Discharge variability in Romania using Palmer indices and a simple atmospheric index of large-scale circulation

ABSTRACT

The aim of this study is to analyse the seasonal characteristics of four Palmer indices calculated on the basis of data from 27 meteorological stations in Romania, and the impact of these indices on river discharges in the period 1931–1998. Our research also tests the influence of large-scale atmospheric circulation on these indices and on discharge. For each season, developments in the empirical orthogonal functions (EOF) and multivariate EOF (MEOF) are achieved. The MEOF representation highlights the overall characteristics of the four Palmer indices. It maximizes specific information for each season compared with individual information of each Palmer index. We then identify geographical areas with homogeneous distribution, taking into account both the discharge distribution and the rotated EOF components of each Palmer index. Finally, we analysed the impact of large-scale atmospheric circulation on hydro-climatic events in Romania by means of the Greenland-Balkan Oscillation Index (GBOI), which is shown to have a greater influence on southeastern Europe than the North Atlantic Oscillation Index (NAOI).     

QANAT SYSTEMS IN IRAN

Abstract

A fuzzy rule-based technique is used for modelling the relationship between climatic forcing and droughts in a Central/Eastern European country, Hungary. Two types of climatic forcing'called premises'are considered: atmospheric circulation patterns (CP) and El Niño Southern Oscillation (ENSO). Both the Hess-Brezowsky CP types and ENSO events influence the occurrence of droughts, but the ENSO signal is relatively weak in a statistical sense. The fuzzy rule-based approach is able to learn the high space—time variability of monthly Palmer Drought Severity Index (PDSI) and results in a proper reproduction of the empirical frequency distributions. The “engine” of the approach, the fuzzy rules, are ascertained from a subset called the learning set of the observed time series of premises (monthly CP frequencies and Southern Oscillation Index) and PDSI response. Then an independent subset, the validation set, is used to check how the application of fuzzy rules reproduces the observed PDSI.     

Spatiotemporal analysis of climate variability (1971–2010) in spring and summer on the Loess Plateau,China

By using the Variable Infiltration Capacity model with Palmer Drought Severity Index (VIC‐PDSI) model and Standardized Precipitation Index (SPI), spatiotemporal trends of climate variation during the main growing seasons for plants of Loess Plateau between 1971 and 2010 were detected and characterized. The VIC‐PDSI model is established by combining the VIC model with PDSI. The simulation results and the grids system of VIC were applied to substitute for the two‐layer bucket‐type model to do the hydrological accounting, which could improve the physical mechanism of PDSI and expand its application range. Our results suggest that the climate of the study area has experienced a drying and warming trend during the past four decades. Apart from some individual years and regions, there was a perpetuation of water deficit over the Plateau both in spring and summer. The drought frequency increased from southeast to northwest in spring, while the drought frequency decreased from southeast to northwest in summer. The climate in the southern part of the Loess Plateau, accounting for 23.3% of the study region, showed a significant drying and warming trend in spring over the past four decades. The climate variability detected by VIC‐PDSI model shows good agreement with that monitored by SPI. Since a large part of the study region frequently suffered from water shortage during the main growing seasons for plants, people living in such drought‐prone areas should take measures to prevent the negative effects on agricultural production, reforestation, and regional food security caused by drought. Copyright © 2013 John Wiley & Sons, Ltd.     

Comparative study of statistical methods to identify a predictor for discharge at Orsova in the Lower Danube Basin

ABSTRACT

The aims of this study are to investigate the influence of large-scale atmospheric circulation quantified by indices such as the North Atlantic Oscillation index (NAOI), the Greenland-Balkan Oscillation index (GBOI) and blocking-type indices on the Lower Danube discharge. We separately analysed each season for the 1948–2000 period. In addition to the statistical linear procedure, we applied methods to quantify nonlinear connections between variables, as mutual information between predictors and predictand, using Shannon’s information entropy theory. The nonlinear correlation information between climate indices and discharge is higher than that obtained from the linear measure, providing more insight into real connections. Also, the non-stationarity of the link between variables is highlighted by spectral coherence based on wavelet analysis. For the physical interpretation, we analyse composite maps over the Atlantic-European region. The most significant influence on the discharge of the Lower Danube Basin is given by the GBOI and blocking-type atmospheric circulation over Europe.     

