Principal Component Time Series Research Articles

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.

Interannual variability of winter precipitation over northwest India and adjoining region: impact of global forcings

The role of El Nino/Southern Oscillation (ENSO) and the mechanism through which ENSO influences the precipitation variability over northwest India and the adjoining (NWIA) region is well documented. In this study, the relative role of North Atlantic Oscillation (NAO)/Arctic Oscillation (AO) and ENSO in modulating the Asian jet stream in the Northern Hemisphere winter and their relative impact on the precipitation variability over the region have been estimated through analysis of observed data. It is seen that interannual variations of NWIA precipitation are largely influenced by ENSO. An empirical orthogonal function (EOF) analysis has been carried out to understand dominant modes of interannual variability of zonal wind at 200 hPa of the Northern Hemisphere. The EOF-1 pattern in the tropical region is similar to that of an ENSO pattern, and the principal component (PC) time series corresponds to the ENSO time series. The EOF-2 spatial pattern resembles that of NAO/AO with correlation of PC time series with AO and NAO being 0.74 and 0.62, respectively. The precipitation anomaly time series over the region of interest has marginally higher correlation with the PC-2 time series as compared to that of PC-1. Regression analysis of precipitation and circulation parameters indicates a larger contribution of the second mode to variability of winds and precipitation over the NWIA. Moisture transport from the Arabian Sea during the active phase of NAO/AO and the presence of a cyclonic anomaly lead to higher precipitation over the NWIA region.

Physical‐biological synchrony in the global ocean associated with recent variability in the central and western equatorial Pacific

Synchrony in the second modes of an empirical orthogonal function (EOF) analysis of global physical and biological properties is described for the 1993–2010 time period. High correlations are found between the El Niño Modoki index and principal component time series of sea surface temperature, sea surface salinity, photosynthetically active radiation, precipitation, surface currents, surface chlorophyll concentration, and equatorial temperature profiles. Spatial patterns indicate that the second mode is also associated with the North Pacific Gyre Oscillation (NPGO). Biological changes during traditional El Niños (first EOF mode) have been explained on the basis of strong and coherent variations in thermocline depth, wind‐driven upwelling and light, but changes associated with the second mode are more subtle and complex. Equatorial temperature profiles indicate that the warming is confined to the mixed layer and that changes in thermocline depth are small. The biological changes associated with the second mode may be driven by a combination of weak perturbations to vertical nutrient supply and the strength of subtropical gyres. In the western tropical Pacific, the site of some of the strongest perturbations, the biological changes can be associated with the occurrence and thickness of barrier layers and to island effects downstream of the Kiribati Islands. Globally integrated impacts of the second mode are much weaker than those associated with the traditional ENSO. During the positive phase, chlorophyll is strongly enhanced in the tropics and weakly enhanced at the global scale in sharp contrast to traditional El Niño effects. The analysis improves our understanding of global‐scale physical and biological coupling associated with the so‐called El Niño Modoki and the NPGO.

Responses of an African wading bird community to resource pulses are related to foraging guild and food‐web position

Summary1. The immediate impacts of nutrient inputs on aquatic systems are well documented, but the ways in which resource pulses affect the wider food web of water‐associated vertebrates remain obscure.2. We monitored the wading bird community of Barberspan, a natural freshwater lake and Ramsar wetland in South Africa, before, during and after the addition of a pulse of nutrients in the form of a sewage overspill from an upstream processing facility. We counted waders at 13 points around the lake over 3 years, every 2 months from March 2007 to March 2010, and sampled water quality during all counting periods from January 2008 to January 2010.3. We used our data to test the hypothesis that wading birds that forage directly at lower trophic levels and/or on prey populations that have fast turnover rates, such as those of phytoplankton and invertebrates, will be more heavily influenced by nutrient addition than birds that forage on species with lower turnover rates and/or at higher trophic levels (such as frogs and fish).4. During the sampling period Barberspan experienced a significant, nutrient‐driven decline and subsequent recovery in dissolved oxygen and pH. This trend was mirrored by significant changes in the wading bird community. Partial Mantel tests and Canonical Correspondence Analysis (CCA) showed that the nutrient pulse had marked short‐term, negative impacts on both the diversity and the abundance of medium‐sized, shoreline foragers such as scolopacids (e.g. sandpipers).5. Our analysis supports the proposal that both food‐web position and the turnover rate of the prey population are strong influences on ecological responses to resource pulses. Analysis of time series of principal components that describe community composition suggested that recovery of the prey base was rapid and that the bird community was able to respond via immigration. These results must, of course, be considered provisional in the absence of replicated experimental data.6. More generally, we interpret our results as suggesting that two different mechanisms act in different directions to determine the sensitivity of secondary and tertiary consumers to changes in their prey. First, for ‘earlier’ consumers (i.e. that forage lower in the food web), there are (on average) fewer generalist consumers and fewer stored nutrients in the intervening trophic levels. This increases the sensitivity of earlier consumer populations to changes in the composition of the primary consumer community. Second, the dynamics of prey populations lower in the food web tend to be faster, making recovery faster and serving to decrease the sensitivity of earlier consumers to perturbations. These dynamics may obscure the impacts of nutrient pulses in cases where additional analysis of system trajectories is not undertaken.

