The spatial and temporal uncertainty of water availability is one of the major challenges in water resource management. Understanding patterns and identifying trends in seasonal to annual precipitation are very important for water infrastructure management. In particular, forecasts that incorporate such information can be used to inform decisions about the operation of multipurpose reservoir systems in the face of changing climate conditions.

Success in making useful forecasts is often achieved by considering climate teleconnections such as the El Niño–Southern Oscillation (ENSO) as related to sea surface temperature variations and air pressure over the globe using empirical data (Amarasekera et al., 1997; Denise et al., 2017; Korecha and Sorteberg, 2013; Seibert et al., 2017). Also, modes of variability of other tropical oceans can be related to regional precipitation (Dettinger and Diaz, 2000; Eden et al., 2015; Maity and Nagesh Kumar, 2006; Malmgren et al., 2007; Ranatunge et al., 2003; Suppiah, 1996; Roplewski and Halpert, 1996). For example, the effect of the Indian Ocean Dipole (IOD) is identified as independent of the ENSO effect (Eden et al., 2015). Pacific decadal oscillation (PDO), Atlantic multi-decadal mode oscillation (AMO), ENSO, and IOD teleconnections to precipitation have been found by many studies over the globe. Variations in precipitation in the United States are explained by ENSO, PDO, and AMO (Eden et al., 2015; National Oceanic and Atmospheric Administration, 2017; Ward et al., 2014); in African countries by ENSO, AMO, and IOD (Reason et al., 2006); and in southeastern Asian countries by ENSO: Indonesia (Lee, 2015; Nur'utami and Hidayat, 2016), Thailand (Singhrattna et al., 2005b), China (Cao et al., 2017; Ouyang et al., 2014; Qiu et al., 2014), Australia (Bureau of Meteorology, 2012; Verdon and Franks, 2005), and central and south Asia (Gerlitz et al., 2016).

The impact of ENSO and IOD on the position of the intertropical convergence zone (ITCZ) has been identified as a primary factor driving south Asian tropical climate variations. South Asian countries get precipitation from two monsoons from the movements of ITCZ in boreal summer (2∘ N) and boreal winter (8∘ S). The southwestern monsoon (summer monsoon) is during June–August and the northeastern monsoon (winter monsoon) is during December–February (Schneider et al., 2014). Climate teleconnections have been studied for summer monsoons (Singhrattna et al., 2005a; Surendran et al., 2015) and winter monsoons (Zubair and Ropelewski, 2006). A negative correlation of ENSO with the Indian summer monsoon has been identified (Jha et al., 2016; Surendran et al., 2015).

The objective of this study is to explore the climate teleconnection to dual monsoons and inter-monsoons. Water resource management decisions are typically based on precipitation throughout the year, and it is extremely important to explore the possibility that rainfall might be related to teleconnection indices for which seasonal forecasts are available. Sri Lanka is a south Asian country that gets rainfall from two monsoons and two inter-monsoons. We explore ENSO and IOD climate teleconnection to Sri Lanka precipitation throughout the year. Past studies have identified a climate teleconnection linking precipitation to climate indices for several months and monsoon seasons, and shown the importance of these for forecasting rainfall in river basins (Chandimala and Zubair, 2007; Chandrasekara et al., 2017). We extend these analyses across monsoon and inter-monsoon seasons.

Although rainfall anomalies may be correlated strongly with teleconnection indices, the scatter in the data can be large, making predictions from regression models have high uncertainty. However, water managers may act on information about whether rainfall is expected to be abnormally low or high. Seasonal precipitation is generally forecasted in broad categories. For example, the US National Weather Service forecasts seasonal precipitation as above normal, below normal, and normal (National Oceanic and Atmospheric Administration, 2018). The International Research Institute for Climate and Society also forecasts seasonal precipitation as above, below, and near normal (International Research Institute for Climate Society, 2018). We chose to follow a similar approach and investigate river basin rainfall teleconnections to climate indices with classification models. If reasonably accurate relationships can be developed, they will be useful for water resource management. For example, in Sri Lanka decisions about allocations of water for irrigation and hydropower could be improved with estimates of when low rainfall seasons are likely.

Sri Lanka is an island in the Indian Ocean (latitude 5∘55′–9∘50′ N, longitude 79∘40′–81∘53′ E). Mean annual rainfall varies from 880 to 5500 mm across the island. The rainfall distribution is determined by the monsoon system of the Indian Ocean interacting with the elevated land mass in the interior of the country. The country is divided into three climatic zones according to the rainfall distribution: humid zone (wet zone) (annual rainfall > 2500 mm), intermediate zone (2500 mm < rainfall < 1750 mm), and arid zone (dry zone) (rainfall < 1750 mm) (Department of Agriculture Sri Lanka, 2017).

