Long Lead-Time Streamflow Forecasting Using Oceanic-Atmospheric Oscillation Indices

Pacific and Atlantic Ocean influence on the spatiotemporal variability of heavy precipitation in the western United States

Groundwater level response in U.S. principal aquifers to ENSO, NAO, PDO, and AMO

The ability to predict the future variability of groundwater resources in time and space is of critical importance in society’s adaptation to climate variability and change. Periodic control of large-scale ocean-atmospheric circulations on groundwater levels serves as a potentially effective source for relatively long-term forecasting. In this study, as a first national-scale assessment, we use the continuous wavelet transform, global power spectrum, and wavelet coherence analyses to quantify the linkage between the Atlantic Multidecadal Oscillation (AMO), Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), and El Niño Southern Oscillation (ENSO) and the representative groundwater levels of the 24 principal aquifers, scattered across 14 different climate zones of Iran. Aquifer storage variations are found to be partially derived from annual to interdecadal climate variability and not solely a function of pumping variability. Moreover, teleconnections are observed to be both frequency and time-specific. The significant coherence patterns between the climate indices and groundwater levels are concentrated at five frequency bands of the annual (~ 1 year), short-interannual (2–4 years), medium-interannual (4–6 years), decadal (8–12 years), and interdecadal (14–18 years), consistent with the dominant oscillations of climate indices. AMO’s strong linkage to groundwater variability is found to be at interdecadal and annual modes of groundwater levels while PDO’s highest imprint is concentrated in interannual, decadal, and interdecadal periodicities. Unlike ENSO for which the highest modulating influence is found to be across the decadal and interannual modes, the NAO’s footprint in aquifers is marked at annual and interdecadal frequency bands. Findings further show that the groundwater variability is driven primarily by a combination of multiple large-scale atmospheric circulations rather than a single atmospheric circulation index. The decadal and interdecadal oscillation modes are the dominant modes in Iranian aquifers. Findings also mark the unsaturated zone contribution in damping and lagging of the climate variability modes, particularly for the higher-frequency indices of ENSO and NAO so that groundwater variability is more correlated with those climate circulations that have lower frequency such as PDO and AMO. Finally, the data length is found to have a significant effect on the teleconnections if the time series are not contemporaneous and for each particular series, only one value of coherence/correlation is computed instead of separate computations for different frequency bands and different time spans.

We present a data‐driven model, Support Vector Machine (SVM), for long lead time streamflow forecasting using oceanic‐atmospheric oscillations. The SVM is based on statistical learning theory that uses a hypothesis space of linear functions based on Kernel approach and has been used to predict a quantity forward in time on the basis of training from past data. The strength of SVM lies in minimizing the empirical classification error and maximizing the geometric margin by solving inverse problem. The SVM model is applied to three gages, i.e., Cisco, Green River, and Lees Ferry in the Upper Colorado River Basin in the western United States. Annual oceanic‐atmospheric indices, comprising Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), Atlantic Multidecadal Oscillation (AMO), and El Nino–Southern Oscillations (ENSO) for a period of 1906–2001 are used to generate annual streamflow volumes with 3 years lead time. The SVM model is trained with 86 years of data (1906–1991) and tested with 10 years of data (1992–2001). On the basis of correlation coefficient, root means square error, and Nash Sutcliffe Efficiency Coefficient the model shows satisfactory results, and the predictions are in good agreement with measured streamflow volumes. Sensitivity analysis, performed to evaluate the effect of individual and coupled oscillations, reveals a strong signal for ENSO and NAO indices as compared to PDO and AMO indices for the long lead time streamflow forecast. Streamflow predictions from the SVM model are found to be better when compared with the predictions obtained from feedforward back propagation artificial neural network model and linear regression.