Regional applicability of seven meteorological drought indices in China

The definition of a drought index is the foundation of drought research. However, because of the complexity of drought, there is no a unified drought index appropriate for different drought types and objects at the same time. Therefore, it is crucial to determine the regional applicability of various drought indices. Using terrestrial water storage obtained from the Gravity Recovery And Climate Experiment, and the observed soil moisture and streamflow in China, we evaluated the regional applicability of seven meteorological drought indices: the Palmer Drought Severity Index (PDSI), modified PDSI (PDSI_CN) based on observations in China, self-calibrating PDSI (scPDSI), Surface Wetness Index (SWI), Standardized Precipitation Index (SPI), Standardized Precipitation Evapotranspiration Index (SPEI), and soil moisture simulations conducted using the community land model driven by observed atmospheric forcing (CLM3.5/ObsFC). The results showed that the scPDSI is most appropriate for China. However, it should be noted that the scPDSI reduces the value range slightly compared with the PDSI and PDSI_CN; thus, the classification of dry and wet conditions should be adjusted accordingly. Some problems might exist when using the PDSI and PDSI_CN in humid and arid areas because of the unsuitability of empiricalparameters. The SPI and SPEI are more appropriate for humid areas than arid and semiarid areas. This is because contributions of temperature variation to drought are neglected in the SPI, but overestimated in the SPEI, when potential evapotranspiration is estimated by the Thornthwaite method in these areas. Consequently, the SPI and SPEI tend to induce wetter and drier results, respectively. The CLM3.5/ObsFC is suitable for China before 2000, but not for arid and semiarid areas after 2000. Consistent with other drought indices, the SWI shows similar interannual and decadal change characteristics in detecting annual dry/wet variations. Although the long-term trends of drought areas in China detected by these seven drought indices during 1961–2013 are consistent, obvious differences exist among the values of drought areas, which might be attributable to the definitions of the drought indices in addition to climatic change.     

Improving drought predictability in Arkansas using the ensemble PDSI forecast technique

Drought prediction is important for improved water resources management and agriculture planning. Although Arkansas has suffered severe droughts and economic loss in recent years, no significant study has been done. This study proposes a local nonparametric autoregressive model with designed stochastic residual-resampling approach to produce ensemble drought forecasts with associated confidence. The proposed model utilizes historical climate records, including drought indices, temperature, and precipitation to improve the quality of the short-term forecast of drought indices. Monthly forecasts of Palmer Drought Severity Index (PDSI) in Arkansas climate divisions show remarkable skills with 2–3 month lead-time based on selected performance measure such as, Normalized Root Mean Square Error (NRMSE) and the Kuiper Skill Score (KSS). Rank histograms also show that the model captures the natural variability very well in the produced drought forecasts. The incorporation of categorical long-term precipitation prediction significantly enhances the performance of the monthly drought forecasts.     

Drought monitoring

Drought detection, monitoring and indices are closely related to its definition. The specific definition chosen for a particular drought analysis will affect the procedures one uses in drought detection and monitoring. The traditional Palmer Drought Severity Index (PDSI) has been proven to be ineffective in regions of predominantly irrigated agriculture.The recently developed ALERT (Automated Local Evaluation in Real Time) system is proposed for use in monitoring the spatial and temporal variations of drought in real time. The ALERT system uses standardized instruments, radio frequencies, software and hardware. It was originally developed as a flash flood waming system by local flood control districts and the National Weather Service. However, now it has expanded to over 100 other uses in the areas of natural and man-made disaster detection and warning. The successful ALERT system indicates the need for the continued development of a national drought monitoring index that is applicable to a wide range of climate, hydrologic and water resource environments.     

Drought monitoring

Drought detection, monitoring and indices are closely related to its definition. The specific definition chosen for a particular drought analysis will affect the procedures one uses in drought detection and monitoring. The traditional Palmer Drought Severity Index (PDSI) has been proven to be ineffective in regions of predominantly irrigated agriculture.The recently developed ALERT (Automated Local Evaluation in Real Time) system is proposed for use in monitoring the spatial and temporal variations of drought in real time. The ALERT system uses standardized instruments, radio frequencies, software and hardware. It was originally developed as a flash flood waming system by local flood control districts and the National Weather Service. However, now it has expanded to over 100 other uses in the areas of natural and man-made disaster detection and warning. The successful ALERT system indicates the need for the continued development of a national drought monitoring index that is applicable to a wide range of climate, hydrologic and water resource environments.     