Sri Lanka, a water-rich country, has 103 river basins varying from 9 to 10 448 km2. A large fraction of the water resource management infrastructure of the country is associated with the Mahaweli and Kelani River basins. The catchment areas of the Mahaweli and Kelani are 10 448 and 2292 km2, respectively. The two rivers start from the central highlands. Mahaweli, the longest river, travels to the ocean 331 km in the eastern direction and the Kelani 145 km in the western direction. Average annual discharge volume for the Mahaweli and Kelani basins is 26368×106 and 8660×106 m3 respectively (Manchanayake and Madduma Bandara, 1999). The Kelani River basin is totally inside the humid zone, whereas the Mahaweli River basin migrates through all three climate zones (Fig. 1).

The temporal pattern of rainfall in Sri Lanka can be divided into four seasons as follows.

Generally there is low precipitation across the country from the northeastern monsoon (NEM), which includes most precipitation during January to February. The arid zone of the country receives significant precipitation from the NEM, while the humid zone receives very little rainfall during this period.

Reflecting the rainfall seasons, the country has two agriculture seasons: “Yala” (April–September) and “Maha” (October–March). Because the arid zone receives minimal precipitation during the SWM, the agricultural systems (165 000 ha) developed under the Mahaweli multipurpose project depend on irrigation water during the Yala season. The country depends on stored water to drive hydropower year round. The Mahaweli and Kelani hydropower plants of 810 and 335 MW capacity serve as peaking and contingency reserve power to the power system (Ceylon Electricity Board, 2015). Management of reservoir systems is performed to cater to both irrigation and hydropower requirements.

Seasonal values of MEI and DMI were used as the predictors to classify seasons into the three rainfall classes. The total data set is divided into 75 % for training the model and 25 % for testing model performance. Quadratic discriminant analysis (QDA) and classification trees were selected for the analyses. A random forest model also was applied to investigate the reliability of a cross-validated statistical forecast tool based on an advance estimate of MEI and DMI. We used the R programming language to carry out the statistical analyses. R packages caret, tree, randomForest, fitdistriplus, devtools, and quantreg are used for the studies.

Monthly rainfall box plots of eight subbasins over the year for 1950–2013 illustrate the seasonal and the spatial variation in rainfall patterns (Fig. 2). The largest fraction of total rainfall in the arid zone occurs at the end of the SIM (December) and during the NEM (January–February) with correspondingly high variability, whereas there is little rainfall in the arid zone during the SWM (May–September) with correspondingly little variability (Fig. 2h). The intermediate zone receives approximately 60 % of total rainfall from the SIM and NEM. Although the variability of the rainfall is low in the intermediate zone, high rainfall can occur in all seasons (Fig. 2c and d). In the humid zone, a large portion of rainfall occurs in SWM and early months of the SIM (October–November). High variability of humid zone rainfall is observed at the end of the FIM (April), in the SWM (May–September), and at the start of the SIM (October) (Fig. 2a, b, e–g).

Similar to other investigators, we observe several strong correlations between rainfall anomalies and the climate indices (Tables A1 and A2 and Appendix). Higher correlation values between MEI and rainfall anomalies can be seen compared to the correlation with other ENSO indices (Table 1). In addition, rainfall in the SWM is very important for stations in the humid zone of the country which is the source of a large amount of water stored in reservoirs (Table A2). Correlation coefficients between SWM rainfall at Norton Bridge are negative and strong, -0.31 for MEI (p=0.01) and -0.37 for DMI (p<0.01). The strength of the correlation notwithstanding, the residuals from a regression model indicate that high uncertainty would be associated with any forecast (Fig. 3). Thus, we are led to explore the efficacy of classification methods (Appendix).

We present classification results for two subbasins, one that has the highest rainfall during the NEM, Manampitiya, and one that has the highest rainfall for the SWM, Norton Bridge (Fig. 4). Norton Bridge represents the areal rainfall of reservoir catchments in the wet zone and Manampitiya represents the rainfall that contributes to irrigation tanks in the dry zone. Results of other subbasins are presented in the Appendix (Figs. A4–A7, Appendix). Because MEI has a higher correlation with rainfall anomalies than other ENSO indices, classification was done with only MEI and  MI.

The SWM is a season when the humid zone receives the bulk of rainfall. At Norton Bridge, the occurrences of the dry rainfall anomaly class in the SWM are seen to “clump” in the region of relatively high MEI and DMI. Both the classification tree and the QDA successfully identify the pattern (Fig. 4a and c) with an overall accuracy of 73 %, 19 and 16 correct out of 22 occurrences (Table 2). In the arid zone the NEM season is one of the most important for rainfall. At Manampitiya, the MEI provides the primary variable in the classification, with the dry anomaly class being correctly selected in 52 % by tree model and 95 % with the QDA model. The results suggest that it may be possible to identify seasons when it is expected to be anomalously dry. The correct classification of “average” conditions likely has less importance for water managers. We explored classification using two classes, “dry” and “not dry”. In this case, the classification model again correctly classifies 86 % of the anonymously dry cases and gets more than 69 % of the not dry cases correct (Fig. 5).

Classification trees are known to be unstable. That is, small changes in the observations can lead to large changes in the decision tree. The random forest approach overcomes the issue by building a “bag” of trees from bootstrap samples. The robustness of the model can then be checked by considering the “out-of-bag” error. The results of the random forest indicate that predictions of three rainfall anomaly classes using MEI and DMI are not feasible (Table 3). The out-of-bag error rate is close to two-thirds, which for three categories is equivalent to a random selection.

