https://orcid.org/0000-0003-0866-7435
Figures
Abstract
This study investigated the influence of the Antarctic Oscillation (AAO) on the interannual variability of precipitation in the Indian monsoon region, aiming to address discrepancies in the existing literature. The degree of correlation between the AAO and moisture transport from the main oceanic sources supplying moisture to India, as well as precipitation over the region, was assessed using Pearson and partial correlation coefficients. The latter was employed to account for the potential confounding influence of the El Niño-Southern Oscillation (ENSO), the dominant mode of climatic variability. Five datasets were used: moisture sinks derived from the FLEXPART model driven by ERA-Interim data (1980-2012), and precipitation data (1984-2016) from two surface rain gauge datasets (IITM-IMR and IMD4), ERA5 reanalysis, and PERSIANN-CDR. In addition, spatial patterns of climate teleconnections associated with the AAO were analyzed using a composite approach. Seasonal time series were constructed under two analytical frameworks: the first (scenario A1) was based on patterns linked to seasonal variations in solar radiation between hemispheres due to Earth’s orbital movement around the Sun, while the second (scenario A2) focused on patterns associated with the strengthening and weakening of the Antarctic polar vortex, considering the specific climatic characteristics of India. Among the key findings, scenario A2 yielded improved results compared to scenario A1. The strongest correlation emerged over the Peninsular region, where the AAO showed a significant in-phase relationship with precipitation during October, November, and December (OND), and an out-of-phase relationship with precipitation during January-February (JF), relative to the OND period. Another notable result was found for the Northwestern region, where a positive relationship was identified between the AAO during the June-September (JJAS) period and precipitation during the subsequent OND season. These findings align with the “coupled ocean-atmosphere bridge” mechanism, which involves anomalies in the high-pressure system over the Indian Ocean and its interaction with key components of the Indian monsoon system.
Citation: Campos P, Castillo R (2025) Antarctic oscillation: Its influence on the interannual variability of precipitation in the Indian monsoon region. PLOS Clim 4(7): e0000671. https://doi.org/10.1371/journal.pclm.0000671
Editor: Andrea Storto, National Research Council, ITALY
Received: January 2, 2025; Accepted: June 18, 2025; Published: July 16, 2025
Copyright: © 2025 Campos and Castillo. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All original datasets are available freely from the original sources as stated in the Data section. Processed data from the article are available in the repository kerwa.ucr.ac.cr of the University of Costa Rica (https://hdl.handle.net/10669/102367).
Funding: The authors gratefully acknowledge the support provided by projects VI-C0074, VI-B9609, VI-B9454, VI-B7605, VI-B6147 and VI-C5455 of the University of Costa Rica, which enabled this research.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The influence of the Antarctic Oscillation (AAO) on precipitation in India has been the focus of several studies, employing diverse methodologies to explore this relationship. Noteworthy contributions include the works of Viswambharan & Mohanakumar [1], Prabhu et al. [2], Dou et al. [3], and Pal et al. [4].
Viswambharan & Mohanakumar [1], using data from the NCEP/NCAR (National Centers for Environmental Prediction/National Center for Atmospheric Research) reanalysis [5,6] alongside surface rain gauge observations, identified an in-phase correlation between the AAO Index (AAOI) and precipitation for June, as well as for the July-August period. Similarly, Prabhu et al. [2] demonstrated that the February-March AAOI strongly correlates with precipitation during the Indian summer monsoon, indicating that a positive (negative) AAOI increases (decreases) rainfall from June to September. Dou et al. [3] reported a positive correlation between the May AAOI and precipitation in June and July, while Pal et al. [4], using NCEP/NCAR reanalysis data, observed a correlation between July-September precipitation and the June AAOI.
Although these studies collectively suggest that the AAO influences interannual precipitation patterns in India, discrepancies in the reported periods of influence raise questions about the reliability of the AAOI as a predictive index for precipitation behavior. Furthermore, these four studies focus solely on the summer monsoon, leaving the potential impact of the AAO on other periods of the annual precipitation cycle unexplored.
These discrepancies can be attributed to various methodological factors, such as differences in the selected analysis periods, variations in the sources of AAOI data, and divergent approaches to temporal averaging (e.g., monthly or bi-monthly). To address these inconsistencies, the present study investigates the modulation of the AAO on India’s precipitation patterns through a more comprehensive and standardized approach. Specifically, this methodology involves a comparative analysis using multiple datasets from diverse sources, the standardization of the time periods analyzed, and significant enhancements to the spatial resolution of the data. Improving spatial resolution is especially crucial as it facilitates a more detailed and homogeneous regionalization study at the sub-regional level in India.
This research comprises two key components. The first is a moisture transport analysis, which examines the potential modulation of the AAO on moisture transport from the main oceanic sources supplying moisture to India, using moisture sinks data calculated by Castillo [7] in his study on global moisture sources. The second is a precipitation and teleconnection analysis, which measures the relationship between the AAO and interannual precipitation in India and analyzes the spatial teleconnection patterns associated with the AAO in the region.
The findings from this study offer valuable insights for decision-making in sectors heavily dependent on precipitation seasonality, including agriculture, food security, industry, tourism, and hydroelectric power generation. Additionally, these results offer valuable insights that can support operational forecasting efforts conducted by India’s meteorological services.
Interannual variability of precipitation over India
The interannual variability of precipitation patterns across India has been extensively studied, particularly in relation to the summer monsoon. A well-established relationship exists between monsoonal rainfall variability and the El Niño-Southern Oscillation (ENSO). During this period, an El Niño (La Niña) phase is typically associated with negative (positive) rainfall anomalies over India [8–12]. However, recent research suggests a weakening of this relationship in recent decades, implying that other factors might also influence monsoonal rainfall variability [13,14].
Among these factors, the Indian Ocean Dipole (IOD) has gained significant attention due to its capacity to influence precipitation patterns across the Indian region [15,16]. The Quasi-Biennial Oscillation (QBO) has also been a subject of interest. Studies, such as that by Chattopadhyay & Bhatla [17], suggest that an easterly QBO over the Niño 3 region may amplify ENSO phase effects on the monsoon, whereas a westerly QBO could lead to a normal monsoon. Additionally, the North Atlantic Oscillation (NAO) has been recognized as a potential modulator of Indian monsoon dynamics, interacting with oscillations like the QBO and ENSO [18–20].
Beyond the monsoon season, winter precipitation in northern India has been linked to oscillations such as ENSO, NAO, and the Arctic Oscillation (AO). These oscillations influence precipitation patterns through their impact on northern wind circulation [21–25]. Furthermore, during the OND period, precipitation variability has been associated with sea surface temperature (SST) anomalies and the IOD [26–29].
Role of the AAO in boreal climate variability
The AAO is the primary mode of large-scale climate variability in the Southern Hemisphere. It is characterized by atmospheric pressure anomalies between mid- and high latitudes, driving a north-south shift in the westerly wind belt encircling the South Pole, resulting in mass exchanges between these regions [30]. Using principal component analysis, Thompson & Wallace [31] demonstrated that the geopotential height anomalies associated with the AAO exhibit a symmetric and equivalent barotropic structure. They identified a dipole pattern with opposing pressure anomaly centers between high and mid-latitudes, with a node near 45∘S at lower levels. The polarity is defined as positive when polar anomalies are below normal and negative otherwise. They also observed that the structural characteristics of each polarity amplify and extend into the lower stratosphere during specific times of the year, distinguishing active from inactive seasons.
The AAO’s active season occurs in November, coinciding with the breakdown of the Southern Hemisphere’s polar vortex, a feature that differentiates it from the AO in the Northern Hemisphere. The inactive season comprises two subperiods: June-August (JJA), when the polar vortex is active and westerly winds strengthen, and February-March (FM), marked by an inactive vortex and prevailing easterly winds.
Building on this foundation, subsequent studies explored the AAO’s role in atmospheric circulation and precipitation anomalies across the Southern Hemisphere. Research on its impacts on Northern Hemisphere regions has emerged more recently [32].