Influence of climate on alpine stream chemistry and water sources

The resilience of alpine/subalpine watersheds may be viewed as the resistance of streamflow or stream chemistry to change under varying climatic conditions, which is governed by the relative size (volume) and transit time of surface and subsurface water sources. Here, we use end‐member mixing analysis in Andrews Creek, an alpine stream in Rocky Mountain National Park, Colorado, from water year 1994 to 2015, to explore how the partitioning of water sources and associated hydrologic resilience change in response to climate. Our results indicate that four water sources are significant contributors to Andrews Creek, including snow, rain, soil water, and talus groundwater. Seasonal patterns in source‐water contributions reflected the seasonal hydrologic cycle, which is driven by the accumulation and melting of seasonal snowpack. Flushing of soil water had a large effect on stream chemistry during spring snowmelt, despite making only a small contribution to streamflow volume. Snow had a large influence on stream chemistry as well, contributing large amounts of water with low concentrations of weathering products. Interannual patterns in end‐member contributions reflected responses to drought and wet periods. Moderate and significant correlations exist between annual end‐member contributions and regional‐scale climate indices (the Palmer Drought Severity Index, the Palmer Hydrologic Drought Index, and the Modified Palmer Drought Severity Index). From water year 1994 to 2015, the percent contribution from the talus‐groundwater end member to Andrews Creek increased an average of 0.5% per year (p < 0.0001), whereas the percent contributions from snow plus rain decreased by a similar amount (p = 0.001). Our results show how water and solute sources in alpine environments shift in response to climate variability and highlight the role of talus groundwater and soil water in providing hydrologic resilience to the system.     

A new standardized Palmer drought index for hydro‐meteorological use

Accepting the concept of standardization introduced by the standardized precipitation index, similar methodologies have been developed to construct some other standardized drought indices such as the standardized precipitation evapotranspiration index (SPEI). In this study, the authors provided deep insight into the SPEI and recognized potential deficiencies/limitations in relating to the climatic water balance it used. By coupling another well‐known Palmer drought severity index (PDSI), we proposed a new standardized Palmer drought index (SPDI) through a moisture departure probabilistic approach, which allows multi‐scalar calculation for accurate temporal and spatial comparison of the hydro‐meteorological conditions of different locations. Using datasets of monthly precipitation, temperature and soil available water capacity, the moisture deficit/surplus was calculated at multiple temporal scales, and a couple of techniques were adopted to adjust corresponding time series to a generalized extreme value distribution out of several candidates. Results of the historical records (1900–2012) for diverse climates by multiple indices showed that the SPDI was highly consistent and correlated with the SPEI and self‐calibrated PDSI at most analysed time scales. Furthermore, a simple experiment of hypothetical temperature and/or precipitation change scenarios also verified the effectiveness of this newly derived SPDI in response to climate change impacts. Being more robust and preferable in spatial consistency and comparability as well as combining the simplicity of calculation with sufficient accounting of the physical nature of water supply and demand relating to droughts, the SPDI is promising to serve as a competent reference and an alternative for drought assessment and monitoring. Copyright © 2013 John Wiley & Sons, Ltd.     

亚洲和北美干湿变化及其与海表温度异常的关系

利用多通道奇异谱方法（MSSA）分析了1953~2003年亚洲和北美Palmer干旱指数（PDSI）与热带和北半球温带海洋海表面温度异常（SSTA）的主要周期振荡特征及其相互联系.结果表明：亚洲和北美PDSI以及SSTA均存在明显的3~6年的年际以及10年左右的年代尺度振荡;此外,亚洲PDSI还存在显著的6~8年的年际振荡.SSTA的年际振荡主要体现了ENSO的变化特征,而其年代尺度振荡的空间分布具有热带太平洋和北太平洋共同作用的类ENSO型.同时,MSSA的分析结果给出了亚洲和北美主要振荡信号的时间和空间演变特征.相关性分析表明,亚洲和北美PDSI的年际及年代尺度振荡均显示明显的对SSTA强迫信号的响应.对于年际振荡,亚洲PDSI对SSTA响应强于北美,但年代尺度振荡则反之.此外,亚洲和北美PDSI对于SSTA信号响应的关键区域也随时间尺度的不同而发生变化.亚洲的西西伯利亚、青藏高原东西两侧以及中西伯利亚东部在年际和年代尺度上均为受SSTA影响最显著的区域;在年际尺度上,北美中部地区的干湿变化与SSTA存在显著相关,而在年代尺度上,美国西部更易受SSTA年代尺度振荡的影响.     