Bay of Bengal wave forecast based on genetic algorithm: A comparison of univariate and multivariate approaches

Prediction of significant wave height (SWH) field is carried out in the Bay of Bengal (BOB) using a combination of empirical orthogonal function (EOF) analysis and genetic algorithm (GA). EOF analysis is performed on 4years (2005–2008) of numerical wave model generated SWH field, and analyzed fields of zonal (U) and meridional (V) winds. This is to decompose the space-time distributed data into spatial modes ranked by their temporal variances. Two different variants of GA are tested. In the first one, univariate GA is applied to the time series of the first principal component (PC) of SWH in the training dataset after a filtering with singular spectrum analysis (SSA) for effecting noise reduction. The generated equations are used to carry out forecast of SWH field with various lead times. In the second method, multivariate GA is applied to the SSA filtered time series of the first PC of SWH, and time- lagged first PCs of U and V and again forecast equations are generated. Once again the forecast of SWH is carried out with same lead times. The quality of forecast is evaluated in terms of root mean square error of forecast. The results are also compared with buoy data at a location. It is concluded that the method can serve as a cost-effective alternate prediction technique in the BOB.

Complexity analysis of spatial–temporal precipitation system by PCA and SDLE

In this paper, an ensemble technique combining the principal component analysis (PCA) with scale-dependent Lyapunov exponent (SDLE) is used to characterize complexity of precipitation dynamical system. The spatial–temporal precipitation data is decomposed by employing PCA method and then the SDLE for the first few principal components (PCs) time series are computed. The first few PCs time series are found to exhibit the different scaling laws on different time scales. The study illustrate that the spatial–temporal precipitation data is chaotic and the precipitation system is truly multiscaled and complex.

Search for high-amplitude δ Scuti and RR Lyrae stars in Sloan Digital Sky Survey Stripe 82 using principal component analysis

We propose a robust principal component analysis framework for the exploitation of multiband photometric measurements in large surveys. Period search results are improved using the time-series of the first principal component due to its optimized signal-to-noise ratio. The presence of correlated excess variations in the multivariate time-series enables the detection of weaker variability. Furthermore, the direction of the largest variance differs for certain types of variable stars. This can be used as an efficient attribute for classification. The application of the method to a subsample of Sloan Digital Sky Survey Stripe 82 data yielded 132 high-amplitude δ Scuti variables. We also found 129 new RR Lyrae variables, complementary to the catalogue of Sesar et al., extending the halo area mapped by Stripe 82 RR Lyrae stars towards the Galactic bulge. The sample also comprises 25 multiperiodic or Blazhko RR Lyrae stars.

A Deterministic Algorithm for Robust Location and Scatter

Most algorithms for highly robust estimators of multivariate location and scatter start by drawing a large number of random subsets. For instance, the FASTMCD algorithm of Rousseeuw and Van Driessen starts in this way, and then takes so-called concentration steps to obtain a more accurate approximation to the MCD. The FASTMCD algorithm is affine equivariant but not permutation invariant. In this article, we present a deterministic algorithm, denoted as DetMCD, which does not use random subsets and is even faster. It computes a small number of deterministic initial estimators, followed by concentration steps. DetMCD is permutation invariant and very close to affine equivariant. We compare it to FASTMCD and to the OGK estimator of Maronna and Zamar. We also illustrate it on real and simulated datasets, with applications involving principal component analysis, classification, and time series analysis. Supplemental material (Matlab code of the DetMCD algorithm and the datasets) is available online.