Nan & Li [33] were among the first to examine the relationship between boreal spring AAO and summer precipitation in China’s Yangtze River Valley. They found that a positive AAO phase during boreal spring strengthens and shifts the North Pacific subtropical high westward, weakening the East Asian summer monsoon and increasing precipitation in the Yangtze River Valley. Subsequent studies (e.g., Xue et al. [34]; Fan [35]; Sun et al. [36]; Nan et al. [37]) confirmed these findings.
Nan et al. [37] further explored the physical mechanisms underpinning these relationships, finding that a positive spring AAO phase is associated with positive SST anomalies between 20∘S-30∘S that persist into summer, as well as anomalies in the Bay of Bengal and equatorial Indian Ocean. These anomalies correlated with a weaker East Asian monsoon, although their work left unanswered questions about the processes connecting mid-latitude SST anomalies to tropical maritime regions and precipitation changes.
Years later, Liu et al. [38] addressed some of these gaps by studying the influence of boreal autumn AAO on boreal winter precipitation in the Northern Hemisphere. They concluded that a positive autumn AAO phase increases precipitation in equatorial and mid-latitudes of the Northern Hemisphere while decreasing it in subtropical regions, with opposite effects for a negative phase. This relationship was attributed to the “coupled ocean-atmosphere bridge”, where AAO-driven wind anomalies alter ocean-atmosphere heat exchange, generating SST anomalies that persist across seasons, modulating baroclinicity and meridional circulation in both hemispheres.
Zheng et al. [39] examined the boreal winter AAO’s impact on spring circulation and precipitation patterns, also emphasizing the “coupled ocean-atmosphere bridge”. Their analysis showed that AAO-induced SST anomalies generate temperature gradients that alter atmospheric baroclinicity, triggering meridional wave adjustments and redistributing heat and momentum through eddy fluxes. These dynamics produce latitudinal bands of altered precipitation, with a positive AAO phase associated with negative anomalies in mid-latitudes (50∘S-37∘S) and tropics (12∘S-2∘S) of the Southern Hemisphere, as well as subtropical Northern Hemisphere (22∘N-38∘N). Positive anomalies occur in subtropical Southern Hemisphere (35∘S-25∘S) and tropical Northern Hemisphere (5∘N-18∘N).
Additionally, the AAO influences atmospheric components such as low-level jets. Gao et al. [40] linked boreal winter AAO to the Somali Low-Level Jet, demonstrating that a positive AAO phase strengthens high-pressure systems in the southern Indian Ocean (Mascarene and Australian highs), increasing the pressure gradient and jet intensity, leading to earlier Indian summer monsoon onset. Shi et al. [41] confirmed these relationships.
In summary, the AAO exerts significant influence on atmospheric circulation and precipitation patterns across both hemispheres, mediated through SST anomalies and meridional flow adjustments. This study focuses on analyzing these teleconnection patterns through surface anomalies in wind and precipitation, emphasizing the AAO’s role in boreal climate variability.
Data and methods
Data set
This research employed five distinct datasets:
- Moisture Sinks Data: The first dataset contains moisture sinks values [mm/day] calculated by Castillo [7] over a 33-year period, from December 1979 to November 2012 (1980-2012), with a spatial resolution of
. These values were derived using the FLEXPART model, which utilized ERA-Interim reanalysis data [42] provided by the European Centre for Medium-Range Weather Forecasts (ECMWF).
- ERA5 Reanalysis Data: The second dataset consists of ERA5 reanalysis data [43] spanning 33 years, from December 1983 to December 2016 (1984-2016). These data feature a monthly temporal resolution and a spatial resolution of
. The analyzed meteorological variables include precipitation [mm], vertically integrated moisture flux divergence [
], and the zonal and meridional wind components [m/s] at 925 hPa.
- PERSIANN-CDR Precipitation Data: The third dataset is PERSIANN-CDR, which provides daily precipitation records [mm] converted to monthly values for this study [44]. This database spans from January 1983 to the present; 33 years of data, from December 1983 to December 2016 (1984-2016), were used for this research. It has a spatial resolution of
and covers latitudes between 60∘S and 60∘N. The data were generated through an artificial neural network that incorporated satellite-derived infrared and microwave imagery, ground-based meteorological stations, radar data, topographic features, and other sources. Bias correction was performed using the Global Precipitation Climatology Project (GPCP) dataset [45–48].
- IITM-IMR Precipitation Data: The fourth dataset includes monthly regional and subdivisional precipitation data [mm] from the Indian Institute of Tropical Meteorology (IITM-IMR) spanning 33 years, from December 1983 to December 2016 (1984-2016). These data are derived from a network of 306 rain gauge stations distributed across 90% of India’s total area. The data are organized into six series:
- (i) The “entire” Indian region, encompassing all 306 stations and 90% of the country’s territory, excluding the northern mountainous regions of Jammu & Kashmir, Himachal Pradesh, Uttaranchal, and Arunachal Pradesh.
- (ii) Five homogeneous monsoon precipitation regions [49].
This classification is based on 30 meteorological subdivisions, each represented by an area-weighted average monthly precipitation series. Subdivisions with similar precipitation patterns were grouped by analyzing their associations with regional and global circulation parameters, culminating in the identification of five homogeneous monsoon precipitation regions in India [50]. - IMD4 Rainfall Data: The fifth dataset is a high spatial resolution (
) daily gridded rainfall product [mm], covering a long period from 1901 to 2024 over the Indian mainland. The historical rainfall data are archived at the National Data Center (NDC) of the India Meteorological Department (IMD) in Pune, based on records from 6995 rain gauge stations across India. The gridded dataset was developed following rigorous quality control procedures applied to the original rain gauge data [51]. For this study, daily precipitation values were aggregated to monthly resolution, and a 33-year subset (December 1983 to December 2016) was used.
Methodology for analyzing moisture sources
This analysis utilized moisture sinks data derived from the Doctoral Thesis by Castillo [7], which explored global moisture sources and their contribution to precipitation in various regions using the Lagrangian particle dispersion model FLEXPART. This model divides a section of the atmosphere into small parcels of air, referred to as particles [52]. As these particles travel from their source to their sink, the model quantifies the gain (e, evaporation) and loss (p, precipitation) of moisture by measuring specific humidity variations along their trajectories, as expressed in Eq 1:
Here, q represents specific humidity, and m is the mass of each particle. Summing the individual contributions of all particles within the atmospheric column yields the term E–P (Eq 2):
In this equation, E and P represent total evaporation and precipitation, respectively; A is the specific area of the atmospheric column; and K is the total number of particles in the column [53]. Regions where E–P > 0 are identified as moisture sources, while regions where E–P < 0 are considered sinks.
The global atmosphere was divided into approximately 2 million particles. Each particle was tracked for 10 days, as this is the average residence time of water vapor in the atmosphere [54]. The trajectories were calculated using reanalysis data at six-hour intervals (00, 06, 12, and 18 UTC) and with a spatial resolution of 1∘ latitude by 1∘ longitude. The 61 vertical levels of the reanalysis data were used, ranging from 0.1 to 1000 hPa, with approximately 14 levels below 1500 m and 23 between 1500 m and 5000 m. These particles were tracked forward in time from the oceanic regions under study to calculate the E–P field every six hours during the ten days of transport and then averaged to obtain the daily field used to establish the quarterly field, which overcomes computational sensitivity due to the time period used, as explained in more detail in Castillo et al. [55].
For this study, the contributions of four of the twelve oceanic moisture sources identified by Castillo [7] were quantified in terms of their role in moisture transport over India. These sources include the Zanzibar Current (ZAN), the Arabian Sea (ARAB), the Indian Ocean (IND), and the Red Sea (REDS) (Fig 1).
The red contours highlight regions with values exceeding 750 mm/year, based on the climatology of vertically integrated moisture flux divergence for the 1980-2012 period. Data derived from ERA-Interim. Adapted from Castillo [7].
To assess the seasonal contribution of these sources to moisture transport over India, precipitation data () from 33 years were averaged across four climatic seasons: June-August (JJA, boreal summer), December-February (DJF, boreal winter), March-May (MAM, boreal spring), and September-November (SON, boreal autumn).
