMA) and leading edge slope (LES)], signal incident angle, and the European Centre for Medium-Range Weather Forecasts Reanalysis V5 (ERA5) reference wind speed are used in the EOF analysis to establish two retrieval models. The remaining 20% of the data are used for accuracy evaluation after getting the final wind speed by the minimum variance (MV) estimator. As a result, when using three 0–20-m/s wind speeds of ERA5, Advanced Scatterometer (ASCAT), and the Modern-Era Retrospective Analysis for Research and Applications V2 (MERRA2) as contrasts, the root mean squared errors (RMSEs) are 1.51, 1.45, and 1.43 m/s, respectively. Compared with CYGNSS wind product, the performance of this algorithm is closer to the L2 Climate Data Record (CDR) V1.1 product than V1.0. The results demonstrate that the EOF algorithm has a good performance in retrieving SSWS and can better retain the influence of the incident angle on the observations. [ABSTRACT FROM AUTHOR]

The Lancang‐Mekong River Basin (LMRB) is home to ~70 million people whose life and livelihood are mostly dependent upon precipitation as the primary freshwater source. Hence, identifying potential oceanic–atmospheric drivers of regional precipitation variability is becoming increasingly important for the sustainable development of the LMRB. This study first investigated spatio‐temporal variability and trends in extreme precipitation characteristics (in terms of intensity, frequency, and duration) throughout the LMRB during 1952–2015, using gauge‐based gridded daily precipitation time series. Then, the associations between the historical extreme precipitation characteristics and seven teleconnection and five summer monsoon indices were explored. On the basin scale, no statistically significant (p <.05) trends were detected in annual extreme precipitation intensity, frequency, and duration indices. The number of wet days (R1mm) significantly increased in both the Mekong River Basin (MRB) and the Lancang River Basin (LRB), predominantly leading to longer wet spells in these two sub‐basins. Spatially, the relatively high extreme precipitation intensity and frequency indices, as well as consecutive wet days (CWD), significantly increased in the south, east, and northwest of MRB, while decreased in the west of MRB and the north of LRB. The intensity and frequency of historical extreme precipitations over the LMRB were most significantly correlated with the East Asian Summer Monsoon Index, North Atlantic Oscillation, and East Pacific/North Pacific pattern. However, the wet/dry spells showed the strongest associations with the Atlantic Multi‐decadal Oscillation/the Southern Oscillation Index on the interannual/decadal time scales (3–4/8–15 years) during 1986–1999/1968–2002, respectively. [ABSTRACT FROM AUTHOR]

Precipitation time series exhibit complex fluctuations and statistical changes. Existing research stops short of a simple and feasible model for precipitation forecasting. In this article, the authors investigate and forecast precipitation variations in South Korea from 1973 to 2021 using cyclostationary empirical orthogonal function (CSEOF) and regression methods. First, empirical orthogonal function (EOF) and CSEOF analyses are used to examine the periodic changes in the precipitation data. Then, the autoregressive integrated moving average (ARIMA) method is applied to the principal component (PC) time series derived from the EOF and CSEOF precipitation analyses. The fifteen leading EOF and CSEOF modes and their corresponding PC time series clearly reflect the spatial distribution and temporal evolution characteristics of the precipitation data. Based on the PC forecasts of the EOF and CSEOF models, the EOF–ARIMA composite model and CSEOF–ARIMA composite model are used to obtain quantitative precipitation forecasts. The comparison results show that both composite models have good performance and similar accuracy. However, the performance of the CSEOF–ARIMA model is better than that of the EOF–ARIMA model under various measurements. Therefore, the CSEOF–ARIMA composite forecast model can be considered an efficient and feasible technology representing an analytical approach for precipitation forecasting in South Korea. [ABSTRACT FROM AUTHOR]

