Disparity in spatiotemporal variability and risk of compound coastal extremes between India’s East-West coasts – npj Climate and Atmospheric Science

Seasonal variability of extremes

The seasonal pattern of independent extremes (exceeding 95th percentile) of rainfall, sea level, and skew surge at the coastal stations was examined (Fig. S1). Along the east coast, the extreme rainfall days occur mostly in the June–November months at Haldia, Paradip, and Visakhapatnam, and in the October–December months at Chennai. The extreme rainfall witnessed at (i) Haldia and Paradip (during July and August) is attributable to southwest monsoon, (ii) Visakhapatnam (during September-October) to the combined impact of southwest monsoon and northeast monsoon, and (iii) Chennai (during November) to northeast monsoon and intense LPS (viz. depressions, deep depressions, cyclones). The extreme rainfall events in the post-monsoon season are also impacted by tropical depressions and cyclones for all the stations on the east coast. Along the west coast, the rainfall extremes are witnessed primarily in the June–September months, and this is due to the impact of the southwest monsoon in this region. The sea level extremes show seasonality in the (i) monsoon for Kandla, Haldia, and Paradip, (ii) post-monsoon for Visakhapatnam and Chennai, and (iii) winter (December–January) for Kochi, Marmagao, and Mumbai. The seasonality of the sea level along the Indian coast resembles the seasonal cycle of coastal currents and the large-scale circulation in the North Indian Ocean^38,39. The seasonal variation of the coastal currents in the North Indian Ocean is shown in (Fig. S4). During the pre-monsoon season, the East India Coastal Current (EICC) is strong and poleward whereas the West India Coastal Current (WICC) is weak and equatorward⁴⁰. During the summer monsoon, both the WICC and EICC are weak due to local alongshore winds. The EICC is unorganized, whereas the WICC is equatorward during this season. During post-monsoon and winter, both EICC and WICC are strong and equatorward and poleward respectively. Equatorward (poleward) WICC and poleward (equatorward) EICC result in Ekman transport away (towards) from the coast and associated coastal upwelling (downwelling)⁴¹. Owing to the seasonal variation of EICC and WICC the east coast stations (i) Haldia and Paradip witness extreme sea level (ESL) peaks during August-October, (ii) Visakhapatnam and Chennai witness these peaks during November-December and west coast stations except Kandla witness these peaks during December-February. Furthermore, the equatorward WICC and subsequent upwelling at the western coast during the southwest monsoon leads to a lower sea level during this season despite freshwater discharge from some rivers^42,43.

The mean sea level is higher on the east coast compared to the west coast, and this is attributed to the large-scale wind-forced circulation and alongshore gradient in salinity (Figs. S8, S9)⁴⁰. The extreme skew surge (ESS) peaks occur mostly during the southwest monsoon (June–September) on the west coast, with some peaks during post-monsoon owing to storms and depressions. For Haldia and Paradip, the monthly ESS frequencies are bimodal in nature, with one peak during May–June and another during October–November, whereas for Visakhapatnam and Chennai, the peaks in monthly ESS frequencies are observed in October–November only (Fig. S1). This observation is attributed to the intense LPS (viz. tropical depressions and cyclones) during these periods in the Bay of Bengal. The tidal range, mean and variance of ESL at the stations located up in the northern part of the peninsula are higher than that in the stations located towards the southern part of the peninsula (Figs. S2, S3, S5). The reason for this is a wider continental shelf at the head of the bay for both coasts, leading to the amplification of semidiurnal tidal constituents.