It is noteworthy that these averages were computed using the South Asia (SAS) continental climatic region as a reference, which encompasses the Indian region. This area, identified by Giorgi & Francisco [56], is a climate change hotspot as indicated by the Regional Climate Change Index (RCCI). The results were plotted, considering only precipitation values exceeding the threshold of 0.05 mm/day ( mm/day) within the SAS region.
Methodology for calculating correlation coefficients
The Pearson correlation coefficient [57] was employed to quantify the degree of relationship between the Antarctic Oscillation Index (AAOI) and estimated precipitation derived from various datasets, including moisture sinks, surface rain gauge datasets (IITM-IMR and IMD4), ERA5 reanalysis, and PERSIANN-CDR. Subsequently, the partial Pearson correlation coefficient [57] was computed between these precipitation datasets and the AAOI, whilst controlling for the effect of the Oceanic Niño Index (ONI).
The ONI, obtained from the National Oceanic and Atmospheric Administration/Climate Prediction Center (NOAA/CPC) [58], corresponds to the Niño 3.4 region (5∘N-5∘S, 120∘-170∘W). It is derived using a three-month moving average of SST anomalies from the ERSSTv5 database [59]. Due to its relevance for monthly-scale analyses, the ONI serves as a key index for monitoring and forecasting the ENSO [32,53,60].
Incorporating ONI as a third variable in partial correlation calculations aimed to discern whether the observed correlations between AAOI and precipitation were influenced by a shared driver, ENSO. Castillo [32] emphasized that this method helps mitigate spurious correlations potentially arising from concurrent variability linked to ENSO. Only correlations with a significance level of 95% or higher, as determined by the t-test over a 33-year period, were considered. Coefficients exceeding ±0.34 were retained for analysis.
Partial correlation analyses were conducted on precipitation data from surface rain gauges (IITM-IMR and IMD4), ERA5 reanalysis, and PERSIANN-CDR, examining their relationships with the AAOI and ENSO indices. The study covered the All-India region, encompassing 90% of India’s territory, excluding the northern mountainous regions due to the limited availability of surface rain gauge stations. It also considered five homogeneous monsoon precipitation zones (Fig 2): Northwest, Central Northeast, Central West, Northeast, and Peninsular [50].
The selection of these regions was guided by surface rain gauge data (IITM-IMR), with the regional divisions based on similarities in precipitation patterns and their association with both regional and global circulation patterns affecting India [50,61].
Seasonal analysis framework.
Correlation analyses were based on seasonal averages under two analytical frameworks:
- Analytical Framework 1 (scenario A1): Seasonal divisions were determined by global circulation patterns representative of boreal winter (DJF), spring (MAM), summer (JJA), and autumn (SON). These patterns are primarily driven by variations in solar radiation across hemispheres due to Earth’s orbit around the Sun.
- Analytical Framework 2 (scenario A2): Seasonal periods were aligned with patterns of polar vortex strengthening and weakening, a key feature of the Antarctic Oscillation (AAO) [31], alongside the unique climate characteristics of India. These periods were classified into four climate seasons: Winter (DJF), Pre-Monsoon (MAM), Summer Monsoon (JJAS), and Post-Monsoon (OND) [62].
Methodology of analysis using the composite technique
Eq 3 outlines the mathematical structure of the composite technique employed to estimate the average state of a field conditioned by an external index [32,53].
Here, denotes the composite of the spatiotemporal signal
, conditioned by an index i across a number of observations j [53].
In this study, the technique was applied to achieve two primary objectives: first, to analyze the role of the AAO in the interannual variability of oceanic sources influencing moisture sinks; and second, to deduce the spatial teleconnection patterns driven by the AAO’s modulation of precipitation in the Indian monsoon region. Implementing this technique required initially obtaining the AAOI time series and utilizing them to identify the AAO’s extreme events.
Three methodologies are commonly used to calculate the AAOI: (i) those based on principal component analysis (PCA) of meteorological variables such as sea-level pressure or geopotential height [31,33]; (ii) those that calculate the difference between normalized zonal mean pressures at two latitudes using reanalysis data [30,63]; and (iii) those derived from surface data recorded by meteorological observation stations [64,65]. However, inconsistencies arise in determining extreme positive and negative phases when these methodologies are compared. These discrepancies primarily result from variations in the data types used to define the indices. To address this, the approach outlined by Nieto et al. [66], Castillo et al. [53], and Castillo [32] was adopted. This method incorporates two indices: one derived from normalized sea-level pressure data recorded by observation stations and another derived from reanalysis data of the same variable. Using both indices is crucial due to known limitations. Reanalysis-based indices are considered unreliable for periods prior to 1979 because satellite data were unavailable. Additionally, reanalysis data have demonstrated higher effectiveness in examining and understanding their relationship to associated impacts [67]. Consequently, employing both indices facilitates meaningful and robust comparisons [32].
The AAOI derived from observation station data utilized an index based on 12 stations, calculating the zonal mean sea-level pressure at 40∘S and 65∘S [64] (http://www.nerc-bas.ac.uk/icd/gjma/sam.html). Conversely, the AAOI derived from reanalysis data (NCEP/NCAR) [5,6] was defined as the normalized monthly difference in zonal mean sea-level pressure between 40∘S and 70∘S [33] (http://lijianping.cn/dct/page/65609).
Seasonal averages of the AAOI were computed annually for scenarios A1 and A2, generating time series for both observation-based and reanalysis-based indices. Extreme positive and negative phases were identified by selecting values exceeding the thresholds of the 18th and 82nd percentiles, respectively (Table 1). These thresholds approximate standard deviation values under the assumption of a normal statistical distribution. Neutral years were defined as those with values closest to zero. The results of this classification are summarized in Tables 2, 3, and 4. Once extreme episodes were identified, composites were constructed for years classified as positive, negative, and neutral phases (depicted in red, blue, and yellow, respectively) for the anomaly fields of key analysis variables, including: (i) moisture sinks derived from the FLEXPART model driven by ERA-Interim data, (ii) precipitation, vertically integrated moisture flux divergence, and 925 hPa winds from ERA5 reanalysis, and (iii) precipitation anomaly fields from PERSIANN-CDR. The spatial signals of AAO extreme events were extracted from the time series by calculating the differences between composites of positive and negative phases relative to the neutral phase composite for atmospheric anomaly fields. Statistical significance was assessed using a bootstrapping test [68], which involved permuting the time series 1000 times, with a confidence level exceeding 90%.
Discussion of results
Climatological analysis
Fig 3 displays the monthly climatological precipitation distribution derived from surface rain gauge data (IITM-IMR) spanning 1984-2016 across the homogeneous monsoon precipitation regions defined by Kothawale & Rajeevan [50]. Most regions exhibit a unimodal precipitation pattern with a peak in July, corresponding to the Indian summer monsoon during JJAS. However, the Peninsular region deviates, showing a bimodal pattern with relative peaks in July and October and a minimum in September. Moreover, annual total precipitation generally decreases from east to west, with the Northwest region experiencing the lowest monsoon season peak compared to others.
The seasonal climatological wind circulation patterns at 925 hPa, overlaid with precipitation fields in mm/day for scenarios A1 and A2, are presented in Fig 4. These patterns underscore notable features of each period, enriching the comparative analysis of the two scenarios.
During DJF and JF periods, prevailing winds in the region predominantly have a northeasterly component. After crossing the equator, they veer northwestward and converge with southeasterly winds in a zone between 20∘S and 5∘N. This convergence aligns with average precipitation maxima and the ITCZ’s climatological position over the Indian Ocean during this time [69]. Over land, northwesterly winds dominate in the north, while stronger southeasterly winds originating from the Bay of Bengal create a ridge over eastern India.
In the MAM period, the ITCZ appears less distinct, with only slight convergence of northwesterly and southeasterly winds near the equator between 65∘E and 80∘E. Winds are generally weak near the equator and southern Arabian Sea. A notable feature is a trough-like circulation enveloping the Indian peninsula, accompanied by winds entering from the Arabian Sea and Bay of Bengal. Coastal winds, particularly along the western coast, are stronger and linked to the onset of Shamal winds originating in the Persian Gulf and Gulf of Oman, typically commencing in late May [70]. Rainfall patterns reveal a primary maximum in the northeast and a secondary maximum in southern Peninsular India, with a pronounced east-west gradient.
