Cyclostationary Empirical Orthogonal Function Research Articles

Regional sea level variability in the Bohai Sea, Yellow Sea, and East China Sea

The regional sea level variability in the Bohai Sea (BS), Yellow Sea (YS) and East China Sea (ECS) is investigated based on tide gauge, satellite altimeter data and an independent oceanic general circulation model for the Earth Simulator (OFES) model outputs. It is found that atmospheric forcing significantly affects local sea level variability in the BS and YS and local sea level variability at the Southern ECS is highly correlated with along-shore currents. Particularly, the annual sea level fluctuations potentially change inundation risk and the frequency and magnitude of flooding in regions with high annual sea level. Hence, the cyclostationary empirical orthogonal function (CSEOF) analysis is carried out to investigate the variations of annual sea level cycle amplitude. Similar spatial distribution characteristics of annual sea level amplitude fluctuations are presented from satellite altimeter data and model outputs. The variability of annual sea level amplitude estimated from the satellite altimeter data agrees well with that from the tide gauge data, and positively (negatively) correlates with Southern Oscillation Index (Pacific Decadal Oscillation). The OFES model, however, underestimates the fluctuation of the annual cycle. After removing the annual signal, the low-passed (i.e., 13-month running mean) tide gauge data shows high correlations with SOI and PDO on time scales over 8 years in the BS and ECS.

Variability of sea surface height in the South China Sea and its relationship to Pacific oscillations

The spatio-temporal variability modes of the sea surface height in the South China Sea (SCS-SSH) are obtained using the Cyclostationary Empirical Orthogonal Function (CSEOF) method, and their relationships to the Pacific basin scale oscillations are examined. The first CSEOF mode of the SCS-SSH is a strongly phase-locked annual cycle that is modulated by a slowly varying principal component (PC); the strength of this annual cycle becomes reduced during El Nino events (at largest by 30% off in 1997/98) and enhanced during La Nina events. The second mode is a low frequency oscillation nearly on decadal time scale, with its spatial structure exhibiting an obscure month-dependence; the corresponding PC is highly correlated with the Pacific Decadal Oscillation (PDO) index. Five independent oscillations in the Pacific are isolated by using the independent component (IC) analysis (ICA) method, and their effects on the SCS-SSH are examined. It is revealed that the pure ENSO mode (which resembles the east Pacific ENSO) has little effect on the low frequency variability of the SCS-SSH while the ENSO reddening mode (which resembles the central Pacific ENSO) has clear effect. As the ENSO reddening mode is an important constituent of the PDO, this explains why the PDO is more important than ENSO in modulating the low frequency variability of SCS-SSH. Meridional saddle like oscillation mode, the Kuroshio extension warming mode, and the equatorial cooling mode are also successfully detected by the ICA, but they have little effect on the low frequency variability of the SCS-SSH. Further analyses suggest the Pacific oscillations are probably influencing the variability of the SCS-SSH in ways that are different from that of the sea surface temperature (SST) in the SCS.

Characteristics of formation and growth of atmospheric nanoparticles observed at four regional background sites in Korea

Measurements of the number concentration and size distribution of atmospheric nanoparticles were conducted at four sites on the west coast of the Korean Peninsula by using identical scanning mobility particle sizers (SMPSs) in October 2012. The new particle formation and subsequent growth (NPF) of atmospheric nanoparticles, which were identified by the cyclostationary empirical orthogonal function (CSEOF) analysis technique, was observed on 11 out of 21days at the Baengnyeong-do Comprehensive Monitoring Observatory (BCMO); and on 10 out of 21days at the Korea Global Atmosphere Watch Center (KGAWC) from October 9 to 29, 2012. We also observed NPF events for 9 out of 21days at both the Gosan Climate Observatory (GCO) and the Jeju Comprehensive Monitoring Observatory (JCMO). During the study period, NPF was simultaneously observed for five days at all four sites, which indicates that the NPF event had a spatial extent of at least 540km.A cold, dry and cloud-free continental air mass originated from northern China, formed favorable environmental conditions (e.g., increasing solar insolation at the surface) on simultaneous NPF at the four sites. These synoptic weather patterns were closely associated with an extraordinary typhoon passing over the south of Japan. The mean values of particle formation rates at BCMO (1.26cm−3s−1) and KGAWC (1.49cm−3s−1) were relatively higher than those at GCO (0.39cm−3s−1) and JCMO (0.74cm−3s−1), however, the growth rate showed a similar level among four sites. An increase in the spatial homogeneity and inter-site correlation of atmospheric particles among the four sites was apparent for small particles (diameter<30nm) on simultaneous NPF event days.