In the northern South China Sea of western Pacific Ocean during winter, clouds, sun glint, and other factors block optic sensors, leading to a high missing rate and hence a major concern in ocean color products such as the chlorophyll-a (CHL) data. These constraints inhibit the understanding of CHL variabilities at short (< seasonal) scales. Here, we introduce a new gap-filling method to reconstruct data gaps in a daily CHL remote sensing product. We applied discrete cosine transform with penalized least square (DCT-PLS) approach in the adjacent Luzon Strait, yielding a 15-year full-coverage daily 4-km CHL product. Against the cross-validation set and an independent observational data set collected from 34 cruises, evaluations suggest that DCT-PLS has outperformed the widely applied classical data-interpolating empirical orthogonal function (DINEOF) method. Besides, the DCT-PLS method is characterized by more efficient computation. The complete CHL product was analyzed with a particular focus on the intraseasonal (~30–60 days) control on the winter bloom by the Madden-Julian Oscillation (MJO). The MJO’s local signature on the CHL presents asymmetry. The CHL peaks at the late phases of MJO events, which could be explained by the relaxation after the MJO-induced wind strengthening. This gap-filling approach can be promisingly applied in other remote sensing gap-filling problems, which could shed light on the short-term variability of biological and physical dynamics in the ocean. [ABSTRACT FROM AUTHOR]

Missing data is a critical pitfall in the investigation of remotely sensed displacement measurement because it prevents from a full understanding of the physical phenomenon under observation. In the sight of reconstructing incomplete displacement data, this letter presents a data-driven spatiotemporal gap-filling method, which is an extension of the expectation–maximization-empirical orthogonal function (EM-EOF) method. The presented method decomposes an augmented spatiotemporal covariance of a displacement time series into EOF modes and then selects the optimal set of EOF modes to reconstruct the time series. This selection is based on the cross-validation root-mean-square error and a confidence index associated with each eigenvalue. The estimated missing values are then iteratively updated until convergence. Results on displacement time series derived from cross correlation of Sentinel-2 optical images over Fox Glacier in New-Zealand’s Alps show that the reconstruction accuracy is improved compared with the EM-EOF method. The proposed extension can tackle challenging cases, i.e., short time series with heterogeneous displacement behaviors corrupted by a large amount of missing data and noise. [ABSTRACT FROM AUTHOR]

Copyright of China Sciencepaper is the property of China Sciencepaper and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The sound speed profile (SSP) is an important factor affecting the acoustic propagation characteristics of the ocean, making the accurate acquisition of SSP a crucial step in the interdisciplinary research of oceanography and underwater acoustics. Limited by the cost of in-situ measurement and the performance of the instrument itself, direct measurement of SSP inevitably leads to insufficient depth or even missing information. In this paper, we propose using partial underwater prior information (UWPI) only including underwater sound speed to obtain preliminary reconstruction results of global SSP for the first time. The empirical orthogonal function (EOF) reconstruction algorithm is optimized by employing assimilated SSP as the background SSP to further reduce reconstruction errors. The maximum global average reconstruction error and root mean square error (RMSE) after optimization decrease by >51% and 71%, respectively, which indicates that the performance of the optimized algorithm combined with partial UWPI is further improved. Finally, the performance of the optimized algorithm is discussed from the perspective of acoustic propagation. This research provides a reliable technical approach for SSP reconstruction under incomplete depth conditions, which can be applied in underwater sound field prediction and acoustic detection in the future. [Display omitted] • Global scale SSP large-depth fast reconstruction is realized based on partial UWPI for the first time. • The algorithm is optimized and the maximum reconstruction error and RMSE decrease by >51% and 71%, respectively. • Providing a reliable technical approach for SSP reconstruction under incomplete depth conditions. [ABSTRACT FROM AUTHOR]