During JJA and JJAS periods, southwesterly winds dominate the region and surrounding seas, intensifying with the Somali Low-Level Jet’s establishment and persistence [71,72]. Rainfall peaks over the Indian subcontinent, particularly in the northeast, east, central regions, and along the western coast adjacent to the Arabian Sea. While JJA experiences the highest rainfall, JJAS shows slightly reduced precipitation and weaker winds.
Finally, a comparison of SON and OND periods highlights key distinctions. SON circulation over the Indian Ocean mimics the summer monsoon due to the Somali Low-Level Jet’s persistence through September. However, over the subcontinent and surrounding seas, wind magnitudes decrease, with a northwesterly component prevailing over the Arabian Sea. Conversely, OND exhibits a winter-like circulation pattern, with northeasterly winds over India and surrounding seas north of the equator and summer monsoon-like southeasterly winds south of the equator. Wind convergence during OND occurs near the equator, coinciding with precipitation maxima. In contrast, SON lacks a well-defined convergence, with rainfall peaks occurring in northeastern India and southwestern Asia. These differences emphasize the importance of examining the AAO’s influence on Indian precipitation under both scenarios.
Correlation analysis for major oceanic moisture sources affecting the Indian monsoon region and associated climate teleconnection patterns
Fig 5 highlights the four main oceanic moisture sources influencing the SAS region during different climate seasons: ARAB, IND, ZAN, and REDS. During JJA, the four sources contribute to most of India, except for REDS, which has minimal impact on the eastern areas of the SAS region. In the other seasons, only ARAB exhibits contributions exceeding 0.05 mm/day. During DJF, ARAB’s influence is minimal and concentrated in the northwest, coinciding with one of the rainiest areas during this time due to “western disturbances” [73–75]. In MAM, ARAB provides moisture to the northwest, east, and western coastal areas, while in SON, its influence is primarily concentrated in the eastern region, encompassing the northeast, central east, and peninsular zones.
Only values () exceeding 0.05 mm/day were plotted and are shown using the same color as their respective oceanic source.
Table 5 summarizes the correlation and partial correlation coefficients, both in-phase and lagged, between the time series of moisture sinks from major oceanic sources and the AAOI derived on reanalysis data. No significant correlations were found when using the AAOI based on observational data. Notably, significant in-phase correlations (95% confidence level) were identified solely between ARAB-derived moisture sinks and the AAOI during SON. These correlations remained significant after controlling for ENSO influences.
In terms of lagged correlations, positive and significant relationships (95% confidence level) were observed for three sources (ARAB, ZAN and REDS) during JJA, lagged with MAM. These relationships also remained robust after adjusting for ENSO forcing.
Fig 6 illustrates spatial climate teleconnection patterns linked to the AAO, derived from the main oceanic moisture sources contributing to India. It emphasizes variations in moisture sinks magnitudes among positive and negative phase composites relative to the neutral phase composite for the AAO for in-phase and lagged relationships with the AAOI. DJF results are excluded due to a lack of significant values over India or adjacent areas.
Only values () exceeding 0.05 mm/day and within a statistical confidence level above 90% were plotted, based on a bootstrapping test with 1000 permutations of the time series.
While MAM correlations, both in-phase and lagged with the AAO, were not statistically significant, spatial teleconnection analyses for this period revealed positive and significant patterns. These patterns suggest increased moisture transport to India during AAO(+) phases from IND, ARAB, and ZAN, with ZAN showing the most pronounced contribution during MAM in-phase.
For MAM lagged with DJF, similar results were observed across IND, ARAB, and ZAN, with greater moisture transport toward areas near India during AAO(+) phases compared to AAO(-). No significant patterns were detected for REDS.
During JJA, significant correlations were confined to lagged analyses with MAM. Spatial teleconnection patterns during JJA revealed a notable contrast: negative patterns for JJA in-phase and positive patterns for JJA lagged. In-phase analysis indicated greater moisture transport to India during AAO(-), while lagged analysis (MAM to JJA) showed higher transport during AAO(+).
The negative pattern during JJA in-phase is associated with a weakened Indian Ocean high-pressure system during AAO(-) [34,76]. This weakening reduces the intensity of the Somali Low-Level Jet and its ability to transport moisture toward India’s east and northeast, redirecting it to the western and southern coasts. Conversely, the lagged pattern for JJA reflects increased moisture sink activity around India during AAO(+) [77,78].
For SON in-phase, positive correlation coefficients for ARAB align with observed moisture sinks anomalies, displaying positive patterns for ARAB, REDS, and ZAN. This indicates greater moisture transport to eastern and southern India during AAO(+). The underlying mechanism involves the weakening of the Somali Low-Level Jet’s east-southwest component, which persists climatologically over the Indian peninsula, north of Madagascar, and the Bay of Bengal until September [79,80].
Although SON lagged correlations with JJA were not significant, spatial teleconnection patterns demonstrated a positive signal, with enhanced moisture transport to India during AAO(+) in JJA. Notably, ZAN’s contribution surpassed that of the other three sources. This suggests that during AAO(+) in JJA, the oscillation’s influence extends across seasons, amplifying moisture transport to India during SON from ZAN. This inter-seasonal signal persistence is consistent with the “coupled ocean-atmosphere bridge” mechanism [38,39]. Additionally, ZAN’s dominance highlights the reinforcement of the Somali Low-Level Jet’s south-southwest component crossing the equator, possibly driven by strengthened Indian Ocean high pressure during AAO(+) [31].
Comparative analysis of correlations between scenarios A1 and A2 for seasonal periods and associated climate teleconnection patterns
Boreal Winter (DJF) vs. Indian Winter Season (JF).
The calculation of statistically significant Pearson correlation and partial correlation coefficients at the 95% confidence level for DJF in scenario A1 and JF in scenario A2 is summarized in Table 6. The vast majority of significant results were observed for lagged correlation coefficients. The near absence of in-phase signals during these periods can be attributed to the inactivity of the AAO and the mature phase of ENSO [60].
In the in-phase analysis, a negative and statistically significant correlation was observed in the Northeast during DJF, using precipitation series from surface rain gauge stations (IMD4) alongside the AAOI derived from reanalysis data. This signal remained significant even after accounting for potential ENSO forcing.
Regarding lagged results, correlations were stronger and more consistent across the four precipitation datasets in scenario A2 compared to scenario A1. For JF lagged with OND, correlations were consistently observed in the Peninsular region. However, when accounting for the potential influence of ENSO, the significance of correlations between the AAOI derived from reanalysis data was lost. Conversely, for DJF lagged with SON, a positive correlation was identified between the PERSIANN-CDR precipitation series and the AAOI derived from observational data, although this significance diminished after considering ENSO’s potential impact.
To identify the spatial patterns of climate teleconnections associated with the AAO in scenario A1 during the in-phase DJF season, differences in precipitation anomaly fields, 925 hPa wind patterns, and vertically integrated moisture flux divergence were analyzed. This analysis used ERA5 reanalysis data and PERSIANN-CDR precipitation data (Fig 7), comparing composite fields for the positive and negative AAO phases relative to the neutral phase.
The inferred spatial teleconnection patterns closely align with the detected correlations for the northeastern region. This association is particularly evident during the negative phase, when northwesterly wind anomalies favor the intrusion of “western disturbances”, enhancing moisture transport into the area [73–75]. These findings underscore the relevance of the negative AAO phase during DJF, which appears to support positive precipitation anomalies and negative anomalies in vertically integrated moisture flux divergence. In contrast, although the positive AAO phase may not exhibit a strong relationship with the interannual variability of DJF precipitation over northeastern India, the negative phase clearly demonstrates such a connection.