The seasonal frequency of compound SL-RF and SS-RF extremes show different patterns for the East Coast and the West Coast stations (Fig. 2). Along the East Coast, SS-RF extremes occur mostly in the June–November months at Haldia, Paradip, and Visakhapatnam, and in the October–December months at Chennai. Along the West Coast, these extremes are witnessed primarily in the June–September months. During the monsoon (June–September), the SS-RF extremes at the northeastern coastal stations (Haldia, Paradip, and Visakhapatnam) are driven primarily by monsoon LPS forming in the Bay of Bengal, whereas during post-monsoon (October–December) those extremes are driven primarily by intense LPS (viz. tropical depressions and cyclones). The SS-RF extremes at Chennai in post-monsoon are driven primarily by the northeast monsoon, tropical depressions, and cyclones near the equatorial Indian Ocean. There is a gradual shift in the peak frequency month of SS-RF extremes from June to November as we proceed southwards from Haldia (north) to Chennai (south) along the East Coast (Fig. 2). For the west coast stations, the compound SS-RF extremes occur primarily during monsoon season (June–September), with few events occurring in October. This is attributed to the southwest monsoon and associated monsoon trough (in central India and Kandla on Gujarat coast) and offshore trough (along the west coast encompassing Kochi, Marmagao and Mumbai) during this season that gives rise to extreme rainfall and storm surge. The composite maps of means of total column rain water and 850 hPa wind anomalies show these synoptic patterns for SS-RF extremes on both the coasts (Fig. S12).

The frequency of SL-RF extremes was higher during (i) August- October at Haldia and Paradip, (ii) October at Visakhapatnam and (iii) October–December at Chennai on the East Coast. The seasonal variation in these extremes on the East Coast stations is attributed to the effect of seasonality of coastal currents on ESL along with seasonal variation of LPS activity on the Bay of Bengal (monsoon LPS during June–September and tropical depressions/cyclones during October–December) driving extreme rainfall and extreme sea level. Fifty percent of the witnessed SS-RF extremes contributed to SL-RF extremes at Chennai and Paradip (Fig. S10). The percentage was relatively less (25%) at Haldia and Visakhapatnam due to shift in seasonality of ESS and ESL. The frequency of SL-RF extremes was lower at the West Coast stations due to differences in seasonality of WICC and monsoon winds driving ESL and extreme rainfall respectively, at these stations. However, the weaker equatorward WICC at Kandla and Mumbai in monsoon (June–September) leads to a moderately higher frequency of SL-RF extremes at these stations compared to Marmagao and Kochi in this season. Considerably low percentage (S10).

Interannual variability in frequency of extremes

The annual frequencies of sea level, rainfall, and skew surge extremes were analyzed for the East and the West Coast stations over the period 1980–2020 (Figs. S6, S7). For extreme rainfall, there was no significant divergence in its variability between the East and West coasts. Among all stations, Haldia witnessed a significantly lower variance in the annual frequency of ESL days. Furthermore, Chennai, Marmagao, and Kochi witnessed higher variance in the annual frequency of ESL days, which is attributed to an increase in the annual frequency of ESL in the current decade due to a higher rate of mean sea level rise²⁹. The West Coast stations (except Kochi) also witness an increasing trend in the annual frequency of ESS in the current decade due to the rising frequency of TCs (Vayu, Nisarga, Biparjoy, Tauktae, Tej to name a few) in the Arabian Sea⁴⁴. On the contrary, the annual frequency of ESS shows a decreasing or negligible trend at the East Coast stations, owing to an insignificant change in the frequency of LPS over the study period.

The interannual variability of the frequency of SL-RF and SS-RF extremes was analyzed for all the stations (Fig. 3). The East Coast stations have a significantly higher mean annual frequency of SL-RF extremes than the West Coast stations. Among all the stations, Chennai has the highest mean annual frequency of SL-RF days (4.6 days/year) followed by Paradip and Visakhapatnam. Haldia has the least mean and variance of annual frequency of SL-RF extremes among the East Coast stations. The interannual variability of SL-RF extremes (considered as one standard deviation) is higher at Chennai and Visakhapatnam than the other stations. The East Coast stations except Paradip experience a reduced frequency of SL-RF extremes in the current decades than the decades in 20th century. The significantly reduced frequency of these extremes at Visakhapatnam in the current decades is attributed to the reduced frequency of extreme rainfall during the post-monsoon (OND) season contributing to these extremes. On the other hand, the reduced frequency of ESL concurring with extreme rainfall at Haldia and Chennai during October–November in the current decade (2010–2020) led to reduced frequency of SL-RF extremes in this decade. The strikingly higher frequency of SL-RF events in 1985 at Paradip and in 1987 at Visakhapatnam and Chennai is attributed to the severe cyclonic storms with a longer lifespan of 4–5 days (cyclonic storm BOB 04 in 1985⁴⁵, cyclonic storm 04B and 06b in 1987⁴⁶) making landfall at the East Coast during October and November causing extensive torrential rain along with ESL in these regions. The West Coast stations show a negligible frequency of compound SL-RF days due to the differences in the seasonality of extreme sea level and extreme rainfall.