The wind and temperature fields at 20 to 55 km above the Antigua launch site (17°N, 61°W) were analyzed by using sounding rocket data published by the research organization on Stratosphere-Troposphere Processes and their Role in Climate (SPARC). The results showed distinct variations in the wind and temperature fields at different heights from the 1960s to the 1990s. The overall zonal wind speed showed a significant increasing trend with the year, and the overall change in meridional wind speed showed a falling trend from 1976 to 1990, whereas the temperature field showed a significant cooling trend from 1964 to 1990. The times the trends mutated varied at different levels. By taking the altitudes at 20, 35, and 50 km as representatives, we found that the zonal wind speed trend had mutated in 1988, 1986, and 1986, respectively; that the meridional wind speed trend had mutated in 1990, 1986, and 1990, respectively; and that the temperature trend had mutated separately in 1977, 1973, and 1967, respectively. Characteristics of the periodic wind and temperature field variations at different heights were also analyzed, and obvious differences were found in time scale variations across the different layers. The zonal and meridional wind fields were basically characterized as having a significant periodic variation of 5 years across the three layers, and each level was characterized as having a periodic variation of less than 5 years. Temperature field variation at the three levels was basically characterized as occurring in 10-year and 5-year cycles.

The seasonal and interannual variabilities of sea surface wind (SSW) in the South China Sea (SCS), especially in coral reef regions such as Nansha Islands, Xisha Islands, Zhongsha Islands and Dongsha Islands were investigated in detail using the Blended Sea Winds dataset (1988–2017). Annual and monthly variations of SSW and sea surface temperature (SST) in the four zones were investigated. Empirical Orthogonal Function (EOF) analysis of wind field was performed to aid in better understanding the different spatial patterns. The results indicate that, as observed in the spatial distribution of the first mode of monthly mean wind speed anomaly, the magnitudes in the four island zones are all negative and are similar to each other, showing that the variations of SSW in the four island zones are consistent. In the second mode, the magnitudes in Nansha Islands are opposite to those in the other three zones. The spatial distribution of the third mode reflects regional differences. The maximum annual SSW appears in Dongsha Islands, and the minimum appears in Nansha Islands. The interannual variations of SSW in all island zones are basically concurrent. The island zones with high SSW mostly have low SST, and vice versa. There may be an inverse relationship between SSW and SST in coral reef regions in the SCS. The multi-year monthly variations of SSW in the island zones present a 'W'-shaped structural variation. Each island undergoes two months of minimum SSW every year, one during March–May (MAM) and the other during September–November (SON). Both months are in monsoon transition periods. During the months with low SSW, high SST appears. The SST peaks almost correspond to the SSW troughs. This further indicates that SSW and SST may have opposite changes in coral reef regions. Coral bleaching events often correspond to years of high SST and low SSW. [ABSTRACT FROM AUTHOR]

Ecologists and oceanographers inform population and ecosystem management by identifying the physical drivers of ecological dynamics. However, different research communities use different analytical tools where, for example, physical oceanographers often apply rank‐reduction techniques (a.k.a. empirical orthogonal functions [EOF]) to identify indicators that represent dominant modes of physical variability, whereas population ecologists use dynamical models that incorporate physical indicators as covariates. Simultaneously modeling physical and biological processes would have several benefits, including improved communication across sub‐fields; more efficient use of limited data; and the ability to compare importance of physical and biological drivers for population dynamics. Here, we develop a new statistical technique, EOF regression, which jointly models population‐scale dynamics and spatially distributed physical dynamics. EOF regression is fitted using maximum‐likelihood techniques and applies a generalized EOF analysis to environmental measurements, estimates one or more time series representing modes of environmental variability, and simultaneously estimates the association of this time series with biological measurements. By doing so, it identifies a spatial map of environmental conditions that are best correlated with annual variability in the biological process. We demonstrate this method using a linear (Ricker) model for early‐life survival ("recruitment") of three groundfish species in the eastern Bering Sea from 1982 to 2016, combined with measurements and end‐of‐century projections for bottom and sea surface temperature. Results suggest that (a) we can forecast biological dynamics while applying delta‐correction and statistical downscaling to calibrate measurements and projected physical variables, (b) physical drivers are statistically significant for Pacific cod and walleye pollock recruitment, (c) separately analyzing physical and biological variables fails to identify the significant association for walleye pollock, and (d) cod and pollock will likely have reduced recruitment given forecasted temperatures over future decades. [ABSTRACT FROM AUTHOR]