We investigate the potential impact of Stratospheric Aerosol Intervention (SAI) on the spatiotemporal behavior of large-scale climate teleconnection patterns represented by the North Atlantic Oscillation (NAO), Pacific Decadal Oscillation (PDO), El Niño/Southern Oscillation (ENSO) and Atlantic Multidecadal Oscillation (AMO) indices using simulations from the Community Earth System Models (CESM1 and CESM2). The leading Empirical Orthogonal Function of sea surface temperature (SST) anomalies indicates that greenhouse gas forcing is accompanied by increases in variance across both the North Atlantic (i.e., AMO) and North Pacific (i.e., PDO) and a decrease over the tropical Pacific (i.e., ENSO); however, SAI effectively reverses these global warming-imposed changes. The projected spatial patterns of SST anomaly related to ENSO show no significant change under either global warming or SAI. In contrast, the spatial anomaly patterns pertaining to AMO (i.e., in the North Atlantic) and PDO (i.e., in the North Pacific) changes under global warming are effectively suppressed by SAI. For AMO, the low contrast between the cold-tongue pattern and its surroundings in the North Atlantic, predicted under global warming, is restored under SAI scenarios to similar patterns as in the historical period. The frequencies of El Niño and La Niña episodes increase with greenhouse gas emissions in the models, while SAI tends to compensate for them. All climate indices’ dominant modes of inter-annual variability are projected to be preserved in both warming and SAI scenarios. However, the dominant decadal and interdecadal variability mode changes induced by global warming are exacerbated by SAI, particularly in the Atlantic-based AMO.

Teleconnections between El Niño/Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Indian Ocean Dipole (IOD) and Pacific Decadal Oscillation (PDO) and seasonal precipitation regimes over the Yangtze River basin have been analysed based on the rotated empirical orthogonal functions. Results show that ENSO is the leading driver of seasonal precipitation variability over the Yangtze River basin, and the spring precipitation has been influenced by the PDO and ENSO, the summer and autumn precipitation has been influenced by the ENSO and IOD, the winter precipitation has been influenced by the ENSO, IOD and NAO. Furthermore, changes for the seasonal occurrence and intensity of wet days linked to the ENSO, NAO, IOD and PDO indices have also been investigated to discover which is the dominant mechanism driving seasonal precipitation changes. And results indicated that the influences of ENSO, NAO, IOD and PDO on the seasonal occurrence and intensity of precipitation events are complex, such as that the negative PDO event at the same year tends to increase the spring occurrence of precipitation events in the southwestern part of the Yangtze River basin while the positive ENSO event a year earlier tends to increase the spring intensity of precipitation events in the east part of the Yangtze River basin.

Floods in the middle reaches of the Yangtze River threaten thousands of million people, causing casualties and economic loss. Yet, the prediction of floods in this region is still challenging. To better understand the floods in this region, we investigate the interdecadal-interannual rainfall variation of the flood season (April–September) in Hunan province. The relationship between the rainfall and the Pacific decadal oscillation (PDO), Atlantic Multidecadal Oscillation (AMO), and El Niño-Southern Oscillation (ENSO) are also analyzed. The results show that the precipitation in the flood seasons shows an interdecadal oscillation with a period of about 20 years, which is caused by the joint effect of the PDO and AMO. When the PDO and AMO are in the same phase, the corresponding flood season is characterized by more precipitation, and conversely, it is less precipitation. Further analyses show that in the year after El Niño, when the PDO and AMO are both in the positive phase, it is favorable for the west Pacific subtropical high (WPSH) to be stronger and more southward than normal. Such circulation anomaly is conducive to the water vapor transport to the southern China, and as a result there is more precipitation in Hunan. When the PDO and AMO are both in the negative phase, the WPSH is weaker than normal, but the India-Burma trough is obviously stronger, which is also favorable for the southwesterly water vapor transport to the southern China. However, in the next year of the La Niña year, regardless of the phase combination of the PDO and AMO, the southern coast of China are controlled by a negative geopotential height anomaly and the WPSH retreats to the sea, which is not conducive to the northward transport of water vapor, and the precipitation in Hunan is less than normal. But if only the cold SST background in the previous stage is considered (without reaching the standard of a La Niña event),is more precipitation in most of the Hunan Province. Therefore, at the interannual scale, the PDO and AMO also have a modulating effect on the precipitation signal. However, the interannual-scale ENSO signal has a greater influence on the precipitation in Hunan flood seasons. Our results will give implications for the predications of floods in Hunan.