Each panel corresponds to a tide gauge station: (a) Haldia, (b) Paradip, (c) Visakhapatnam, (d) Chennai, (e) Kandla, (f) Mumbai, (g) Marmagao, and (h) Kochi. The bars depicting the frequency of these extremes in monsoon (JJAS), post-monsoon (OND), and pre-monsoon (MAM) are stacked to represent the annual frequency. The mean annual frequency and its uncertainty, considered as one standard deviation (SD) from the mean, are depicted by the blue and purple lines. The red line shows the trend of annual frequency of the extremes.

Similarly, for SS-RF extremes too, the stations on the East Coast have a higher mean annual frequency compared to those on the West Coast. Visakhapatnam and Marmagao experienced the highest mean annual frequency of SS-RF extremes on the East and West coasts, respectively. SS-RF extremes have increased in the recent decade (2010–2020) for the West Coast stations and decreased for the East Coast stations except Haldia. The variance in the annual frequency of SS-RF extremes on the West Coast was highest at Kandla, followed by Marmagao, Mumbai, and Kochi. The mean and high variability of the frequency of SS-RF extremes on the West Coast were significantly impacted by the multi-fold (2-3-fold) increase in the frequency in 2019 (Marmagao and Kochi) and 2020 (Kandla and Mumbai). Further, significantly high frequency of SS-RF extremes was observed at Visakhapatnam and Chennai in 2010 and 2020 respectively. The reason behind these intriguing observations is attributed to the effect of ENSO and IOD on the compound extremes, as discussed below.

The signature of ENSO and IOD on the frequency of SL-RF and SS-RF extremes was also observed at the stations located along the southern part of the Indian peninsula. A higher mean annual frequency of these extremes was observed during La Niña years followed by Neutral ENSO and El Niño years on the East Coast (Fig. 4). A significant decreasing trend in the frequency of both the SL-RF and SS-RF extremes was observed during October–December at the East Coast stations for the El Niño and neutral ENSO years. Conversely, an increasing trend was observed for the extreme SL-RF and SS-RF days for the La Niña years, even though it was not statistically significant. El Niño (La Niña) associated with weaker (stronger) Walker Circulation causes a weakening (strengthening) effect on the ISMR and tropical cyclone activity in the North Indian Ocean, thereby reducing (enhancing) the intensity and frequency of extreme rainfall and ESS along the south Indian coastline. Interestingly, however the two distinct phases of El Niño (East Pacific El Niño and Central Pacific El Niño) had contrasting impact on the cyclogenesis in the Bay of Bengal. In the El Niño years during 1980–1990 (viz.1982,1986,1987), frequency of SS-RF extremes associated with tropical cyclones witnessed on the East Indian Coast during the post-monsoon season was higher than that witnessed in the El Niño years post 1990. This can be attributed to the fact that these pre 1990 El Ninos were East Pacific El Niños where the anomalous descent in the climatological mean ascending branch of the Walker circulation in the western Pacific- tropical Indian Ocean did not have significant negative impact on the cyclogenesis potential in the southwestern Bay of Bengal^47,48 (Fig. S15). Post 1990 there were primarily central Pacific El Niño years during which the anomalous descending limb of the Walker Circulation in this region was nearby southwestern Bay of Bengal that significantly affected the Tropical Cyclone genesis in the region. As shown in Fig. S15, during East Pacific El Niño events during 1980–1190, the anomalous descending limb of the Walker Circulation extends westward up to approximately 90°E from 170⁰E, whereas during Central Pacific El Niño events, it reaches further west, extending to around 83°E. To quantify the strength of this anomalous subsidence over the equatorial Indian Ocean -southern Bay of Bengal region (80°–105°E), the vertical velocity anomalies were spatially averaged across all pressure levels within this domain. The resulting mean vertical velocity anomaly is notably higher during Central Pacific El Niño years (0.006 Pa s⁻¹) compared to East Pacific El Niño years (0.002 Pa s⁻¹), indicating stronger anomalous downward motion (subsidence) during the former. This eastward displacement and the relative weakening of the subsiding branch during East Pacific El Niño events imply a reduced suppression of convection over the southwestern Bay of Bengal near the East Indian coast. For the west coast stations, no significant trends were observed for June–September during El Niño and La Niña years, whereas a decreasing (increasing) trend in the frequency of SS-RF extremes was observed during 1980–2000 (post-2000) under neutral ENSO conditions. Further, positive (negative) IOD event associated with increased convection in the western (eastern) tropical Indian Ocean led to stronger winds and rainfall along the southwestern (southeastern) coast of India and higher frequency of SS-RF extremes. The cooccurrence of Positive IOD with both types of El Nino has further suppressing effect on the cyclone assocuated SS-RF extremes on East Coast of India.