Copyright of Arid Zone Research / Ganhanqu Yanjiu is the property of Arid Zone Research Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Ecology & Rural Environment is the property of Journal of Ecology & Rural Environment Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In this paper, a data-adaptive method, namely, principal modes (PM) method, based on the spatially averaged temporal covariance of a time series of InSAR displacement measurement obtained from consecutive SAR acquisitions is proposed to retrieve the displacement signal for decorrelating targets. On wrapped interferogram time series, the PM method can highlight and restore coherent fringe patterns where they are more or less significantly hindered by decorrelation noise, whereas on unwrapped interferogram time series, the PM method provides a satisfactory separation of the displacement signal from the spatially correlated perturbations. A two-stage application of the PM method to both wrapped and unwrapped interferogram time series can significantly improve the retrieval of the displacement signal. Synthetic simulations are first performed to investigate the impact of the choice of the appropriate number of modes to retain in the empirical orthogonal function decomposition and of the time series size on the performance of the PM method, as well as to highlight the efficiency of the PM method. Then, the PM method is applied to time series of wrapped and unwrapped Sentinel 1 A/B interferograms over the Gorner glacier between October 2016 and April 2017. The main characteristics of the PM method, such as realistic assumptions, ease of implementation, and high efficiency, are highlighted. [ABSTRACT FROM AUTHOR]

On the basis of the empirical orthogonal function (EOF) decomposition technique, the spatiotemporal characteristics of the ionospheric total electron content (TEC) from 2007 to 2016 over China and adjacent areas (15°–55°N, 70°–145°E) were studied. The spatial patterns and temporal variations of the TEC were separated by EOF decomposition and can be represented well by the base functions and associated coefficients, respectively. The time variation features of different scales can also be obtained by continually decomposing the coefficients that reflect the temporal variations. Results show that this method is extremely beneficial to the comprehensive analysis of the overall spatiotemporal variations in the ionospheric TEC. The base functions of EOF reflect that the TEC in China and adjacent areas experiences annual and semiannual changes with the latitudinal and longitudinal changes, as well as equatorial anomaly phenomenon. The coefficients of these spatial patterns have evident regularity with the changes in solar activity, season, and local time (LT). After the spatial patterns and the diurnal variations with LT were extracted through a two-layer EOF decomposition, the remaining time variation features were modeled using seasonal variations, solar activity, and geomagnetic activity. Finally, an empirical TEC model was established by incorporating the modeled coefficients and the original EOF base functions. The EOF-based TEC model can demonstrate the temporal and spatial variation characteristics of TEC, and the EOF-based model can obtain significantly smaller modeling errors and achieve better performance in terms of TEC modeling than the International Reference Ionosphere model IRI-2016. [ABSTRACT FROM AUTHOR]