Effect of ENSO modulation by decadal and multi-decadal climatic oscillations on contiguous United States streamflows

Estimating long‐lead time precipitation under the stress of increased climatic variability is a challenging task in the field of hydrology. A modified Support Vector Machine (SVM) based framework is proposed to estimate annual precipitation using oceanic‐atmospheric oscillations. Oceanic‐atmospheric oscillations, consisting of Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), Atlantic Multidecadal Oscillation (AMO), and El Niño–Southern Oscillation (ENSO) for a period of 1900–2008, are used to generate annual precipitation estimates with a 1 year lead time. The SVM model is applied to 17 climate divisions encompassing the Colorado River Basin in the western United States. The overall results revealed that the annual precipitation in the Colorado River Basin is significantly influenced by oceanic‐atmospheric oscillations. The long‐term precipitation predictions for the Upper Colorado River Basin can be successfully obtained using a combination of PDO, NAO, and AMO indices, whereas coupling AMO and ENSO results in improved precipitation predictions for the Lower Colorado River Basin. The results also show that the SVM model provides better precipitation estimates compared to the Artificial Neural Network and Multivariate Linear Regression models. The annual precipitation estimates obtained using the modified SVM modeling framework may assist water managers in statistically understanding the hydrologic response in relation to large scale climate patterns within the Colorado River Basin.

Interannual to Decadal Rainfall Variability over Greater Horn of Africa: Non-Stationary Teleconnections, Physical Mechanisms, and Predictability

Long-range precipitation forecasts are useful when managing water supplies. Oceanic-atmospheric oscillations have been shown to influence precipitation. Due to a longer cycle of some of the oscillations, a short instrumental record is a limitation in using them for long-range precipitation forecasts. The influence of oscillations over precipitation is observable within paleoclimate reconstructions; however, there have been no attempts to utilize these reconstructions in precipitation forecasting. A data-driven model, KStar, is used for obtaining long-range precipitation forecasts by extending the period of record through the use of reconstructions of oscillations. KStar is a nearest neighbor algorithm with an entropy-based distance function. Oceanic-atmospheric oscillation reconstructions include the El Nino-Southern Oscillation (ENSO), the Pacific Decadal Oscillation (PDO), the North Atlantic Oscillation (NAO), and the Atlantic Multi-decadal Oscillation (AMO). Precipitation is forecasted for 20 climate divisions in the western United States. A 10-year moving average is applied to aid in the identification of oscillation phases. A lead time approach is used to simulate a one-year forecast, with a 10-fold cross-validation technique to test the models. Reconstructions are used from 1658-1899, while the observed record is used from 1900-2007. The model is evaluated using mean absolute error (MAE), root mean squared error (RMSE), RMSE-observations standard deviation ratio (RSR), Pearson’s correlation coefficient (R), Nash-Sutcliffe coefficient of efficiency (NSE), and linear error in probability space (LEPS) skill score (SK).The role of individual and coupled oscillations is evaluated by dropping oscillations in the model. The results indicate ‘good’ precipitation estimates using the KStar model. This modeling technique is expected to be useful for long-term water resources planning and management.

The multidecadal variability of Indian Ocean sea surface temperature (IOSST) has an important impact on both the regional Indian Ocean climate and the global climate. Here, we explore multidecadal variability in the annual IOSST. Observational analysis shows that the annual IOSST multidecadal variability is not only related to the Pacific Decadal Oscillation (PDO), but also to the North Atlantic Oscillation (NAO). The NAO leads by 15–20 years the detrended annual IOSST in which the PDO signal of the same period has been removed. Further analysis reveals that the NAO leads the annual IOSST multidecadal variability through its leading effect on the Atlantic Multidecadal Oscillation (AMO). The AMO affects the vertical wind anomaly in the Indian Ocean region through the Atlantic–Indian Ocean multidecadal teleconnection (AIMT), which in turn affects the net longwave radiation in the Indian Ocean region, thus driving the annual IOSST multidecadal variability. A Hasselmann model based on NAO and PDO further verify the joint influence of the NAO and PDO on the multidecadal variability of the IOSST. A PDO-based linear model and a climate model that incorporates the NAO signal are also constructed for the annual IOSST. Results show that the climate model with the NAO signal can better simulate the annual IOSST. This again verifies that the NAO is part of the annual IOSST multidecadal variability source, indicating that the annual IOSST variability may be due to the combined influences of the NAO and PDO.