Intraseasonal atmospheric variability and its interannual modulation in Central Africa

The spatial and temporal structures of the intraseasonal atmospheric variability over central Africa is investigated using 2.5° × 2.5° daily outgoing longwave radiation (OLR) and National Centers for Environmental Prediction (NCEP) Reanalysis zonal winds for the period 1980–2010. The study begins with an overview of the Central African rainfall regime, noting in particular the contrast amongst Western and Eastern parts, with different topography and surface conditions features. The annual mean rainfall and OLR over the region revealed a zone of intense convective activity centered on the equator near 30°E, which extends southward and covers almost all the Congo forest. The annual cycle of rainfall reflects the classical bi-annual shift of Inter-Tropical Convergence Zone across the equatorial belt, between 10°S and 10°N. The result of the empirical orthogonal functions (EOFs) analysis has shown that the three leading EOF modes explain about 45 % of total intraseasonal variability. The power spectra of all the three corresponding principal components (PCs) peak around 45–50 days, indicating a Madden–Julian Oscillation (MJO) signal. The first mode exhibits high positive loadings over Northern Congo, the second over Southern Ethiopia and the third over Southwestern Tanzania. The PCs time series revealed less interannual modulation of intraseasonal oscillations for the Congo mode, while Ethiopian and Tanzanian modes exhibit strong interannual variations. Hovmoller plots of OLR, 200 and 850 hPa NCEP zonal winds found the eastward propagating features to be the dominant pattern in all the three times series, but this propagation is less pronounced in the OLR than in the 850 and 200 hpa zonal wind anomalies. An index of MJO strength was built by averaging the 30–50 day power for each day. A plot of MJO indices and El Nino Southern Oscillation (ENSO) cycle confirm a strong interannual modulation of MJO over Eastern central Africa partially linked with the ENSO events (El Nino and La Nina). Strong MJO activity is observed during La Nina years or during ENSO-neutral years, while weak or absent MJO activity is typically associated with strong El Nino episodes.

Statistical evidence for the natural variation of the central Pacific El Niño

Extensive studies claimed that the central equatorial Pacific (CP) El Niño has occurred more frequently and strongly than the eastern equatorial Pacific El Niño in recent years. To explain this phenomenon, spatial patterns and principal component time series from several sea surface temperature (SST) data sets in the tropical Pacific are analyzed for the period of 1951–2010. Cyclostationary empirical orthogonal function analysis separates two modes of SST variability, which explain about 50% and 10% of the total SST variability, respectively. Their spatial and temporal patterns are similar among the different SST data sets. The first mode captures the typical El Niño pattern, while the second mode is a dipole pattern in the tropical Pacific. The two modes are, by definition, uncorrelated over the analysis period but are in phase since the late 1990s; superposition of the two modes results in a significant warming in the CP region, which is a potential explanation for a more frequent occurrence of the CP El Niño in the recent decades. Similar analysis is conducted based on the 500 year data from the Geophysical Fluid Dynamics Laboratory Climate Model version 2.1 under the preindustrial condition. The result is generally consistent with the observations yielding occasional in‐phase relationship between the two modes. Thus, it cannot be ruled out that a more frequent occurrence of the CP El Niño in recent years is a natural feature of the equatorial climate system.

Simplified metrics for the identification of the Madden–Julian oscillation in models

AbstractWe propose simplified metrics to evaluate the fidelity with which the Madden–Julian oscillation (MJO) is simulated in climate models. These metrics are based on lag correlation analysis of principal component time series (PCs). The PCs are obtained by projecting simulated 20–100 day bandpass filtered daily outgoing longwave radiation onto the two leading empirical orthogonal functions of observed MJO variability. The simplified MJO metrics, the maximum positive correlation and time lag at which it occurs, provide consistent information relative to more complex diagnostics developed by the Madden–Julian Oscillation Working Group (CLIVAR MJOWG) and by Kim et al. Copyright © 2012 Royal Meteorological Society

Forecasting tidal currents from tidal levels using genetic algorithm

Prediction of tidal current in the coastal region is an important activity in marine science. It is useful in taking operation- and planning-related decisions such as towing of vessels and monitoring of oil slick movements. It is also useful for fisheries and recreational activities. General practice is to carry out this prediction using harmonic analysis or numerical hydrodynamic models. However, both these methods have their own limitations and nonlinear data adaptive approaches are gaining increasing acceptance. In this paper, such an approach, known as genetic algorithm (GA), has been employed for this prediction. A preliminary empirical orthogonal function (EOF) analysis has been used to compress the spatial variability into a few eigenmodes, so that GA could be applied to the time series of the dominant principal components (PC). The multivariate version of GA has been used to carry out the forecast using a few tidal levels at the boundary of the domain of study as inputs. The performance of this combined technique has been found to be quite satisfactory.