Spatiotemporal variability of drought and the potential climatological driving factors in the Liao River basin

Spatiotemporal characteristics of drought based on the Standardized Precipitation Index (SPI) in the Liao River basin (LRB) are investigated in this study. High autocorrelation in SPI seems to lend itself to drought prediction. Drought is becoming more frequent, widespread, and severe in the LRB during the past several decades. Major factors affecting drought in this basin are analysed by investigating relationship between SPI and several circulations including western Pacific Subtropical High (WPSH), East Asian Summer Monsoon (EASM) and El Niño‐Southern Oscillation (ENSO) indices. Different correlation patterns between WPSH indices and SPI are obtained. Several significant positive correlations between the area, intensity of WPSH and SPI are observed in the west and the centre of the study area, while negative correlations are observed in the east. Reverse patterns are observed in the correlation between the ridge of westward longitude of WPSH and SPI. Corresponding lag‐correlation is dominated by positive correlations between the area, intensity of WPSH and SPI, and by negative correlation between the ridge of westward longitude of WPSH and SPI. EASM is mainly negative related with drought in the east of the LRB. Significant positive correlation between ENSO and SPI is mainly located in the east while negative correlation is located in west of the basin. Lag‐correlation (with lags of 1 to 12 months) between them is also investigated and results show that significant negative correlation is located in a broad area extending from the west to the centre of the basin, while less positive correlation is observed with the increase of lags. The possibility of employing general circulation models (GCMs) for drought prediction is discussed based on the above analyses. Copyright © 2011 John Wiley & Sons, Ltd.     

Correlation between El Niño–Southern Oscillation (ENSO) and precipitation in South‐east Asia and the Pacific region

The relationship between El Niño–Southern Oscillation (ENSO) events versus precipitation anomalies, and the response of seasonal precipitation to El Niño and La Niña events were investigated for 30 basins that represent a range of climatic types throughout South‐east Asia and the Pacific region. The teleconnection between ENSO and the hydroclimate is tested using both parametric and non‐parametric approaches, and the lag correlations between precipitation anomalies versus the Southern Oscillation Index (SOI) several months earlier, as well as the coherence between SOI and precipitation anomalies are estimated. The analysis shows that dry conditions tend to be associated with El Niño in the southern zone, and part of the middle zone in the study area. The link between precipitation anomalies and ENSO is statistically significant in the southern zone and part of the middle zone of the study area, but significant correlation was not observed in the northern zone. Patterns of precipitation response may differ widely among basins, and even the response of a given river basin to individual ENSO events also may be changeable. Copyright © 2003 John Wiley & Sons, Ltd.     

Identifying non-stationary and long-term river–aquifer interactions as a response to large climatic patterns and anthropogenic pressures using wavelet analysis (Mancha Oriental Aquifer,Spain)

The objective of this study was to analyse periodicities and the long-term variability of monthly Júcar River–Mancha Oriental Aquifer interactions (RAI) and regionally measured precipitation (PP) with special focus on the correlations between these local hydrological variables and the large climatic patterns governing the Iberian Peninsula, represented by their teleconnection indices – the North Atlantic Oscillation index (NAOi) and the Western Mediterranean Oscillation index (WeMOi). To that end, wavelet analysis has been applied since it not only provides insight into the time-series dynamics but also permits statistical interpretation and correlation analysis. As a result, several periodicities have been detected: intermittent semi-annual periodicity in PP and the NAOi and annual periodicity in the RAI, NAOi and WeMOi time series. Long cycles (approximately 14 years) are also observed in the PP and WeMOi time series. The cross-wavelet spectra show a correlation between the RAI and the rest of the variables on the semi-annual and the annual scales, while wavelet coherence detects common behaviour with longer cycles – 5–6 years between the NAOi and the RAI and cycles of both 1–5 years and 7–10 years between PP and the RAI. Furthermore, results show that the periodicities in the teleconnection indices and precipitation propagate into the RAI with certain lead times: 3 months between the RAI and PP and 6 months between the RAI and the NAOi. The results indicate that the detected periodicities and the coherence between the studied variables could have applications in strategic planning on a river basin scale, taking into account the propagation times and the frequency scale. This methodological approach can be applied into strategic water resource planning independently of the geographical location of the hydrogeological system, the basin size and the climate region.     