Quantitative Monthly Precipitation Forecasting Using Cyclostationary Empirical Orthogonal Function and Canonical Correlation Analysis

AbstractAn empirical statistical system for quantitative forecasting of monthly precipitation in Korea has been developed using the cyclostationary empirical orthogonal function (CSEOF) and the canonical correlation analysis (CCA) with sea surface temperature (SST) data as the predictor. Monthly Korean precipitation and SST data are comprehensively analyzed using the empirical orthogonal function (EOF) technique and the CSEOF technique, respectively, and the CSEOF technique can exhibit the spatial distribution and temporal evolution characteristics of variability along with recurrent seasons of precipitation in Korea. Through a multivariate regression method, the CCA technique is used to forecast precipitation with different lead times, and the forecasting results indicate that the CSEOF-CCA forecasting model agrees well with the observation data and is particularly useful in forecasting seasonal precipitation variations in Korea.

The effect of the El Niño‐Southern Oscillation on U.S. regional and coastal sea level

AbstractAlthough much of the focus on future sea level rise concerns the long‐term trend associated with anthropogenic warming, on shorter time scales, internal climate variability can contribute significantly to regional sea level. Such sea level variability should be taken into consideration when planning efforts to mitigate the effects of future sea level change. In this study, we quantify the contribution to regional sea level of the El Niño‐Southern Oscillation (ENSO). Through cyclostationary empirical orthogonal function analysis (CSEOF) of the long reconstructed sea level data set and of a set of U.S. tide gauges, two global modes dominated by Pacific Ocean variability are identified and related to ENSO and, by extension, the Pacific Decadal Oscillation. By estimating the combined contribution of these two modes to regional sea level, we find that ENSO can contribute significantly on short time scales, with contributions of up to 20 cm along the west coast of the U.S. The CSEOF decomposition of the long tide gauge records around the U.S. highlights the influence of ENSO on the U.S. east coast. Tandem analyses of both the reconstructed and tide gauge records also examine the utility of the sea level reconstructions for near‐coast studies.

A Southern Hemisphere sea level pressure‐based precursor for ENSO warm and cold events

AbstractPast studies have described large‐scale sea level pressure (SLP) variations in the Southern Hemisphere that lead to El Nino–Southern Oscillation (ENSO) warm and cold events (WE and CE). By relying on this description and the importance of the related variability in the lead up to WE and CE, Southern Hemisphere SLP variations in May‐June‐July (MJJ) are shown here to be excellent predictors for the peak warm/cold events in sea‐surface temperatures (SST) and sea level pressure that mark the mature phase of a warm/cold event in November‐January of the same year. Cyclostationary empirical orthogonal functions (CSEOFs) are used to extract the variability associated with this description of SLP evolution leading to extreme events, underscoring the importance of this signal in the build‐up to ENSO events. Using the CSEOF decomposition, an MJJ precursor is established and shown to precede impending warm and cold events in the past sixty years. Furthermore, the precursor developed in this study would have suggested that a significant WE for the latter half of 2014 was unlikely.

Decadal changes in the Southern Hemisphere sea surface temperature in association with El Niño–Southern Oscillation and Southern Annular Mode

The spatial structure of El Nino–Southern Oscillation (ENSO) and the Southern Annular Mode (SAM), which are the two most important climate modes affecting sea surface temperature (SST) variability in the Southern Hemisphere (SH), appear to have changed since 1999. The characteristic features of the ENSO- and SAM-related atmospheric and oceanic variability in the SH are compared between two sub-periods (1979–1998 and 1999–2012) using cyclostationary empirical orthogonal function analysis. During the earlier period of 1979–1998, the ENSO is characterized by conventional eastern Pacific type, in which the signals in the SH constitute the Pacific South America teleconnection pattern. In contrast, due to a shift of the active center of ENSO to the central Pacific in the later period (1999–2012), atmospheric circulation and SST variability over the SH significantly vary. Moreover, the SAM-related SST variability also shows remarkable differences before and after 1998–1999. This difference is primarily attributed to differences in the non-annular spatial component of the SAM between the two periods. Due to the changes in the spatial structure of the SAM, as well as those of the ENSO, SST variability in the SH displays a marked change between the two periods. Detailed descriptions of the decadal changes of the SH SST in terms of interaction in the oceanic and atmospheric variability are presented along with the possible implications of this change.