For the case of the spatial teleconnection patterns associated with the AAO in scenarios A1 and A2 for lagged seasonal periods (Fig 8), positive AAO phases are characterized by wind anomalies indicating enhanced northward meridional flow in both scenarios. These anomalies oppose the mean wind circulation during these periods, particularly through strengthened southeasterly components originating around 20∘S and reaching the equator, where they diverge northward toward southern India and the Bay of Bengal. These wind anomalies coincide with positive precipitation anomalies and negative vertically integrated moisture flux divergence anomalies over southern India. Notably, the spatial signature of these positive anomalies forms a diagonal structure extending from 10∘S to southern India, corresponding to the convergence of southeasterly and northwesterly wind anomalies just north of the climatological ITCZ position during these periods. In contrast, negative precipitation anomalies were observed further south.
This pattern suggests that a positive AAO phase during OND may induce a northward shift of the ITCZ in JF, leading to above-average precipitation in the Peninsular region. This relatively strong and coherent signal aligns with the reinforcement of the Indian Ocean High during a positive AAO phase and heightened AAO activity in November. These findings align with Thompson & Wallace [31], who highlighted increased variability in geopotential heights at 50 hPa in the Southern Hemisphere’s high latitudes between mid-October and mid-December, coinciding with the Antarctic polar vortex breakdown. Furthermore, the stronger correlation observed in scenario A2 compared to scenario A1 may result from the more prolonged influence of the OND AAO on the two-month JF atmospheric patterns, allowing a more persistent teleconnection signal compared to a three-month period.
The impact of the OND AAO on Indian precipitation patterns during JF, as well as the persistence of its influence across seasons, can be explained by the “coupled ocean-atmosphere bridge” mechanism described by Liu et al. [38] for boreal autumn. This mechanism impacts circulation and precipitation during boreal winter in the Northern Hemisphere. In summary, a positive (negative) AAO phase in OND may correspond to increased (decreased) northward meridional flow in surface circulation, resulting in higher (lower) precipitation in southern India.
Boreal Spring (MAM) vs. Pre-Monsoon (MAM).
The significant Pearson and partial correlation coefficients at the 95% confidence level for the MAM period in scenario A2 (lagged phase) are presented in Table 7. These correlations were significant only for MAM lagged with JF in scenario A2. Positive correlations were detected in the All-India and Central Northeast regions using precipitation series derived from ERA5 reanalysis data and PERSIANN-CDR. However, for the All-India region, correlations with PERSIANN-CDR data were significant only when using the AAOI derived from reanalysis data. Additionally, in the Peninsular region, positive correlations were observed exclusively with precipitation data from ERA5 reanalysis. Across all these cases, the coefficients remained statistically significant after accounting for potential ENSO forcing.
The detected correlations correspond to inferred spatial teleconnection patterns depicted in Fig 9 for MAM lagged with JF in scenario A2. This association is particularly notable in the negative phase, where the spatial signal of negative precipitation anomalies is more distinct than in the positive phase. Regarding the wind field, northeasterly anomalies are observed over the surrounding seas, fully opposing the mean seasonal circulation and adversely affecting moisture transport to the region.
These findings highlight the relevance of the negative AAO phase in JF, which appears to hinder favorable conditions for precipitation in eastern Indian regions. This is linked to a reduction in sea breeze inflow and the suppression of storm formation over and from the surrounding seas [81–83]. In contrast, while a positive AAO phase may not demonstrate a strong relationship with the interannual variability of MAM precipitation over India, a negative phase of the AAO clearly exhibits such a connection.
Boreal Summer (JJA) vs. Summer Monsoon (JJAS).
Table 8 presents the correlation and partial correlation coefficients, significant at the 95% level for the analysis periods JJA in scenario A1 and JJAS in scenario A2, considering both in-phase and lagged conditions. In the in-phase analysis, negative and statistically significant correlations were observed for JJA using precipitation data from surface rain gauge stations (IMD4) in the Central Northeast, and for JJAS using precipitation data from surface rain gauge stations (IITM-IMR) in the Peninsular region, both in association with the AAOI derived from observational data. Subsequently, positive and significant lagged correlations for JJA and JJAS with MAM were found only when using ERA5 reanalysis data and the AAOI derived from reanalysis in the Peninsular region. In all of the aforementioned cases, the signal weakens and loses statistical significance when ENSO forcing is taken into account.
Examining the spatial teleconnection patterns associated with the AAO for the seasonal periods of JJA and JJAS (both in-phase) under scenarios A1 and A2, respectively (Fig 10), it was observed that these patterns align with the inverse relationship identified in the correlation results. In both cases, during a positive AAO phase, easterly wind anomalies over the Bay of Bengal and northeasterly anomalies over the Arabian Sea promote anticyclonic rotation that opposes the mean monsoon circulation flow for these periods, suggesting a possible decrease in precipitation. The pattern observed for the negative AAO phase displays a clearer signal, with positive precipitation anomalies and negative vertically integrated moisture flux divergence anomalies extending across the Indian region.
It is important to consider that this result could be due to the strong influence of ENSO, given that significant correlation coefficients were scarce and partial correlations lost statistical significance in all cases. Additionally, the AAO is in an inactive season during JJA [31]. In this context, Pal et al. [4] found an inverse relationship (negative correlation) between the AAOI in June and precipitation over India in July. These researchers indicated that their results suggest an amplification effect of the AAO signal associated with ENSO phases. Previously, Viswambharan & Mohanakumar [1] obtained a similar result, with a negative correlation between the AAO in June and precipitation in July and August in India. Upon analyzing this work, Pal et al. [4] highlighted that there could be an underlying ENSO effect in the results of Viswambharan & Mohanakumar [1], given that this signal was not removed from the study of spatial teleconnection patterns.
As for the analysis of spatial teleconnection patterns for both scenarios, A1 and A2, in lagged phases (Fig 11), wind anomalies over the Indian region and surrounding seas oppose the mean climatological circulation flow for each period. However, during a positive AAO, positive precipitation anomalies were recorded in southern and south-central regions of India, while during a negative AAO, negative anomalies were detected across much of the region. Additionally, wind anomalies indicate a weakening of the Somali Low-Level Jet, which reinforces the positive signal recorded in the correlation results.
In Dou et al. [3], a similar result was found, with a positive relationship between the AAOI in May and precipitation in the first two months of the Indian summer monsoon (June and July) over southern and south-central regions of India. However, two significant differences are also highlighted: (i) the correlations obtained by Dou et al. [3] remain significant even when considering possible ENSO forcing, and (ii) Dou et al. [3] associates the AAO in May with precipitation in the early months of the monsoon, without considering August and September. In fact, these researchers observed that the relationship between the May AAO signal and precipitation in the last two months of the monsoon is not significant, which could be due, first, to the great variability of precipitation between and during the four months of the monsoon and, second, to the rapid decay of the May AAO signal, which does not influence August and September.
In summary, the described differences arise from the selection of the analyzed time periods, an aspect that could indicate that the AAO signal is highlighted on monthly and bimonthly time scales compared to longer scales. This is due to the great variability of precipitation patterns in the region during the monsoon season and the impact of the AAO signal, which persists for only two months.
A similar case is that of Prabhu et al. [2], who found a positive relationship where a positive (negative) AAO phase in February and March (FM) leads to an increase (decrease) in precipitation during the Indian summer monsoon. These researchers limited their analysis of the AAO’s influence during FM after examining the annual cycle of the AAOI for the period 1949-2013, observing that the AAOI presented two peaks throughout the year: the first during FM and the second in October and November (ON). After calculating the correlation coefficients with the pre-monsoon periods, they noted that FM showed the highest correlation with monsoon precipitation compared to other months from January to May.
In Prabhu et al. [2], the month of March was considered for the analysis, which is part of the MAM period. This difference in the extent of the analyzed periods reiterates that the AAO analysis could improve if the average signal of two months or less is considered.
Boreal Autumn (SON) vs. Post-Monsoon (OND).
The correlation and partial correlation coefficients, significant at the 95% confidence level, for SON period in scenario A1 and OND period in Scenario A2, both in phase and out of phase, are presented in Table 9. For the SON in-phase period, positive and significant correlations at the 95% level were observed in the regions of All-India, Central West, Central Northeast, and Peninsular. In contrast, for the OND in-phase period, significant correlations were identified in All-India, Northwest, and Peninsular regions.