Copyright of Arid Zone Research / Ganhanqu Yanjiu is the property of Arid Zone Research Editorial Office and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Xi, H.; Zhang, Z.; Lu, Y., and Li, Y., 2019. Long-term and interannual variation of the steric sea level in the South China Sea and the connection with ENSO. Journal of Coastal Research, 35(3), 489–498. Coconut Creek (Florida), ISSN 0749-0208. Multisource observation and model data sets, including satellite altimetry, satellite gravimetry (the Gravity Recovery and Climate Experiment mission [GRACE]), ocean model (Estimation of the Circulation and Climate of the Ocean [ECCO]), and oceanographic reanalysis data (Ishii), are used to explore the long-term and interannual sea-level variation (SLV) in the South China Sea (SCS) and the connection with El Niño–Southern Oscillation (ENSO). From 1993 to 2012, the sea level rose at a rate of 4.7 ± 0.3 mm/y, and the steric component contributed approximately 40–55% of the increase shown in the Ishii data (1.9 ± 0.3 mm/y) and the ECCO model (2.6 ± 0.3 mm/yr). Using the GRACE observations from 2003 to 2012 for independent validation, the ECCO-derived steric trend was consistent with the mass-corrected altimetry result, whereas the Ishii data failed to capture the sea-level rise in the central basins. On the interannual scale, both the empirical orthogonal function (EOF) and the wavelet coherence (WTC) analysis indicate that the total SLV and the steric SLV have positive correlations with ENSO. The correlation is stronger between the Southern Oscillation index (SOI) and the ECCO-derived steric SLV than it is between the SOI and the Ishii-based steric SLV. In the time-frequency domain, the WTC shows a clear in-phase coherence in the 2-year cycle between the SOI and the ECCO-derived steric SLV over the entire time span; no significant coherence appears between the SOI and the Ishii-based steric SLV after 2003. The abnormal northerly winds and the increasing intrusion of low-temperature Kuroshio water into the SCS through the Luzon Strait during El Niño years may explain the connection between ENSO and the interannual steric SLV. [ABSTRACT FROM AUTHOR]

In 2020, when the novel coronavirus disease 2019 (COVID-19) broke out as a global pandemic, cities in Hubei Province first went into lockdown on 23 January and resumed work and production on 20 March. From February to March 2020, human activities in Hubei decreased significantly, with the average particulate matter smaller than 2.5 μm (PM2.5) concentration standing at 40 μg/m3, which is 21% lower than the expected based on a linear fitting trend in thePM2.5 concentration in Hubei. By using the empirical orthogonal function (EOF) method, this paper comparatively analyzes the spatial-temporal variations of Hubei’s PM2.5 concentration anomaly in February and March 2020 and the same periods of 2016–2019. The results show that the daytime peak of the PM2.5 daily variation in Hubei in a low-emission scenario during COVID-19 declined significantly, to which human activities contributed the most. However, during nighttime, the PM2.5 peak became more prominent, and the meteorological conditions had a more noticeable effect on the PM2.5 concentration. In addition, during COVID-19, there was a great drop in PM2.5 pollution accumulated from local sources within the urban circle of Wuhan City, while an increase was seen in central-western Hubei due to the inter-regional pollutant transport. Thus, the high PM2.5 concentration center in the urban circle of Wuhan disappeared, but the pollution transport channel cities in central-western Hubei remained as high-PM2.5-concentration centers.

Floods and droughts are driven, in part, by spatial patterns of extreme rainfall. Heat waves are driven by spatial patterns of extreme temperature. It is therefore of interest to design statistical methodologies that allow the rapid identification of likely patterns of extreme rain or temperature from observed historical data. The standard work-horse for the rapid identification of patterns of climate variability in historical data is Principal Component Analysis (PCA) and its variants. But PCA optimizes for variance not spatial extremes, and so there is no particular reason why the first PCA spatial pattern should identify, or even approximate, the types of patterns that may drive floods, droughts or heatwaves, even if the linear assumptions underlying PCA are correct. We present an alternative pattern identification algorithm that makes the same linear assumptions as PCA, but which can be used to explicitly optimize for spatial extremes. We call the method Directional Component Analysis (DCA), since it involves introducing a preferred direction, or metric, such as “sum of all points in the spatial field”. We compare the first PCA and DCA spatial patterns for U.S. and China winter and summer rainfall anomalies, using the sum metric for the definition of DCA in order to focus on total rainfall anomaly over the domain. In three out of four of the examples the first DCA spatial pattern is more uniform over a wide area than the first PCA spatial pattern and as a result is more obviously relevant to large-scale flooding or drought. Also, in all cases the definitions of PCA and DCA result in the first PCA spatial pattern having the larger explained variance of the two patterns, while the first DCA spatial pattern, when scaled appropriately, has a higher likelihood and greater total rainfall anomaly, and indeed is the pattern with the highest total rainfall anomaly for a given likelihood. The first DCA spatial pattern is arguably the best answer to the question: what single spatial pattern is most likely to drive large total rainfall anomalies in the future? It is also simpler to calculate than PCA. In combination PCA and DCA patterns yield more insight into rainfall variability and extremes than either pattern on its own.