Effects of climate oscillations on wind resource variability in the United States

Natural climate variations in the United States wind resource are assessed by using cyclostationary empirical orthogonal functions (CSEOFs) to decompose wind reanalysis data. Compared to approaches that average climate signals or assume stationarity of the wind resource on interannual time scales, the CSEOF analysis isolates variability associated with specific climate oscillations, as well as their modulation from year to year. Contributions to wind speed variability from the modulated annual cycle (MAC) and the El Nino-Southern Oscillation (ENSO) are quantified, and information provided by the CSEOF analysis further allows the spatial variability of these effects to be determined. The impacts of the MAC and ENSO on the wind resource are calculated at existing wind turbine locations in the United States, revealing variations in the wind speed of up to 30% at individual sites. The results presented here have important implications for predictions of wind plant power output and siting.

Characteristics of submicron aerosol number size distribution and new particle formation events measured in Seoul, Korea, during 2004–2012

The diurnal and seasonal characteristics of the submicron aerosol size distribution measured in Seoul for a long term (2004–2012) were analyzed. The average diurnal variation of aerosol number size distribution showed a single modal distribution but the modal diameter varied around 35 nm and the total aerosol concentration varied diurnally with the maximum during the rush hour. To identify the dominant patterns of the diurnal variation of aerosol number size distribution, the cyclostationary empirical orthogonal function (CSEOF) method was employed. The first CSEOF mode revealed the dominant diurnal pattern that represented an urban life pattern. The principal component (PC) values of the first CSEOF mode showed clear seasonal variation. The PC values were the largest in winter months due to winter monsoon that brought polluted air masses from China, and larger traffic emission amount during the winter season. New particle formation (NPF) events were found to occur on 12.3% of the total measurement days in Seoul when classified with a subjective method. NPF frequency varied seasonally, with the maximum in spring. The second CSEOF mode was found to represent NPF events. A way to identify NPF event days by the CSEOF method was proposed and generally performed well but with some exceptions. The favorable conditions for NPF events in Seoul were found to be low RH, low condensation sink, and low cloud cover, which is consistent with some previous studies. Lastly, the relative contribution of NPF to ultrafine aerosol in Seoul was estimated to be ~32%.

A comparative study of sea level reconstruction techniques using 20 years of satellite altimetry data

AbstractSea level reconstructions extend spatially dense data sets, such as those from satellite altimetry, by decomposing the data set into basis functions and fitting those functions to in situ tide gauge measurements with a longer temporal record. We compare and evaluate two methods for reconstructing sea level through an idealized study. The compared sea level reconstruction methods differ in the technique for calculating basis functions, i.e., empirical orthogonal functions (EOFs) versus cyclostationary EOFs (CSEOFs). Reconstructions are created using Archiving, Validation, and Interpretation of Satellite Oceanographic (AVISO) satellite altimetry data and synthetic tide gauges. Synthetic tide gauge records are simulated using historical distributions and real high‐frequency signal to test reconstruction skill. The CSEOF reconstructions show high skill in reproducing variations in global mean sea level (GMSL) and ocean climate indices, and are affected less by both limited tide gauge distribution and added high‐frequency tide gauge signal than EOF reconstructions. Typically, CSEOF reconstructions slightly underestimate sea level amplitudes while EOF reconstructions overestimate sea level amplitudes, in some cases, significantly. Both of these results are accentuated with decreasing quality of the synthetic tide gauge data set. Additionally, we investigate how the reconstructions differ when reconstructing with more of the variance retained in the basis functions. Increasing the variance explained by the basis functions from 70% to 90% reduces the efficacy of an EOF reconstruction to reproduce common ocean indices when noise is included in the tide gauge data sets. These results show that in the idealized comparative cases examined the CSEOF method of sea level reconstruction creates more robust reconstructions, especially when less than ideal tide gauge data are used.