When analyzing both periods in phase, the strongest correlation signal was primarily detected in the Peninsular region, followed by All-India. In other regions, correlations were either not significant in both studies or lost significance when accounting for ENSO forcing in most cases. In the Central West region, significant in-phase correlations during SON with the reanalysis-based AAOI were observed across the four precipitation datasets, but persisted only for ERA5 and PERSIANN-CDR after adjusting for ENSO influence. For the Northwest region, a positive correlation during OND was identified using surface rain gauge data (IITM-IMR) and reanalysis-derived AAOI, but this disappeared when ENSO variability was included, indicating a lack of complementarity among datasets in this region.
In the All-India region, statistically significant correlations at the 95% confidence level during SON were observed across the four precipitation datasets. For surface rain gauge data (IMD4) and the PERSIANN-CDR dataset, correlations were significant with both AAOI sources and remained so after accounting for ENSO forcing. In contrast, for surface rain gauge data (IITM-IMR) and the ERA5 dataset, correlations were significant only with the reanalysis-derived AAOI and lost significance when ENSO forcing was considered. Conversely, during OND, new significant coefficients emerged when ENSO forcing was excluded, particularly with observation-based AAOI for surface rain gauge datasets (IITM-IMR and IMD4) and the ERA5 data. For ERA5, this behavior was also observed with the reanalysis-derived AAOI.
In the Peninsular region, during SON, positive and significant correlations at the 95% level were observed only between surface rain gauge datasets (IITM-IMR and IMD4) and PERSIANN-CDR with observation-based AAOI, maintaining significance despite ENSO influence. For OND in phase, significant correlations were found using precipitation series from the four datasets with observation-based AAOI, even when accounting for ENSO forcing. PERSIANN-CDR data initially showed non-significant correlations, which intensified upon excluding ENSO effects.
For out-of-phase correlations, notable results were observed for OND out of phase with JJAS, with positive and significant correlations at the 95% level in the Northwest region across all datasets and AAOI sources. This signal persisted and even intensified when ENSO forcing was considered. For SON out of phase with JJA, significant correlations were identified in the Northwest region only with surface rain gauge (IITM-IMR) and ERA5 data using reanalysis-derived AAOI. These results are significant because values were not dampened by ENSO signals; instead, they appeared and strengthened when ENSO forcing was excluded.
When comparing spatial teleconnection patterns associated with the AAO for the two in-phase scenarios with AAOI (Fig 12), positive cases showed reinforced meridional circulation from south to north between 20∘S and the equator. North of the equator and west of 80∘E over the Arabian Sea, key differences were noted. In OND in phase, wind anomalies were more zonally oriented, with a westerly component, compared to SON in phase, where observed anomalies had a northwesterly component. Consequently, OND patterns represented greater moisture transport to western and southern Indian regions, aligning with positive correlation results. This explained positive precipitation anomalies and negative vertically integrated moisture flux divergence anomalies over these areas. Notably, the spatial precipitation pattern was more significant with PERSIANN-CDR data than with ERA5 reanalysis data.
For AAO(-) in phase during OND, wind anomalies over the southern Bay of Bengal opposed the climatological low-pressure circulation in this region, potentially inhibiting storm formation responsible for much of the precipitation in the Peninsular region during this period.
Out-of-phase correlations aligned with spatial teleconnection patterns observed using composite techniques (Fig 13). For AAO(+), a strengthened meridional circulation toward the north from 20∘S to India was observed, with westerly wind anomalies north of the equator, especially over the Arabian Sea. These patterns coincided with positive precipitation anomalies and negative vertically integrated moisture flux divergence anomalies for both SON and OND.
For AAO(-), results were more significant for OND out of phase with JJAS than SON out of phase with JJA. During OND, wind anomalies reinforced southward meridional circulation with northwesterly and westerly components south of the equator, opposing the mean wind flow during this period.
Both the in-phase and lagged results in scenarios A1 and A2 reflect the modulation exerted by the Antarctic polar vortex on surface anomalies, driven by the generation of an atmospheric wave train that accelerates surface southeasterly winds. This process is accompanied by a cold SST anomaly in the central and eastern equatorial Indian Ocean, triggered by wind-evaporation-SST feedback during a AAO(+). The resulting conditions stimulate anticyclonic anomalies over the eastern Arabian Sea, enhancing moisture transport toward India and favoring the formation and maintenance of precipitation [77]. This “coupled ocean-atmosphere bridge” mechanism involves the intensification of the high-pressure system over the Indian Ocean and its interaction with key components of the Indian monsoon system, including the Somali Low-Level Jet, the ITCZ, “western disturbances”, and other contributing systems [38–41].
Conclusion
This study offers valuable insights into the influence of the AAO on interannual precipitation variability in India, while also acknowledging inherent methodological limitations. The key findings are summarized as follows:
- The influence of the AAO on moisture sinks and precipitation patterns was explored across multiple seasons, including the monsoon period, offering a broader temporal perspective than previous studies.
- Correlation results generally aligned with spatial teleconnection patterns inferred from composite analysis, suggesting coherence in the observed relationships.
- Cross-validation using multiple precipitation and AAO datasets provided complementary evidence, particularly where AAO signals appeared strongest.
While the consistency between moisture transport and precipitation results supports the robustness of certain associations, these relationships should be interpreted with caution due to the observational nature of the data and the potential confounding influences of ENSO and the IOD, which are primary modes of Indian monsoon variability. Notably, positive correlations were observed for MAM lagged with JF, and for SON (both in-phase and lagged with JJA), while JJA showed mixed results, negative in-phase and positive when lagged with MAM.
Analysis of seasonal variations highlighted OND as the period with the most consistent correlation signals, followed by JF lagged with OND. Improvements under scenario A2 relative to A1 emphasize the importance of temporal framing and the need for sensitivity to the seasonal variability of both the AAO and the Indian climate system.
AAO influence on regional variability was also evident. The Peninsular and Northwestern regions exhibited relatively strong and coherent patterns across the four precipitation datasets, with correlations that remained significant even after accounting for the potential influence of ENSO, specifically during OND and JF (lagged with OND) in the Peninsular region, and OND lagged with JJAS in the Northwestern region. In contrast, other regions showed more sporadic or weaker signals. For instance, correlations in the Central-Northeastern region, limited to MAM lagged with JF, were only partially consistent and sensitive to the choice of dataset, although they remained statistically significant when ENSO forcing was considered.
Finally, this study highlights the utility, and also the limitations, of using multiple datasets to examine climate teleconnections. While convergence across datasets lends credibility to some of the findings, differences in resolution, coverage, and underlying assumptions also introduce uncertainties. Future work should explore complementary methods, including modeling approaches and longer observational records, to better isolate AAO impacts from other climate drivers.
References
- 1. Viswambharan N, Mohanakumar K. Signature of a southern hemisphere extratropical influence on the summer monsoon over India. Clim Dyn. 2012;41(2):367–79.
- 2. Prabhu A, Kripalani RH, Preethi B, Pandithurai G. Potential role of the February–March Southern Annular Mode on the Indian summer monsoon rainfall: a new perspective. Clim Dyn. 2015;47(3–4):1161–79.
- 3. Dou J, Wu Z, Zhou Y. Potential impact of the May Southern Hemisphere annular mode on the Indian summer monsoon rainfall. Clim Dyn. 2016;49(4):1257–69.
- 4. Pal J, Chaudhuri S, Roychowdhury A, Basu D. An investigation of the influence of the southern annular mode on Indian summer monsoon rainfall. Meteorol Appl. 2017;24(2):172–9.
- 5. Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L, et al. The NCEP/NCAR 40-year reanalysis project. Bull Amer Meteor Soc. 1996;77(3):437–71.
- 6. Kistler R, Collins W, Saha S, White G, Woollen J, Kalnay E, et al. The NCEP–NCAR 50–year reanalysis: monthly means CD–ROM and documentation. Bull Amer Meteor Soc. 2001;82(2):247–67.
- 7.
Castillo R. Global sources of moisture: characterization and study of its variability. Universidade de Vigo; 2015. https://ephyslab.uvigo.es/wp-content/uploads/2019/04/TesisRodrigoCastillo.pdf
- 8.