The Relationship between the North Atlantic Jet and Tropical Convection over the Indian and Western Pacific Oceans

Abstract Boreal winter jet variability over the North Atlantic is investigated using 40-yr European Centre for Medium-Range Weather Forecasts Re-Analysis (ERA-40) data, where the variability is defined by the first EOF of the zonal wind on seven vertical levels. The principal component time series of this EOF is referred to as the jet index. A pattern correlation analysis indicates that the jet index more accurately describes intraseasonal North Atlantic zonal wind variability than does the North Atlantic Oscillation (NAO). A series of composite calculations of the jet index based on events of intraseasonal convective precipitation over the tropical Indian and western Pacific Oceans reveals the following statistically significant relationships: 1) negative jet events lead enhanced Indian Ocean precipitation, 2) positive jet events lag enhanced Indian Ocean precipitation, 3) positive jet events lead enhanced western Pacific Ocean precipitation, and 4) negative jet events lag enhanced western Pacific Ocean precipitation. These intraseasonal relationships are found to be linked through the circumglobal teleconnection pattern (CTP). Implications of the sign of the CTP being opposite to that of the jet index suggest that relationships 1 and 3 may arise from cold air surges associated with the CTP over these oceans. On interdecadal time scales, a much greater increase in the frequency of precipitation events from 1958 to 1979 (P1) to 1980 to 2001 (P2) was found for the Indian Ocean relative to the western Pacific Ocean. This observation, combined with relationships 2 and 4, leads to the suggestion that this change in the frequency of intraseasonal Indian Ocean precipitation events may make an important contribution to the excitation of interdecadal variability of the North Atlantic jet.

Global Modes of Sea Surface Temperature Variability in Relation to Regional Climate Indices

Abstract A century-long EOF analysis of global sea surface temperature (SST) was carried out and the first six modes, independent by construction, were found to be associated with well-known regional climate phenomena: the El Niño–Southern Oscillation (ENSO), the Atlantic multidecadal oscillation (AMO), the Pacific decadal oscillation (PDO), the North Pacific Gyre Oscillation (NPGO), El Niño Modoki, and the Atlantic El Niño. Four of the six global modes are dominated by Pacific changes, the other two (M2 and M6) being associated with the AMO and Atlantic El Niño, respectively. The principal component time series of the ENSO (M1) and North Pacific (M3) modes are coherent at time scales >10 yr, and their interaction results in the traditional PDO pattern and the dominant mode of Pacific multidecadal variability. The M3 and PDO time series are well correlated, but the EOFs have different spatial patterns. The fourth mode (M4) has been strengthening since the 1950s and is related to the NPGO but also to El Niño Modoki, especially at the decadal scale. The fifth global mode (M5) is also spatially and temporally correlated to El Niño Modoki. The Pacific SST modes are further related to atmospheric forcing and the circulation of the North Pacific subpolar and subtropical gyres.

An assessment of monsoon precipitation changes during 1901–2001

Changes of global land monsoon precipitation are assessed by using three sets of rain-gauge precipitation data for the period of 1901–2002 compiled by GPCC, CRU and Dai-dataset, respectively. The three datasets show consistent long-term changes of precipitation over the monsoon region with slightly different amplitudes. During 1901–2001, global land monsoon precipitation (GMI) exhibits multi-decadal variations, with an overall increasing trend from 1901 to 1955, followed by a decreasing trend up to 2001. The upward trends of global and Northern hemispheric land monsoon precipitation during 1901–1955 are mainly resulted from the increased precipitation over the North African, Indian and East Asian monsoon domains. For the whole period of 1901–2001, precipitation averaged over the Northern Hemisphere and global land monsoon areas both exhibit a decreasing trend although it is only statistically significant at the 5% level for the Northern Hemisphere. The robust decreasing trend of Northern hemispheric land monsoon precipitations during the twentieth century mainly comes from the downward trend of North African and eastern part of Indian monsoon precipitation and occurs mainly after the 1950s. The first leading mode of Empirical orthogonal function (EOF) analysis of precipitation annual range features a coherent change of North African, South Asian, Northeast China, southern South African, eastern Australian and western American monsoon, and a coherent change over the equatorial South African monsoon and eastern American monsoon. The corresponding principal component time series also indicate that the majority of global land monsoon precipitation has experienced an increasing tendency from 1901 to 1955 and a decreasing trend since the 1950s. Examination on the impact of station number change indicates a negligible influence on the results, especially after 1905.