Climate change projection in the Northwest Pacific marginal seas through dynamic downscaling

AbstractThis study presents future climate change projections in the Northwest Pacific (NWP) marginal seas using dynamic downscaling from global climate models (GCMs). A regional climate model (RCM) for the Northwest Pacific Ocean was setup and integrated over the period from 2001 to 2100. The model used forcing fields from three different GCM simulations to downscale the effect of global climate change. MIROC, ECHAM, and HADCM were selected to provide climate change signals for the RCM. These signals were calculated from the GCMs using Cyclostationary Empirical Orthogonal Function analysis and added to the present lateral open boundary and the surface forcing. The RCM was validated by comparing hindcast result with the observation. It was able to project detailed regional climate change processes that GCMs were not able to resolve. A relatively large increases of water temperature were found in the marginal seas. However, only a marginal change was found along the Kuroshio path. Heat supply to the atmosphere decreases in most study areas due to a slower warming of the sea surface compared to the atmosphere. The RCM projection suggests that the temperature of the Yellow Sea Bottom Cold Water will gradually increase by 2100. Volume transports through major straits except the Taiwan Strait in the marginal seas are projected to increase slightly in future. Increased northeasterly wind stress in the East China Sea may also result in the transport change.

1979~2011년, 북극진동지수 측면에서의 겨울철 남한지역 신적설과 최저 온도 특성

A characteristic of snowfall and minimum temperature variability in South Korea with respect to the variability of Arctic Oscillation (AO) was investigated. The climatic snowfall regions of South Korea based on daily new fresh snowfall data of 59 Korea Meteorological Administration (KMA) stations data corresponding to the sign of AO index during December to February 1979~2011 were classified. Especially, the differences between snowfalls of eastern regions and that of western regions in South Korea were seen by each mean 1000hPa geopotential height fields, which is one of physical structure, for the selected cases over the East Asia including the Korean Peninsula. Daily minimum temperature variability of 59 KMA station data and daily AO index during the same period were investigated using Cyclo-stationary empirical orthogonal function (CSEOF) analysis. The first CSEOF of wintertime daily AO index and that of minimum temperature of 59 KMA stations explain 33% and 66% of total variability, respectively. Correlation between principal component time series corresponding to the first CSEOF of AO index and that of temperature at the period of 1990s is over about -0.7 when that of AO index leads about 40 days.

Cyclostationary empirical orthogonal function sea‐level reconstruction

Since 1993, satellite altimetry has provided accurate measurements of sea surface height with near‐global coverage. These measurements led to the first definitive estimates of global mean sea‐level (GMSL) rise and have improved understanding of how sea levels are changing regionally at decadal time scales. These relatively short records, however, provide no information about the state of the ocean prior to 1993, and with the modern altimetry record spanning only 20 years, the lower frequency signals that are known to be present in the ocean are difficult or impossible to resolve. Tide gauges, on the other hand, have measured sea level over the last 200 years, with some records extending back to 1807. While providing longer records, the spatial resolution of tide gauge sampling is poor, making studies of the large‐scale patterns of ocean variability and estimates of GMSL difficult. Combining the satellite altimetry with the tide gauges using a technique known as sea‐level reconstruction results in a dataset with the record length of the tide gauges and the near‐global coverage of satellite altimetry. Cyclostationary empirical orthogonal functions (CSEOFs), derived from satellite altimetry, are combined with historical sea‐level measurements from tide gauges to create the Reconstructed Sea Level dataset spanning from 1950 to 2009. Previous sea‐level reconstructions have utilized empirical orthogonal functions (EOFs) as basis functions, but by using CSEOFs and by addressing other aspects of the reconstruction procedure, an alternative sea‐level reconstruction can be computed. The resulting reconstructed sea‐level dataset has weekly temporal resolution and half‐degree spatial resolution.