Chowdary JS, Parekh A, Gnanaseelan C. Indian summer monsoon variability: El Niño-teleconnections and beyond. 1st ed. Elsevier; 2021.
- 9. Walker GT. Correlation in seasonal variations of weather—a further study of world weather1. Mon Wea Rev. 1925;53(6):252–4.
- 10. Webster PJ, Yang S. Monsoon and enso: selectively interactive systems. Quart J Royal Meteoro Soc. 1992;118(507):877–926.
- 11. Gadgil S, Rajeevan M, Francis P. Monsoon variability: links to major oscillations over the equatorial Pacific and Indian oceans. Curr Sci. 2007;93(2):182–94. https://repository.ias.ac.in/94349/1/monsoon_variability.pdf
- 12. Ashok K, Tejavath CT. The Indian summer monsoon rainfall and ENSO. MAUSAM. 2021;70(3):443–52.
- 13. Kumar K, Rajagopalan B, Cane M. On the weakening relationship between the indian monsoon and ENSO. Science. 1999;284(5423):2156–9. pmid:10381876
- 14. Kumar KK, Rajagopalan B, Hoerling M, Bates G, Cane M. Unraveling the mystery of Indian monsoon failure during El Niño. Science. 2006;314(5796):115–9. pmid:16959975
- 15. Ashok K, Guan Z, Yamagata T. Impact of the Indian Ocean dipole on the relationship between the Indian monsoon rainfall and ENSO. Geophys Res Lett. 2001;28(23):4499–502.
- 16. Ashok K, Guan Z, Saji NH, Yamagata T. Individual and combined influences of ENSO and the Indian ocean dipole on the indian summer monsoon. J Climate. 2004;17(16):3141–55.
- 17. Chattopadhyay J, Bhatla R. Possible influence of QBO on teleconnections relating Indian summer monsoon rainfall and sea‐surface temperature anomalies across the equatorial pacific. Int J Climatol. 2002;22(1):121–7.
- 18. Dugam SS, Kakade SB, Verma RK. Interannual and long-term variability in the North Atlantic oscillation and Indian summer monsoon rainfall. Theor Appl Climatol. 1997;58(1–2):21–9.
- 19. Goswami BN, Madhusoodanan MS, Neema CP, Sengupta D. A physical mechanism for North Atlantic SST influence on the Indian summer monsoon. Geophys Res Lett. 2006;33(2).
- 20. Bhatla R, Singh AK, Mandal B, Ghosh S, Pandey SN, Sarkar A. Influence of North Atlantic oscillation on Indian summer monsoon rainfall in relation to quasi-binneal oscillation. Pure Appl Geophys. 2016;173(8):2959–70.
- 21. Dimri AP. Relationship between ENSO phases with Northwest India winter precipitation. Int J Climatol. 2012;33(8):1917–23.
- 22. Kar SC, Rana S. Interannual variability of winter precipitation over northwest India and adjoining region: impact of global forcings. Theor Appl Climatol. 2013;116(3–4):609–23.
- 23. Yadav RK, Ramu DA, Dimri AP. On the relationship between ENSO patterns and winter precipitation over North and Central India. Glob Planetary Change. 2013;107:50–8.
- 24. Thayyil Mandodi M, Pattanaik DR. The dependence of Indian winter precipitation extreme on the North Atlantic Oscillation. Nat Hazards. 2023;117(2):1869–85.
- 25. Sanap SD. Dynamical linkages between Arctic Oscillations and winter precipitation patterns over the Indian Region. Glob Planet Change. 2023;231:104293.
- 26. Kripalani RH, Kumar P. Northeast monsoon rainfall variability over south peninsular India vis‐à‐vis the Indian Ocean dipole mode. Int J Climatol. 2004;24(10):1267–82.
- 27. Yadav RK, Rupa Kumar K, Rajeevan M. Role of Indian Ocean sea surface temperatures in modulating northwest Indian winter precipitation variability. Theor Appl Climatol. 2006;87(1–4):73–83.
- 28. Sreekala PP, Rao SVB, Rajeevan M. Northeast monsoon rainfall variability over south peninsular India and its teleconnections. Theor Appl Climatol. 2011;108(1–2):73–83.
- 29. Rajeevan M, Unnikrishnan CK, Bhate J, Niranjan Kumar K, Sreekala PP. Northeast monsoon over India: variability and prediction. Meteorol Appl. 2012;19(2):226–36.
- 30. Gong D, Wang S. Definition of Antarctic Oscillation index. Geophys Res Lett. 1999;26(4):459–62.
- 31. Thompson DWJ, Wallace JM. Annular modes in the extratropical circulation. Part I: month-to-month variability. J Climate. 2000;13(5):1000–16.
- 32. Castillo R. La oscilación antártica: su influencia en la variabilidad interanual de la precipitación en Costa Rica. Rev Acad Colomb Cienc Ex Fis Nat. 2021;45(176):875–900.
- 33. Nan S, Li J. The relationship between the summer precipitation in the Yangtze River valley and the boreal spring Southern Hemisphere annular mode. Geophys Res Lett. 2003;30(24).
- 34. Xue F, Wang H, He J. Interannual variability of Mascarene high and Australian high and their influences on summer rainfall over East Asia. Chin Sci Bull. 2003;48(5):492–7.
- 35. FAN K. Atmospheric circulation in southern hemisphere and summer rainfall over Yangtze River Valley. Chin J Geophys. 2006;49(3):599–606.
- 36. Sun J, Wang H, Yuan W. A possible mechanism for the co‐variability of the boreal spring Antarctic Oscillation and the Yangtze River valley summer rainfall. Int J Climatol. 2008;29(9):1276–84.
- 37. Nan S, Li J, Yuan X, Zhao P. Boreal spring Southern Hemisphere annular mode, Indian Ocean sea surface temperature, and East Asian summer monsoon. J Geophys Res. 2009;114(D2).
- 38. Liu T, Li J, Zheng F. Influence of the Boreal Autumn Southern Annular Mode on Winter Precipitation over Land in the Northern Hemisphere. J Climate. 2015;28(22):8825–39.
- 39. Zheng F, Li J, Wang L, Xie F, Li X. Cross-seasonal influence of the December–February Southern Hemisphere annular mode on March–May meridional circulation and precipitation. J Climate. 2015;28(17):6859–81.
- 40. Gao H, Liu Y, Wang Y, Li W. Precursory influence of the Antarctic Oscillation on the onset of Asian summer monsoon. Chin Sci Bull. 2012;58(6):678–83.
- 41. Shi W, Xiao Z, Xue J. Teleconnected influence of the boreal winter Antarctic Oscillation on the Somali Jet: bridging role of sea surface temperature in southern high and middle latitudes. Adv Atmos Sci. 2015;33(1):47–57.
- 42. Dee DP, Uppala SM, Simmons AJ, Berrisford P, Poli P, Kobayashi S, et al. The ERA‐Interim reanalysis: configuration and performance of the data assimilation system. Quart J Royal Meteoro Soc. 2011;137(656):553–97.
- 43. Hersbach H, Bell B, Berrisford P, Hirahara S, Horányi A, Muñoz‐Sabater J, et al. The ERA5 global reanalysis. Quart J Royal Meteoro Soc. 2020;146(730):1999–2049.
- 44. Ashouri H, Hsu K-L, Sorooshian S, Braithwaite DK, Knapp KR, Cecil LD, et al. PERSIANN-CDR: daily precipitation climate data record from multisatellite observations for hydrological and climate studies. Bullet Am Meteorol Soc. 2015;96(1):69–83.
- 45. Huffman GJ, Adler RF, Arkin P, Chang A, Ferraro R, Gruber A, et al. The Global Precipitation Climatology Project (GPCP) combined precipitation dataset. Bull Amer Meteor Soc. 1997;78(1):5–20.
- 46. Huffman GJ, Adler RF, Bolvin DT, Gu G. Improving the global precipitation record: GPCP Version 2.1. Geophys Res Lett. 2009;36(17).