Scale Interaction of the Diurnal Cycle of Rainfall over the Maritime Continent and Australia: Influence of the MJO

Abstract The influence of the MJO on the phase and amplitude of the diurnal cycle of rainfall during Australian summer [December–February (DJF)] over the Maritime Continent (MC) and northern Australia is investigated using the Tropical Rainfall Measuring Mission (TRMM) 3B42 and 3G68 datasets. The gridded rainfall was partitioned into MJO categories (active, suppressed, and weak) based on their longitudinal position and by utilizing the real-time multivariate MJO (RMM) index of Wheeler and Hendon. The diurnal cycles were composited and an empirical orthogonal function (EOF) analysis was applied to extract the spatial and temporal variability. Distinct variations in the rainfall distribution pattern among categories of the MJO over land and ocean are seen. The result of the composite-mean rainfall distribution shows that the average daily rainfall rate over islands is higher during suppressed MJO days, while for surrounding oceans and northern regions of Australia, more rainfall occurs during MJO active days. The normalized relative amplitude (NRA) of the diurnal cycle of rainfall shows that morning rainfall near coastal areas during active days of the MJO is 1.5 times greater than the climatological-mean rainfall but is less than or equal to the climatological mean during other phases of the MJO. Similarly, during the suppressed phase of the MJO evening rainfall is greater over the islands than in other MJO phases. The first two modes of the EOF alone explain more than 88% (65%) of the variance for the 3B42 (3G68) rainfall, and the corresponding principal component time series show a marked diurnal cycle. The results show that both the amplitude and phase of the diurnal cycle of rainfall are modulated by the categories of the MJO. In general, the peak in the diurnal cycle for active (suppressed/weak) days of the MJO lags (leads) the peak in the diurnal cycle for total rainfall by 2 h. Over Darwin and its adjacent regions, the active phase of the MJO is responsible for the occurrence of maximum rainfall after midnight, which is unusual in this region.

Spatiotemporal structures of the intraseasonal oscillations of precipitation over northern Eurasia during summer

AbstractThis study identifies intraseasonal oscillations in summer precipitation over northern Eurasia using multi‐channel singular spectrum analysis (MSSA) and a gridded daily precipitation data set for northern Eurasia. The analysis period is June–August of 1979–2002. Empirical orthogonal function analysis was first performed on 8‐day low‐pass‐filtered precipitation anomalies at 2.5° grid resolution for northern Eurasia; MSSA was then carried out on the ten leading principal components (PCs). Three quasi‐periodic oscillations with timescales of 45, 15, and 9 days were identified. The spatiotemporal structures of precipitation oscillations were defined by composite analysis based on the reconstructed time series of the spatiotemporal PCs obtained from the MSSA. The composite life cycle of each mode was classified into eight phase categories. The 45‐day oscillation is characterised by a slow eastward progression of a broad east/west contrastive pattern which is associated with a replacement of elongated dry and wet zones across northern Eurasia. The 15‐day oscillation shows a regular eastward phase propagation of precipitation anomalies with a tripole structure across the domain. The eastward displacement of dry and wet zones occurs in association with this oscillation. Spatiotemporal behaviours of the precipitation anomalies associated with the 9‐day oscillation are similar to those of the 15‐day oscillation, but the spatial scale of this oscillation is somewhat smaller. To explore the connections between the oscillations of precipitation and large‐scale atmospheric circulation patterns, a similar composite method was applied to 300‐hPa geopotential height anomalies over the Northern Hemisphere. The circulation patterns responsible for the 45‐day oscillation are characterised by a wave train extending across the northeastern Atlantic/northern Eurasian sector. The wave train, with zonal wavenumbers of 3–4, indicates slow eastward moving and quasi‐stationary features. The circulation patterns linked to the 15‐day oscillation exhibit an eastward propagating wave train extending from northeastern Europe into the North Pacific/North American sector. In spatial scale, this wave train has zonal wavenumbers of 5–6. The wave trains associated with the 9‐ and 15‐day oscillations have similar behaviour patterns over northern Eurasian. However, the 9‐day wave train has a comparatively smaller spatial scale with zonal wavenumbers of 6–7. Therefore, the intraseasonal oscillations of precipitation are connected to the well‐organised wave trains that extend from the Euro‐Atlantic region to northern Eurasia, suggesting that the oscillation on each intraseasonal time scale must be produced by propagation of the corresponding wave train. Copyright © 2011 Royal Meteorological Society