i shows the frequency of compound events during the post-monsoon season (OND: October–December) along the East Coast under (a) El Niño, (b) La Niña, and (c) Neutral ENSO conditions. ii shows the same during the monsoon season (JJAS: June–September) along the west coast for the same ENSO phases. The orange (blue) bars depict the frequency of SS-F(SL-RF) extremes. The extreme IOD (positive IOD (+IOD) and negative IOD (-IOD)) years are annotated in each plot. The mean, standard deviation (SD), and statistically significant trend in the frequency of SS-RF (SLRF) extremes are annotated in orange (blue). The red (blue) dotted line shows the trend in the cumulative frequency of SS-RF (SL-RF) extremes, whose value is given wherever it is significant at the 10% significance level.

The high frequency of SS-RF extremes observed at Marmagao and Kochi between August and October in 2019 is attributable to strengthened monsoon circulation caused by the strongest recorded positive IOD event⁴⁹ (Fig. 4). The Walker circulation associated with the El Niño Modoki event witnessed in the previous year aided in strengthening the positive IOD event in the Indian Ocean⁵⁰. The extremes in Mumbai and Kandla during 2020 were primarily observed in August that year. A stronger monsoon circulation and offshore monsoon trough, along with cyclonic vortices in the northern Arabian Sea led to those extremes at the stations (Figs. 5c, d, S13). Five well-identified monsoon LPS occurred in the Bay of Bengal during this season, and their west north-westward movement also contributed to widespread rainfall in Kandla and Mumbai (Fig. S11) during the first and last week of the month.

a, b correspond to Kochi and Marmagao for the year 2019, while panels c–f correspond to Kandla, Mumbai, Kochi, and Chennai, respectively, for the year 2020. The maximum value of the mean wind speed anomaly is shown at the top right corner of each plot with a reference vector. The green dot shows the location of the tide gauge station.

The east coast stations Visakhapatnam and Chennai experienced a significantly high frequency of compound extremes in the post-monsoon periods of La Niña years 2010 and 2020. In 2010, a deep depression intensified into a severe cyclonic storm (Jal) during late October-early November, leading to SS-RF extremes at these stations. The extremes at Chennai and Visakhapatnam in the post-monsoon season of 2020 occurred primarily due to very severe cyclonic storm Nivar in the southwest Bay of Bengal during the last week of November and due to cyclonic storm Burevi witnessed in the equatorial Indian Ocean¹². Figure 5f shows the circulation pattern near the Chennai coast, indicating high vorticity during this period.