- 47. Adler RF, Huffman GJ, Chang A, Ferraro R, Xie P-P, Janowiak J, et al. The Version-2 Global Precipitation Climatology Project (GPCP) Monthly Precipitation Analysis (1979–Present). J Hydrometeor. 2003;4(6):1147–67.
- 48. Adler RF, Sapiano M, Huffman GJ, Wang J, Gu G, Bolvin D, et al. The Global Precipitation Climatology Project (GPCP) Monthly Analysis (New Version 2.3) and a Review of 2017 Global Precipitation. Atmosphere (Basel). 2018;9(4):138. pmid:30013797
- 49.
Mooley DA, Parthasarathy B, Rupa Kumar K, Sontakke NA, Munot AA, Kothawale DR. IITM Indian regional/subdivisional Monthly Rainfall data set (IITM-IMR). Indian Institute of Tropical Meteorology (IITM). 2007. https://iridl.ldeo.columbia.edu/documentation/IITM/All_India/Readme_v1871-2016.pdf
- 50.
Kothawale DR, Rajeevan M. Monthly, Seasonal and Annual Rainfall Time Series for All-India, Homogeneous Regions and Meteorological Subdivisions: 1871-2016. Pune, India: Indian Institute of Tropical Meteorology (IITM). 2017. https://www.tropmet.res.in/lip/Publication/RR-pdf/RR-138.pdf
- 51. Pai DS, Rajeevan M, Sreejith OP, Mukhopadhyay B, Satbha NS. Development of a new high spatial resolution (0.25° × 0.25°) long period (1901-2010) daily gridded rainfall data set over India and its comparison with existing data sets over the region. MAUSAM. 2014;65(1):1–18.
- 52. Stohl A, James P. A lagrangian analysis of the atmospheric branch of the global water cycle. Part II: Moisture transports between Earth’s Ocean Basins and River Catchments. J Hydrometeorol. 2005;6(6):961–84.
- 53. Castillo R, Nieto R, Gimeno L, Drumond A. Influencia de los principales modos anulares hemisféricos y El Niño-Oscilación del Sur (ENOS) en las fuentes de humedad globales de Mesoamérica. Rev Acad Colomb Cienc Ex Fis Nat. 2019;43(169):746–63.
- 54. Numaguti A. Origin and recycling processes of precipitating water over the Eurasian continent: experiments using an atmospheric general circulation model. J Geophys Res. 1999;104(D2):1957–72.
- 55. Castillo R, Nieto R, Drumond A, Gimeno L. Estimating the temporal domain when the discount of the net evaporation term affects the resulting net precipitation pattern in the moisture budget using a 3-D Lagrangian approach. PLoS One. 2014;9(6):e99046. pmid:24893002
- 56. Giorgi F, Francisco R. Uncertainties in regional climate change prediction: a regional analysis of ensemble simulations with the HADCM2 coupled AOGCM. Climate Dyn. 2000;16(2–3):169–82.
- 57.
Wilks DS. Statistical methods in the atmospheric sciences. 4th ed. Elsevier; 2019.
- 58. Smith TM, Reynolds RW, Peterson TC, Lawrimore J. Improvements to NOAA’s Historical Merged Land–Ocean Surface Temperature Analysis (1880–2006). J Climate. 2008;21(10):2283–96.
- 59. Huang B, Thorne PW, Banzon VF, Boyer T, Chepurin G, Lawrimore JH, et al. Extended Reconstructed Sea Surface Temperature, Version 5 (ERSSTv5): upgrades, validations, and intercomparisons. J Climate. 2017;30(20):8179–205.
- 60. Castillo R, Nieto R, Drumond A, Gimeno L. The role of the ENSO cycle in the modulation of moisture transport from major oceanic moisture sources. Water Resour Res. 2014;50(2):1046–58.
- 61.
Parthasarathy B, Munot AA, Kothawale DR. Monthly and seasonal rainfall series for all-India, homogeneous regions and meteorological subdivisions: 1871–1994. Pune, India: Indian Institute of Tropical Meteorology; 1995. https://www.tropmet.res.in/lip/Publication/RR-pdf/RR-65.pdf
- 62.
Shea DJ, Sontakke NA. The annual cycle of precipitation over the Indian subcontinent: daily, monthly and seasonal statistics. University Corporation for Atmospheric Research; 1995. https://dx.doi.org/10.5065/D65M63M8
- 63. Li J, Wang JXL. A modified zonal index and its physical sense. Geophys Res Lett. 2003;30(12):1632.
- 64. Marshall GJ. Trends in the Southern annular mode from observations and reanalyses. J Climate. 2003;16(24):4134–43.
- 65. Visbeck M. A station-based Southern annular mode index from 1884 to 2005. J Climate. 2009;22(4):940–50.
- 66. Nieto R, Castillo R, Drumond A. The modulation of oceanic moisture transport by the hemispheric annular modes. Front Earth Sci. 2014;2.
- 67. Ho M, Kiem AS, Verdon-Kidd DC. The Southern annular mode: a comparison of indices. Hydrol Earth Syst Sci. 2012;16(3):967–82.
- 68. Efron B. Second thoughts on the bootstrap. Statist Sci. 2003;18(2).
- 69. Zhang H, Seager R, Xie S-P. How does sea surface temperature drive the intertropical convergence zone in the Southern Indian Ocean? J Climate. 2022;35(16):5415–32.
- 70. Yu Y, Notaro M, Kalashnikova OV, Garay MJ. Climatology of summer Shamal wind in the Middle East. JGR Atmospheres. 2016;121(1):289–305.
- 71. Findlater J. A major low‐level air current near the Indian Ocean during the northern summer. Quart J Roy Meteoro Soc. 1969;95(404):362–80.
- 72. JOSEPH PV, RAMAN PL. Existence of low level westerly Jet Stream over peninsular India during July. MAUSAM. 1966;17(3):407–10.
- 73. Yadav RK, Rupa Kumar K, Rajeevan M. Characteristic features of winter precipitation and its variability over northwest India. J Earth Syst Sci. 2012;121(3):611–23.
- 74. Sen Roy S. Spatial variations in the diurnal patterns of winter precipitation in India. Theor Appl Climatol. 2008;96(3–4):347–56.
- 75. Nageswararao MM, Mohanty UC, Ramakrishna SSVS, Nair A, Prasad SK. Characteristics of winter precipitation over Northwest India using high-resolution gridded dataset (1901–2013). Glob Planet Change. 2016;147:67–85.
- 76. XUE F, WANG H, HE J. Interannual variability of mascarene high and Australian high and their influences on East Asian summer monsoon. J Meteorol Soc Jpn. 2004;82(4):1173–86.
- 77. Tang Y, Duan A, Hu J. Surface heating over the tibetan plateau associated with the Antarctic Oscillation. JGR Atmospheres. 2022;127(17).
- 78. Tang Y, Duan A, Hu D. Influence of the Antarctic oscillation on summer precipitation over East Asia. Atmosph Res. 2023;292:106847.
- 79. Roja Raman M, Venkat Ratnam M, Rajeevan M, Jagannadha Rao VVM, Bhaskara Rao SV. Intriguing aspects of the monsoon low-level jet over peninsular india revealed by high-resolution GPS radiosonde observations. J Atmosph Sci. 2011;68(7):1413–23.
- 80. Thapliyal R. The monsoon low-level jet: climatology and impact on monsoon rainfall over the west coast and central peninsular India. J Meteorol Res. 2023;37(1):112–25.
- 81. Sathiyamoorthy V, Shukla BP, Pal PK. Increase in the pre‐monsoon rainfall over the Indian summer monsoon region. Atmosph Sci Lett. 2010;11(4):313–8.
- 82. Mitra C, Shepherd JM, Jordan T. On the relationship between the premonsoonal rainfall climatology and urban land cover dynamics in Kolkata city, India. Int J Climatol. 2011;32(9):1443–54.
- 83. Vishnu TR, Kumar GNS, Kalyan S, Nadh BP, VenkateswaraRao M, Umakanth N. Study of pre-monsoon convection-based parameters over four stations of Northern India. Arab J Geosci. 2022;15(19):1553.