An insight into the Palmer drought mechanism based indices: comprehensive comparison of their strengths and limitations

Considering the drawbacks of the original Palmer drought severity index (PDSI) in terms of its simplified hydrologic algorithm and spatio-temporal inconsistency, we compare six variants of PDSI derived from different combinations of two hydrologic algorithms and three standard processes so as to provide deep insights into the individual impacts of hydrological processing and standardization on final PDSI values as well as their combined effects. Investigations are conducted in whole Yellow River basin. On basis of 52 years’ (1961–2012) hydro-meteorological data, comprehensive analysis on multiple drought characteristics are carried out for each PDSI variant, combined with comparison of three crucial intermediate variables of PDSI. Results show that variable infiltration capacity (VIC) model based modification in the hydrologic accounting section significantly improve drought trends with more reasonable spatial distributions presented. For the statistical characteristics of drought areas and frequency, comparable performance is found between VIC-based modification and self-calibrating standard procedure-based modification, though they are derived from different mechanisms. However, in case of the coupling of these two modifications, indices derived from combined modifications perform poorly than single modification-based indices with unexpected high frequency of extreme events detected in certain regions. This reflects the complicated mechanism of PDSI and it is essential to propose an appropriate standardization to match the hydrological algorithm and further improve the performance of relevant drought index. With the crucial findings mentioned above, this study is promising to provide some theoretical supports and serve as a competent reference for future PDSI based researches.     

The influence of hydroclimatic variability on flood frequency in the Lower Rhine

Climate change is expected to significantly affect flooding regimes of river systems in the future. For Western Europe, flood risk assessments generally assume an increase in extreme events and flood risk, and as a result major investments are planned to reduce their impacts. However, flood risk assessments for the present day and the near future suffer from uncertainty, coming from short measurements series, limited precision of input data, arbitrary choices for particular statistical and modelling approaches, and climatic non‐stationarities. This study demonstrates how historical and sedimentary information can extend data records, adds important information on extremes, and generally improves flood risk assessments. The collection of specific data on the occurrence and magnitude of extremes and the natural variability of the floods is shown to be of paramount importance to reduce uncertainty in our understanding of flooding regime changes in a changing climate. For the Lower Rhine (the Netherlands and Germany) estimated recurrence times and peak discharges associated with the current protection levels correlate poorly with historical and sedimentary information and seem biased towards the recent multi‐decadal period of increased flood activity. Multi‐decadal and centennial variability in flood activity is recorded in extended series of discharge data, historical information and sedimentary records. Over the last six centuries that variability correlates with components of the Atlantic climate system such as the North Atlantic Oscillation (NAO) and Atlantic Multi‐decadal Oscillation (AMO). These climatic non‐stationarities importantly influence flood activity and the outcomes of flood risk assessments based on relatively short measurement series. Copyright © 2016 John Wiley & Sons, Ltd.     

Book Review

The objective of the work discussed in this paper was to seek possible links between surface hydrology in Southwestern (SW) Poland and El Ni?o/Southern Oscillation (ENSO). Although the impact of ENSO on hydrology in Europe has been investigated by many researchers, no clear picture demonstrating spatial variability of such a teleconnection has yet been unequivocally reported. In particular, there is no comprehensive study on ENSO–streamflow links for Polish rivers. Herein, discharge time series from 15 sites located at lowland and mountain rivers in SW Poland and different ENSO indices are examined. They include atmospheric time series (axial component of atmospheric angular momentum, Southern Oscillation Index), oceanic indices (Ni?o 3.4 Index, Global SST Index), geodetic data (length-of-day), and the combined index (Multivariate ENSO Index). The data span the period from November 1971 to October 2006. On the basis of cross-correlation and wavelet analyses it was found that there is a weak but significant link between ENSO and surface hydrology in SW Poland. It is inferred that ENSO episodes may be among a few factors affecting winter and early spring discharges of rivers in SW Poland and may have a (probably limited) impact on snow-melt flood generation.     