Interannual variability in the intensity of extremes

The annual mean intensity plot of extreme rainfall, ESL, and ESS are shown for all the stations (Figs. S8 and S9). The West Coast stations exhibit a significantly lower intra-annual and interannual variability in ESL and ESS magnitude compared to the east coast stations. The relatively stronger coastal currents during post monsoon along the east coast could be a reason for the higher ESL variability along the east coast⁴³. Further, the stations at the northern part of the Indian peninsula exhibit higher variability in ESL and ESS than at the southern part due to a wider continental shelf and shallower continental slope in this region, leading to amplified tidal ranges and storm surges. Concerning extreme rainfall, Mumbai and Marmagao exhibit higher annual mean intensities and significantly higher variability than other stations which could be largely attributed to the variability in the Indian Summer Monsoon Rainfall (ISMR) (Fig. S4). The interannual variability in the intensity of extreme rainfall, ESL and ESS at some coastal stations is also impacted by the effect of ENSO on ISMR^34,43,51.

The interannual variability in the annual mean intensity of ESL (ESS) and cooccurring extreme rainfall were analyzed for SL-RF (SS-RF) extremes at all the stations (Fig. 6). Among the East Coast stations, the annual mean intensity of ESL, ESS and its uncertainty were highest for Haldia and relatively lower for Paradip, Visakhapatnam, and Chennai (in that order). No significant trends were observed in the annual mean intensity of ESL and ESS for any of the East Coast stations. However, a change point was noted in 2005 for the time series of the annual mean intensity of ESL at Chennai (Fig. 6d). A decreasing (increasing) trend was observed before (after) the change point. The ESL at Haldia is highly influenced by discharge from the rivers Ganga and Brahmaputra along with a high tidal range which influences it interannual variability. The abnormally high ESL and ESS in the SL-RF and SS-RF event of 1985 at this station is attributable to the combined effect of spring tide and cyclonic storm (Tropical Cyclone 04B) witnessed during 12–17 October in the northwestern Bay of Bengal⁴⁵. The swell waves from this storm also had a signature in the ESL at Paradip. On the other hand, the high peak intensity of SL-RF event at Haldia during 1997 was due to the cooccurrence of a deep depression and a spring tide along with the east-northeast flowing longshore coastal current⁵². Further, the annual mean intensity of ESS at Haldia was higher in 2017 and 2019, which could be attributed to cyclones that made landfall on the West Bengal coast in post-monsoon (very severe cyclonic storm Ockhi) and pre-monsoon (very severe cyclone Fani) seasons, respectively. The variability and uncertainty of annual mean intensity of extreme rainfall for the SL-RF and SS-RF events was considerable for all the East Coast stations. Abnormally high values of rainfall were observed in Haldia in September 1986 due to a deluge³⁰ in that region and in June 1982 at Paradip, corresponding to cyclone BOB 02⁵¹. The extreme rainfall showed a significant decreasing trend at Paradip (−2.05 mm/yr) and Visakhapatnam (−6.3 mm/yr from 1980 to 2000) corresponding to SL-RF event a decreasing trend at Haldia (−2.3 mm/year) corresponding to SS-RF events. For the West Coast stations, the long-term average of annual mean extreme rainfall intensity corresponding to SS-RF extremes was higher at Mumbai and Marmagao compared to Kandla and Kochi by 2 folds. Further, this extreme rainfall also exhibited high interannual variability in Mumbai along with a significant increasing trend.

Each row corresponds to a tide gauge station: (a) Haldia, (b) Paradip, (c) Visakhapatnam, (d) Chennai, (e) Kandla, (f) Mumbai, (g) Marmagao, and (h) Kochi. In each row, the first plot shows the intensity of ESL, the second plot shows the intensity of ERF during SL-RF events, the third plot shows the intensity of ESS, and the fourth plot shows the intensity of ERF during SS-RF events. Each green dot represents the annual mean intensity of a variable (ESL, ESS, ERF) and the blue bar shows its uncertainty, which is considered as one standard deviation (SD) from the mean value. The black dotted line shows the trend in the annual mean intensity, whose value is given wherever it is significant at the 10% significance level.