Recent climate variation in the Bering and Chukchi Seas and its linkages to large-scale circulation in the Pacific

The thermal state of the Bering Sea exhibits interdecadal variations, with distinct changes occurred in 1997–1998. After the unusual thermal condition of the Bering Sea in 1997–1998, we found that the recent climate variability (1999–2010) in the Bering Sea is closely related to Pacific basin-scale atmospheric and oceanic circulation patterns. Specifically, warming in the Bering and Chukchi Seas in this period involves sea ice reduction and stronger oceanic heat flux to the atmosphere in winter. The atmospheric response to the recent warming in the Bering and Chukchi Seas resembles the North Pacific Oscillation (NPO) pattern. Further analysis reveals that the recent climate variability in the Bering and Chukchi Seas has strong covariability with large-scale climate modes in the Pacific, that is, the North Pacific Gyre Oscillation and the central Pacific El Niño. In this study, physical connections among the recent climate variations in the Bering and Chukchi Seas, the NPO pattern and the Pacific large-scale climate patterns are investigated via cyclostationary empirical orthogonal function analysis. An additional model experiment using the National Center for Atmospheric Research Community Atmospheric Model, version 3, is conducted to support the robustness of the results.

Statistical multisite simulations of summertime precipitation over South Korea and its future change based on observational data

In this study, a weather generator for summer (May 19 – September 15) precipitation over South Korea is developed. Precipitation data for 33 years (1979–2011) observed at 57 stations of Korea Meteorological Administration (KMA) are used to develop a new weather generator. Using the cyclostationary empirical orthogonal function (CSEOF) technique, the observed precipitation data is described as a linear combination of deterministic evolution patterns and corresponding stochastic amplitude (principal component) time series. An autoregressive-moving average (ARMA) model is used to generate one hundred sets of synthetic amplitude time series for the period of 1979–2061 (83 years) with similar statistical properties of the original amplitude time series. Based on these synthetic time series and the annually repeating evolution patterns, one hundred sets of synthetic summer precipitation were generated. Statistical characteristics of the synthetic datasets are examined in comparison with those of the KMA observational record for the period of the observational record. Characteristic changes of synthetic precipitations for a future period are also examined. The seasonal cycle in the synthetic precipitation is reproduced faithfully with typical bimodal peaks of summer precipitation. The spatial correlation patterns of the synthetic precipitation are fairly similar to that of the observational data. The frequency-intensity relationship of the synthetic precipitation also looks similar to that of the observational data. In the future period, precipitation amount increases except in the precipitation range of (0,10) mm day−1 with nearly no change in the frequency of no-rain days; frequency increase is particularly conspicuous in the range of (100,500) mm day−1.

Global warming, low-frequency variability, and biennial oscillation: an attempt to understand the physical mechanisms driving major ENSO events

Three primary global modes of sea surface temperature (SST) variability during the period of 1871–2010 are identified through cyclostationary empirical orthogonal function analysis. The first mode exhibits a clear trend and represents global SST warming with an ‘El Niño-like’ SST pattern in the tropical Pacific. The second mode is characterized by considerable low-frequency variability in both the tropical Pacific and the North Pacific regions, indicating that there is a close connection between the two regions on interannual and decadal time scales. The third mode shows a seesaw pattern between El Niño and La Niña within a two-year period; this mode is derived by the oscillatory tendency of the tropical Pacific ocean–atmosphere coupled system. A SST reconstruction based on these three modes captures a significant portion of the SST variability in the raw data, which is primarily associated with El Niño-Southern Oscillation (ENSO) events in the tropical Pacific. Additionally, this study attempts to interpret the major ENSO events that have occurred since the 1970s in terms of the interplay originating from these three modes of variability. In particular, two key points are derived from this analysis: (1) the most extreme El Niño events occurred in 1982/1983 and 1997/1998 are attributed to the positive contributions of all three modes; and (2) the central Pacific (CP) El Niño events in the 1990s and 2000s have different physical mechanisms, that is, the CP El Niño events in the early 1990s originated mainly from the low-frequency mode, while those in the early 2000s derived mainly from the global warming mode.