Abstract. Studies addressing climate variability during the last millennium generally focus on variables with a direct influence on climate variability, like the fast thermal response to varying radiative forcing, or the large-scale changes in atmospheric dynamics (e.g. North Atlantic Oscillation). The ocean responds to these variations by slowly integrating in depth the upper heat flux changes, thus producing a delayed influence on ocean heat content (OHC) that can later impact low frequency SST (sea surface temperature) variability through reemergence processes. In this study, both the externally and internally driven variations of the OHC during the last millennium are investigated using a set of fully coupled simulations with the ECHO-G (coupled climate model ECHAMA4 and ocean model HOPE-G) atmosphere–ocean general circulation model (AOGCM). When compared to observations for the last 55 yr, the model tends to overestimate the global trends and underestimate the decadal OHC variability. Extending the analysis back to the last one thousand years, the main impact of the radiative forcing is an OHC increase at high latitudes, explained to some extent by a reduction in cloud cover and the subsequent increase of short-wave radiation at the surface. This OHC response is dominated by the effect of volcanism in the preindustrial era, and by the fast increase of GHGs during the last 150 yr. Likewise, salient impacts from internal climate variability are observed at regional scales. For instance, upper temperature in the equatorial Pacific is controlled by ENSO (El Niño Southern Oscillation) variability from interannual to multidecadal timescales. Also, both the Pacific Decadal Oscillation (PDO) and the Atlantic Multidecadal Oscillation (AMO) modulate intermittently the interdecadal OHC variability in the North Pacific and Mid Atlantic, respectively. The NAO, through its influence on North Atlantic surface heat fluxes and convection, also plays an important role on the OHC at multiple timescales, leading first to a cooling in the Labrador and Irminger seas, and later on to a North Atlantic warming, associated with a delayed impact on the AMO.

A study of the influence of interdecadal and interannual oceanic / atmospheric influences on seasonal streamflow in the U.S. is presented. Unimpaired streamflow was identified for 639 stations in the U.S. for the period 1951 – 2002. Pacific Ocean [El Nino-Southern Oscillation (ENSO) and Pacific Decadal Oscillation (PDO)] and Atlantic Ocean [Atlantic Multidecadal Oscillation (AMO)] oceanic / atmospheric phases (e.g., warm or cold) were identified for the year (or season) prior to the year of the winter-spring season streamflow (i.e., long lead-time). Statistical significance testing of the difference in means (and medians) of streamflow, based on the interdecadal and interannual oceanic / atmospheric phase (warm or cold), was performed applying the two-sample t-test and the rank-sum test. Additionally, the coupled effects of the oceanic / atmospheric influences were evaluated, based on the long-term phase (warm or cold) of the interdecadal variables (PDO and AMO) and ENSO, and streamflow regions in the U.S. were identified that respond to these climatic couplings. The results show that, in addition to the well-established ENSO signal, the PDO and AMO influence streamflow variability in the United States. Additionally, the phase (warm or cold) of the PDO and AMO enhance (or dampen) the ENSO signal in several streamflow regions in the United States. By utilizing the winter-spring streamflow season (e.g., typical period of peak runoff) and the long lead-time for the oceanic / atmospheric variables, useful information can be provided to streamflow forecasters and water managers.

Max-stable based evaluation of impacts of climate indices on extreme precipitation processes across the Poyang Lake basin, China