Mass variations in terrestrial water storage (TWS) obtained from eight years of satellite data from the Gravity Recovery and Climate Experiment (GRACE) are used to describe low frequency TWS through Empirical Orthogonal Function (EOF) analysis. Results of the second seasonal EOF mode show the influence of the Meiyu season. Annual variability is clearly shown in the precipitation distribution over China, and two new patterns of interannual variability are presented for the first time from observations, where two periods of abrupt acceleration are seen in 2004 and 2008. GRACE successfully measures drought events in southern China, and in this respect, an association with the Arctic Oscillation and El Nino-Southern Oscillation is discussed. This study demonstrates the unique potential of satellite gravity measurements in monitoring TWS variations and large-scale severe drought in China.

In this paper, terahertz time-domain spectroscopy (THz-TDS) is used for the first time to detect fabricated defects in a glass fiber-skinned lightweight honeycomb composite panel. A novel amplitude polynomial regression (APR) algorithm is proposed as a preprocessing method. This method segments the amplitude–frequency curves to simulate the heating and the cooling monotonic behavior as in infrared thermography. Then, the method of empirical orthogonal function (EOF) imaging is applied on the APR preprocessed data as a postprocessing algorithm. Signal-to-noise ratio analysis is performed to verify the image improvement of the proposed APR-EOF modality from a quantitative point of view. Finally, the experimental results and the physical analysis show that THz is more suitable with respect to the detection of defects in glass fiber lightweight honeycomb composites. [ABSTRACT FROM AUTHOR]

Self-organizing map (SOM) method combined with the empirical orthogonal function was used to reconstruct the subsurface temperature field by using sea surface data in the Northwestern Pacific Ocean. In contrast to the traditional method, SOM method can extract nonlinear relations from the data and is more suitable for nonlinear dynamics in the ocean. Error statistics show that SOM method provides reconstructions of the subsurface temperature field with the majority of relative errors below 20% at 0–1000-m depth. [ABSTRACT FROM AUTHOR]

Abstract The synoptic monthly gridded (SMG) world ocean database (WOD) were constructed with 1° × 1° horizontal resolution and 28 standard vertical levels from the surface to 3000 m deep using the optimal spectral decomposition technique. Monthly acceleration potential (AP) relative to 1000-dbar is calculated for the isopycnal surface 26.4 kg m−3 from January 1960 to December 2014 to investigate the spatial and temporal variability of the California Current. A composite analysis was conducted to obtain the total-time mean AP field and the climatological monthly mean AP variability, which is two orders of magnitude smaller than the total-time mean. Residual data were used to examine interannual variations of the AP field. An empirical orthogonal function (EOF) analysis was conducted to analyze the AP anomaly data. The EOF-1 (40% variance) represents strengthening/weakening of the North Pacific Gyre and California Current. EOF-2 (20% variance) represents onshore/offshore geostrophic flow. The first principal component was negatively correlated with the PDO; the second principal component was negatively correlated with both the PDO and ENSO, which implied that the AP anomaly contains climate variability signals. [ABSTRACT FROM AUTHOR]