Energy budget change in the tropics according to the SRES A1B scenario in the IPCC AR4 models

The IPCC AR4 model data sets were investigated in order to address how the energy balance will change in the tropics (30°S–30°N) under the SRES A1B scenario. A climate change signal with a well‐defined trend was extracted using cyclostationary empirical orthogonal function analysis; this signal depicts near‐linear warming in the tropics. This “warming” signal, together with the seasonal cycle, explains most of the variance (91.32–94.86%) in the AR4 model datasets. In addition to the warming of the atmospheric column and the surface, the cloud fraction decreased over most of the tropics, except over the equatorial Pacific. Specific humidity generally increased over the entire troposphere. The decreased cloud fraction and the increased specific humidity resulted in a net increase in shortwave radiation (by ~3.58 (±1.92) W m–2) in the atmospheric column. Simultaneously, the increased atmospheric and surface temperatures (resulting from positive water vapor feedback) caused enhanced longwave radiation exchange between the surface and the atmospheric column; net downward longwave radiation increased by ~19.22 (±3.85) W m–2, and net upward longwave radiation increased by ~14.50 (±3.04) W m–2 over 100 years. The second largest change was found in the heat flux leaving the surface, which amounted to ~4.55 (±1.72) W m–2. As a result of the radiation budget change associated with warming and meridional energy transport by the atmosphere and the ocean, net energy gains were found for both the tropical atmosphere (~5.82 (±3.23) W m–2) and the tropical surface (~0.48 (±0.30) W m–2).

Seasonal scale variability of the East Asian winter monsoon and the development of a two-dimensional monsoon index

This study investigates the seasonal scale variability of the East Asian winter monsoon (EAWM), which is distinguished from the seasonal cycle with temporal variation throughout winter. Winters lasting 120 days (Nov. 17–Mar. 16) for a period of 64 years from the NCEP daily reanalysis data set are used to study the seasonal scale variability of the EAWM. Cyclostationary empirical orthogonal function (CSEOF) analysis is adopted to decompose the variability of the EAWM. The second CSEOF mode of 850-hPa temperature exhibits a seasonal scale variation, the physical mechanism of which is explained in terms of physically consistent variations of temperature, geopotential height, sea level pressure, wind, and surface heat fluxes. The seasonal-scale EAWM exhibits a weak subseasonal and a strong interannual variability and has gradually weakened during the 64 years. In a weak EAWM phase, the land-sea contrast of sea level pressure declines in East Asia. Consistent with this change, low-level winds decrease and warm thermal advection increases over the eastern part of mid-latitude East Asia. Latent and sensible heat fluxes are reduced significantly over the marginal seas in East Asia. However, during a strong EAWM phase, the physical conditions in East Asia reverse. A large fraction of the variability of the EAWM is explained by the seasonal cycle and the seasonal scale variation. A two-dimensional EAWM index was developed to explain these two distinct components of the EAWM variability. The new index appears to be suitable for measuring both the subseasonal and the interannual variability of the EAWM.

Regional Sea Level Reconstruction in the Pacific Ocean

As a result of the spatial and temporal sampling shortcomings of tide gauge and satellite altimetry data, forming a sea level record that has the quality and duration necessary to compare past to present sea level is a challenge. Here, we build on the recently developed method for reconstructing sea level involving cyclostationary empirical orthogonal functions (CSEOFs) and, in particular, examine how the reconstruction varies as a result of the domain used to compute the CSEOF basis functions. While not affecting the ability to capture the dominant climate signals, the choice of basis function is found to have a significant effect on the regional distribution of the trend in sea level. We also demonstrate the use of a CSEOF reconstructed sea level dataset for climate monitoring, focusing on climate signals in the Pacific Ocean.

Near 13 day barotropic ocean response to the atmospheric forcing in the North Pacific

In the Kuroshio Extension System Study (KESS) east of Japan, bottom pressure observations over the 2 year study period exhibit strong high‐frequency variability near 13 days. The first cyclostationary empirical orthogonal function mode for the band‐pass‐filtered KESS bottom pressure explains about 57% of the near 13 day variance and exhibits almost in‐phase variability in space with a hint of westward propagation. The 13 day variability is strong during the winter and is driven by the large‐scale wind stress curl over a broad region of the North Pacific. Modeling results over the North Pacific closely follow the observations and indicate that topography confines the barotropic response to the west of Emperor Seamount Chain and slows the westward propagation of the near 13 day bottom pressure variability.