However, the results of the random forest for a classification as either dry or not dry suggest that there may be skill in such a prediction. The out-of-bag error rates for this case range from 22 % to 38 % for Norton Bridge and Manampitiya (Table 3) and from 20 % to 39 % across all stations (Table A7).

The QDA method produces results that are promising with respect to identification of extreme dry events as indicated by seasonal rainfall (Table 5).

Understanding seasonal rainfall variability across the spatially diverse Mahaweli and Kelani River basins is important for irrigation and hydropower water planning. The SWM and SIM are the key rainfall seasons for subbasins in the humid zone (Norton Bridge, Morape, Peradeniya, and Laxapana), delivering 80 % of annual rainfall (Fig. 2a, b, e, f). For the arid zone (Manampitiya) and intermediate zone (Randenigala, Bowatenna) subbasins, the major season is SIM, which delivers more than 40 % of annual rainfall (Fig. 2c, d, h). The arid zone also receives rainfall during the NEM (24 % of annual rainfall at Manampitiya) and the intermediate zone receives rainfall during the SWM (25 %–30 % of annual rainfall at Randenigala and Bowatenna).

Climate teleconnection indices are related to rainfall anomalies observed during the two main growing seasons, Yala and Maha. The Maha agriculture season (October–March) depends on rain from the SIM and NEM. During El Niño events rainfall increases for the first 3 months of the Maha season (SIM: October–December) (Figs. A4, A5, A6, A8) (Ropelewski and Halpert, 1995) and decreases during the last 3 months (NEM: January–March) (Fig. 4b). In the Yala season (April–September), La Niña events enhance the rainfall during the SWM (Figs. 4a, c, A4, A5, A6, A8) (Whitaker et al., 2001). During El Niño events the SWM rainfall is reduced (Figs. 4a, c, A8, A9) (Chandrasekara et.al, 2017; Chandimala and Zubair, 2007; Zubair, 2003). The El Niño impact during the SWM is not as significant as it is during the NEM season (International Research Institute, 2017a). We find, however, that there is an interaction between two teleconnection indices, MEI and IOD for SWM rainfall. During the Yala season there is a high probability of having a drought when both the IOD and MEI are positive (Fig. 5). Also, not having drought is probable when both the IOD and MEI are negative (Figs. 5, A8, A9).

Classification of wet, average, and dry rainfall anomalies using the MEI and DMI indices is successful. For example, a dry SWM season for Norton Bridge (Table 2) and other humid zone stations (Table A4) is classified correctly with greater than 70 % accuracy with QDA and tree models. However, a random forest approach demonstrates that there is little skill in identifying a full wet–average–dry classification. However, a random forest model using only two rainfall categories shows more than 60 % accuracy in identifying dry and not dry classes of key rainfall seasons of the humid zone (Tables 4 and A7). Similarly, for arid zone locations such as Manampitiya, the dry rainfall class identification for NEM and SIM seasons is about 60 % (Tables 4 and A7).

Our statistical classification models can be combined with MEI and DMI forecasts to indicate the season-ahead expectation for rainfall. ENSO forecasts are available from the International Research Institute for Climate and Society (International Research Institute, 2017b) and IOD forecasts are available in the Bureau of Meteorology (BOM), Australian Government (Bureau of Meteorology, 2017). ENSO and IOD predictions are also associated with the uncertainty. Therefore, final forecast accuracy is a combination of the MEI, DMI forecast uncertainties, and model's accuracy rate in each class. Although overall prediction accuracy is not extremely high, a forecast of an anomalously low rainfall season can have value for risk-averse farmers (Cabrera et al., 2007) and can guide plans for hydropower management (Block and Goddard, 2012).

The electricity and agriculture sectors of Sri Lanka heavily rely on Mahaweli and Kelani River water resources so season-ahead forecasts of abnormally low rainfall should be useful for decisions on adaptation measures. For example, water availability of the first 3 months of a growing season is important for crop selection and the extent of land to be cultivated. Hydropower planning and scheduling of maintenance of the power plants can also benefit from season-ahead forecasts. The damage that can occur due to incorrect rainfall forecasts in the agriculture and energy sectors can be minimized with emergency planning during the season, which is the usual practice.

Although the accuracy of predicting low or not low seasonal rainfall is not very high, decisions based on forecasts that are improvements over climate averages should be an improvement over current practices. The accuracy of statistical models can be improved with longer records, which are important to train the classification models. Also, models can be fine-tuned for important shorter periods such as crop planting months and harvesting months for irrigation water planning.

ENSO and IOD phenomena teleconnections with river basin rainfall provide potentially useful information for water resource management. Relationships identified between teleconnection indices and river basin rainfall agree with other research findings. Prediction of seasonal rainfall classes from ENSO and IOD indices can inform water resource managers in reservoir operation planning for both hydropower and irrigation releases.