A global empirical orthogonal function (EOF) model of the tropospheric weighted mean temperature called GEOFM_Tm was developed using high-precision Global Geodetic Observing System (GGOS) Atmosphere T m data during the years 2008–2014. Due to the quick convergence of EOF decomposition, it is possible to use the first four EOF series, which consists base functions U k and associated coefficients P k , to represent 99.99% of the overall variance of the original data sets and its spatial-temporal variations. Results show that U 1 displays a prominent latitude distribution profile with positive peaks located at low latitude region. U 2 manifests an asymmetric pattern that positive values occurred over 30° in the Northern Hemisphere, and negative values were observed at other regions. U 3 and U 4 displayed significant anomalies in Tibet and North America, respectively. Annual variation is the major component of the first and second associated coefficients P 1 and P 2 , whereas P 3 and P 4 mainly reflects both annual and semi-annual variation components. Furthermore, the performance of constructed GEOFM_Tm was validated by comparison with GTm_III and GTm_N with different kinds of data including GGOS Atmosphere T m data in 2015 and radiosonde data from Integrated Global Radiosonde Archive (IGRA) in 2014. Generally speaking, GEOFM_Tm can achieve the same accuracy and reliability as GTm_III and GTm_N models in a global scale, even has improved in the Antarctic and Greenland regions. The MAE and RMS of GEOFM_Tm tend to be 2.49 K and 3.14 K with respect to GGOS T m data, respectively; and 3.38 K and 4.23 K with respect to IGRA sounding data, respectively. In addition, those three models have higher precision at low latitude than middle and high latitude regions. The magnitude of T m remains at the range of 220–300 K, presented a high correlation with geographic latitude. In the Northern Hemisphere, there was a significant enhancement at high latitude region reaching 270 K during summer. GEOFM_Tm is capable to represent the spatiotemporal variations of T m , with the high accuracy and reliability in a global scale, therefore, will be of great significance to the real-time GNSS water vapor inversion and climate studies. [ABSTRACT FROM AUTHOR]

The computerized ionospheric tomography is a method for imaging the Earth’s ionosphere using a sounding technique and computing the slant total electron content (STEC) values from data of the global positioning system (GPS). The most common approach for ionospheric tomography is the voxel-based model, in which (1) the ionosphere is divided into voxels, (2) the STEC is then measured along (many) satellite signal paths, and finally (3) an inversion procedure is applied to reconstruct the electron density distribution of the ionosphere. In this study, a computationally efficient approach is introduced, which improves the inversion procedure of step 3. Our proposed method combines the empirical orthogonal function and the spherical Slepian base functions to describe the vertical and horizontal distribution of electron density, respectively. Thus, it can be applied on regional and global case studies. Numerical application is demonstrated using the ground-based GPS data over South America. Our results are validated against ionospheric tomography obtained from the constellation observing system for meteorology, ionosphere, and climate (COSMIC) observations and the global ionosphere map estimated by international centers, as well as by comparison with STEC derived from independent GPS stations. Using the proposed approach, we find that while using 30 GPS measurements in South America, one can achieve comparable accuracy with those from COSMIC data within the reported accuracy (1 × 1011 el/cm3) of the product. Comparisons with real observations of two GPS stations indicate an absolute difference is less than 2 TECU (where 1 total electron content unit, TECU, is 1016 electrons/m2). [ABSTRACT FROM AUTHOR]

The Plum Rains process is a complex process, and its spatiotemporal variations and influencing factors on different time scales still need further study. Based on a dataset on the Plum Rains in the Yangtze River Delta, from 33 meteorological stations during the period of 1960 to 2012, we investigated the spatiotemporal variations of Plum Rains and their relation with the East Asian Summer Monsoon (EASM), the El Niño-Southern Oscillation (ENSO), and the Pacific Decadal Oscillation (PDO) using an integrated approach that combines ensemble empirical mode decomposition (EEMD), empirical orthogonal function (EOF), and correlation analysis. The main conclusions were as follows: (1) the plum rainfall (i.e., the rainfall during the period of Plum Rains) showed a trend of increasing first and then decreasing, and it had a three-year and six-year cycle on the inter-annual scale and a 13-year and 33-year cycle on the inter-decadal scale. The effect of the onset and termination of Plum Rains and the daily intensity of plum rainfall on plum rainfall on the inter-annual scale was greater than the inter-decadal scale, (2) the EOF analysis of plum rainfall revealed a dominant basin-wide in-phase pattern (EOF1) and a north-south out-of-phase pattern (EOF2), and (3) ENSO and EASM were the main influencing factors in the three-year and six-year periods, respectively.