Hydrologic model-based Palmer indices for drought characterization in the Yellow River basin,China

The Palmer indices (PIs) that have been most widely used for drought monitoring and assessment are criticized for two main drawbacks: coarse hydrological accounting processes with a simplified two-stage bucket soil water balance model and arbitrary rules for defining drought properties and standardizing index values through limited calibration and comparison. In this study, we introduce a new proposal of the VIC hydrologic model-based Palmer drought scheme, where traditional PIs (e.g. PDSI) can readily be calculated on the basis of distributed finescale hydrologic simulations. Moreover, recent variants of PI (i.e., SPDI and SPDI-JDI) also provide a preferable standardization strategy that allows probabilistic invariability and better spatio-temporal comparability of computed drought indices. Using gridded meteorological forcing, soil and vegetation data to drive the three-layer VIC model, both non-VIC and VIC-based PIs are investigated to examine their performances for drought characterization and detection. Results indicate that VIC hydrologic model would allow for adjustments in statistical properties of computed PDSI and VIC-based SPDI is also preferable to PDSI for better statistical robustness and spatio-temporal consistency/comparability. Moreover, the joint SPDI-JDI has the strength of integrating multi-scale probabilistic properties and drought information released by individual SPDI, providing overall drought conditions that take into account the onset, persistence and termination of droughts. At proposed 0.25° grid scale, the VIC-based SPDI-JDI indicates high frequency and long total time of drought condition in the Yellow River basin (YRB), China. Although no significant temporal trends are found in identified drought duration and severity, both the seasonal and annual drought index values demonstrate a downward trend (higher drought intensity) for considerable proportions of the YRB. These findings imply high drought risk and potential drying stress for this region. The new framework of hydrologic model-based PIs can help to strengthen our knowledge and/or practices in regional drought monitoring and assessment.     

Is There Any Teleconnection Between Surface Hydrology in Poland and El Niño/Southern Oscillation?

The objective of the work discussed in this paper was to seek possible links between surface hydrology in Southwestern (SW) Poland and El Niño/Southern Oscillation (ENSO). Although the impact of ENSO on hydrology in Europe has been investigated by many researchers, no clear picture demonstrating spatial variability of such a teleconnection has yet been unequivocally reported. In particular, there is no comprehensive study on ENSO–streamflow links for Polish rivers. Herein, discharge time series from 15 sites located at lowland and mountain rivers in SW Poland and different ENSO indices are examined. They include atmospheric time series (axial component of atmospheric angular momentum, Southern Oscillation Index), oceanic indices (Niño 3.4 Index, Global SST Index), geodetic data (length-of-day), and the combined index (Multivariate ENSO Index). The data span the period from November 1971 to October 2006. On the basis of cross-correlation and wavelet analyses it was found that there is a weak but significant link between ENSO and surface hydrology in SW Poland. It is inferred that ENSO episodes may be among a few factors affecting winter and early spring discharges of rivers in SW Poland and may have a (probably limited) impact on snow-melt flood generation.     

Ensemble prediction of regional droughts using climate inputs and the SVM–copula approach

In this study, the climate teleconnections with meteorological droughts are analysed and used to develop ensemble drought prediction models using a support vector machine (SVM)–copula approach over Western Rajasthan (India). The meteorological droughts are identified using the Standardized Precipitation Index (SPI). In the analysis of large‐scale climate forcing represented by climate indices such as El Niño Southern Oscillation, Indian Ocean Dipole Mode and Atlantic Multidecadal Oscillation on regional droughts, it is found that regional droughts exhibits interannual as well as interdecadal variability. On the basis of potential teleconnections between regional droughts and climate indices, SPI‐based drought forecasting models are developed with up to 3 months' lead time. As traditional statistical forecast models are unable to capture nonlinearity and nonstationarity associated with drought forecasts, a machine learning technique, namely, support vector regression (SVR), is adopted to forecast the drought index, and the copula method is used to model the joint distribution of observed and predicted drought index. The copula‐based conditional distribution of an observed drought index conditioned on predicted drought index is utilized to simulate ensembles of drought forecasts. Two variants of drought forecast models are developed, namely a single model for all the periods in a year and separate models for each of the four seasons in a year. The performance of developed models is validated for predicting drought time series for 10 years' data. Improvement in ensemble prediction of drought indices is observed for combined seasonal model over the single model without seasonal partitions. The results show that the proposed SVM–copula approach improves the drought prediction capability and provides estimation of uncertainty associated with drought predictions. Copyright © 2013 John Wiley & Sons, Ltd.