Temporal and spatial patterns of precipitation variability for annual, wet, and dry seasons in Turkey

AbstractIn the present study, annual, wet and dry seasons precipitation records for the period 1961–2008 from 271 stations in Turkey were analysed using the rotated empirical orthogonal function (REOF), the Mann‐Kendall trend test and the continuous wavelet transform (WT) method. Additionally, relationships between time variability of the significant spatial patterns and North Atlantic Oscillation (NAO), Arctic Oscillation (AO), and North‐Sea Caspian Pattern (NCP) are examined. The REOF method was used to analyse the annual, dry, and wet season variability of precipitation patterns over Turkey. The Mann‐Kendall method was used to detect the temporal trend of the rotated principal components (RPCs) time series, and the continuous wavelet method was used to explore the periodicity of precipitation changes. Continuous WT results indicate that the significant 3–4 year, 6–10 year, and 12–16 year bands are the major period components. Precipitation in Turkey is uneven in space and time, and its complex temporal structure and spatial variations are different in each dry and wet season. The Mann‐Kendall test results show that decreasing annual precipitation is the dominating trend throughout Anatolia, including west, and southwest sections. Increasing annual precipitation can be observed in only northeast Black Sea region of Turkey. Decreasing wet/dry season precipitation that we observe throughout the country, except northeast coasts and eastern parts of Turkey, is expected to have a strong impact on the economic livelihood of the region, especially on agricultural production, drinking water supply, and hydroelectricity production. Copyright © 2010 Royal Meteorological Society

Sea surface temperature cooling mode in the Pacific cold tongue

Long‐term variability in sea surface temperature (SST) in the equatorial Pacific and its relationship with global warming were investigated using three SST data sets (Hadley Center Global Sea Ice and Sea Surface Temperature, extended reconstruction sea surface temperature, and Kaplan), atmospheric fields from National Centers for Environmental Prediction/National Center for Atmospheric Research reanalysis, and subsurface sea temperature from the Simple Ocean Data Assimilation data set. A cooling mode in the equatorial Pacific cold tongue is evident in all three SST data sets for two periods: 1870–2007 and 1948–2007. This cooling, which is indicated by the second empirical orthogonal function mode, is characterized by cooling in the Pacific cold tongue and warming elsewhere in the tropical Pacific. Its principal component time series is highly correlated with global mean surface temperature combining air temperature and SST. In association with the SST cooling mode, atmospheric fields and subsurface sea temperature are coupled in the tropical Pacific during recent decades. Moreover, for the coupled models in the 20th century run (20C3M), obtained from the Intergovernmental Panel on Climate Change Fourth Assessment Report database, those with realistic features of El Niño‐Southern Oscillation (ENSO) events can well show the cooling mode. However, the cooling mode is not shown in these coupled models in a preindustrial scenario with no forcing attributed to global warming. Results from observations and models suggest that the cooling mode is very likely caused by global warming. This conclusion is supported by a hypothesis that considers dynamic effects in the equatorial Pacific Ocean in response to global warming.

Reconstruction of Global Monthly Upper-Level Temperature and Geopotential Height Fields Back to 1880

Abstract This work presents statistically reconstructed global monthly mean fields of temperature and geopotential height (GPH) up to 100 hPa for the period 1880–1957. For the statistical model several thousand predictors were used, comprising a large amount of historical upper-air data as well as data from the earth’s surface. In the calibration period (1958–2001), the statistical models were fit using the 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) as the predictand. After the weighting of the predictors, principal component (PC) analyses were performed on both the predictand and predictor dataset. Multiple linear regression models relate each principal component time series from the predictand with an optimal subset of principal component time series from the predictor. To assess the quality of the reconstructions, statistical split-sample validation (SSV) experiments were performed within the calibration period. Furthermore, the reconstructions were compared with independent historical upper-air and total ozone data. Based on the SSV experiment, this study obtained good reconstructions for temperature and GPH in the Northern Hemisphere; however, the skill in the tropics and the Southern Hemisphere was much lower. The reconstruction skill shows a clear annual cycle with the highest values in January. The lower levels were better reconstructed except in the tropics where the highest levels showed the best skill. With the inclusion of a considerable amount of upper-air data after 1939 the skill increased substantially. The fields were analyzed for selected months in the 1920s and 1930s to demonstrate the usefulness of the reconstructions. It is shown that the reconstructions are able to capture regional-to-global dynamical features.