Extensive monitoring of cyanobacterial blooms in lakes and reservoirs can provide important protection for drinking water sources. In most inland waterbodies, phycocyanin (PC) concentrations are the best indicator of cyanobacteria distribution. PC has a characteristic absorption peak near 620 nm; however, reflectance at this wavelength is only available from MEdium Resolution Imaging Spectrometer (MERIS) and Ocean and Land Colour Instrument (OLCI) sensors. MERIS stopped providing data after 2012 and OLCI was only recently launched (February 2016). The Moderate Resolution Imaging Spectroradiometer (MODIS) on Terra and Aqua is currently the only satellite instrument that can provide well-calibrated top-of-atmosphere radiance data over an extended number of years to the present. In this study, we develop and validate a new approach based on empirical orthogonal function (EOF) to quantify PC concentrations in a turbid inland lake (Lake Chaohu, China). Based on Rayleigh-corrected reflectance data ( $R_{{\rm{rc}}}$) at 469, 555, 645, and 859 nm, the concentrations of PC were estimated by regression of 87 concurrent MODIS-field measurements for bloom and nonbloom conditions. The validation (N = 93) showed R2 = 0.40 and unbiased RMS = 60.86%. Application of the algorithm from 2000 and 2014 showed spatial distribution patterns and seasonal changes that confirmed in situ and MERIS-based studies of floating algae mats. The spatial information on PC concentrations in Lake Chaohu had a reduced sensitivity to perturbations from thin aerosols and high sediments. This EOF approach allows us for new insights in the long-term dynamics of shallow lakes and reservoirs where having a better understanding of cyanobacterial blooms is important. [ABSTRACT FROM PUBLISHER]

A new method is proposed to retrieve wave height from an X-band marine radar image sequence, without external measurements for reference. The X-band marine radar image sequence is first decomposed by empirical orthogonal function (EOF), and then the sea surface height profile is reconstructed and scaled from the first EOF mode. The radial profiles that are close to the peak wave direction are used to extract the zero-crossing wave periods and relative wave heights. The spectral width parameter is deduced from the histogram of a dimensionless wave period. Based on a joint probability distribution function (pdf) of a dimensionless wave period and wave height, the theoretical pdf of the wave height is derived. A shape parameter is defined for the theoretical pdf and the histogram of the relative wave heights, and then the calibration coefficient is estimated. The method is validated by comparing the significant wave heights retrieved from two different X-band marine radar systems with those measured by buoy; the correlation coefficient, the root-mean-square error, and the bias between them are 0.78, 0.51 m, and −0.19 m for HH polarization, while they are 0.77, 0.51 m, and 0.19 m for VV polarization, respectively. The sources of error of the method are discussed. [ABSTRACT FROM PUBLISHER]

This letter developes and validates a machine learning approach to forecast sea level anomalies (SLAs) derived from satellite altimetry in the tropical Pacific Ocean. The empirical orthogonal function (EOF), also known as principal component analysis, was used to extract dominant signals and reduce the dimensionality of data sets. Such dimensionality was decreased by describing spatial patterns (EOFs) and the corresponding temporal domains [principal components (PCs)]. Support vector regression (SVR) was employed to predict the time series obtained from the leading PCs. Thereafter, the temporal and spatial SLAs from the proposed EOFs were reconstructed to represent the spatiotemporal SLA prediction. Finally, the prediction result was compared with that of the conventional autoregressive integrated moving average (ARIMA) model. Both models reached satisfactory sea level predictions. Even so, intercomparison of the obtained results showed that the SVR significantly ( $P = 0.012$ ) outperformed the ARIMA model in sea level forecasting. That is, a considerably low root-mean-square error was attained for the differences between the predicted and observed mean SLAs. [ABSTRACT FROM PUBLISHER]