Considerable difference in the mean and variance of ESS and extreme rainfall intensity was observed between the El Niño and the La Niña phases at some of the coastal stations. The means of ESS and extreme rainfall intensity at the East Coast stations were higher during October–December than June–September for both the El Niño and La Niña years. In addition, higher mean intensity of extreme rainfall and ESS were observed during negative (positive) IOD in the East (West) Coast stations during late monsoon and post-monsoon (Fig. S11). During the El Niño years of 1982 and 1986, higher mean intensity of ESS (0.5–0.6 m) and extreme rainfall (~200 mm) corresponding to SS-RF extremes were observed at Paradip and Haldia, respectively, during both monsoon (Fig. 7-I) and post-monsoon (Fig. 7-II) seasons. This is attributed to the severe cyclonic storm (BOB 02) in 1982 at Paradip⁵³ and the deep depression in Haldia in September 1986³⁰. The high means of TCRW anomalies (1.2–1.4 kg/m²) and 850 hPa wind speeds in the region encompassing Haldia and Paradip during El Niño periods also illustrate the synoptic pattern associated with these events (Fig. 8(i)). Further, during the post-monsoon season of East Pacific El Nino years 1982, 1986, 1987 seven TCs occurred in the southwestern BoB leading to SS-RF extremes at the East Coast stations Visakhapatnam and Chennai1^45,46,52,. Stronger wind gusts with higher positive vorticity were also observed near those stations during the occurrence of the extremes (Fig. 8(i)-g, j). Reduction in vertical wind shear and simultaneous increase in mid-tropospheric relative humidity favored the intensification of low-pressure systems to cyclones in the Bay of Bengal during these East Pacific El Nino years, causing intense SS-RF extremes in these years⁴⁷. During 1982 and 1986, SS-RF extremes were associated with low-latitude cyclones. In contrast, the increased occurrence of SS-RF extremes in 1987 at the East Coast stations (Visakhapatnam and Chennai) corresponded to cyclones forming between 10°–20°N. The significant positive anomalies in mid-tropospheric relative humidity in these regions support enhanced cyclogenesis (Fig. 9). The mean TCRW anomalies associated with SS-RF extremes at Haldia, Paradip and Visakhapatnam did not vary significantly between the La Nina and Neutral phases. Unexpectedly, Chennai experienced a higher TCRW anomaly associated with cyclones and deep depressions during La Niña in 2020, largely contributing to the increased frequency of SS-RF extremes in this year (Fig. 8-i(k)).

a–d show ESS (mean), ESS (SD), ERF (mean), and ERF (SD), respectively, for each phase. If no (one) SS-RF event is witnessed at a station, the mean (SD) is represented by gray dot.

Three panels are shown for each of the East Coast stations Haldia (a–c), Paradip (d–f), Visakhapatnam (g–i) and Chennai (j–l) which correspond to El Niño, La Niña and Neutral ENSO phases during the post-monsoon season. Similarly, three panels are shown for the West Coast stations Kandla (a–c), Mumbai (d–f), Marmagao (g–i) and Kandla (j–l) corresponding to the three ENSO phases during the monsoon season. Blue and red shading indicate negative and positive TCRW anomalies, respectively. The maximum value of the mean wind speed anomaly is shown at the top right corner of each plot with a reference vector.

a–c correspond to the years 1982, 1986, and 1987, respectively. Blue shading indicates negative anomalies, while red shading denotes positive anomalies. Anomalies are calculated with respect to the climatological mean over the period 1980–2020.

Among the west coast stations, Kandla experienced a mean intensity of 0.3–0.4 m ESS in monsoon during the La Niña years whereas no ESS was observed during EL Niño (Fig. 7-I). The SS-RF extremes at Kandla in this season are attributable to the monsoon trough (elongated band of high TCRW anomalies and low pressure), causing extensive rainfall and storm surge in this region (Fig. 8-ii(a–c)). The coastal stations of Mumbai, Marmagao and Kochi witnessed these extremes primarily during La Nina and Neutral ENSO years which were attributable to the offshore trough along the west coast of India that transports moisture-laden winds from the Arabian Sea thereby causing heavy rainfall and storm surges at these coastal stations. Higher TCRW anomalies along the coast and low-level jets towards the coast in Fig. 8-ii illustrated these synoptic patterns. However, noticeably higher extreme rainfall intensities corresponding to SS-RF extremes was observed during the year 1997 at Kandla and in 2015 (El Nino) at Mumbai which was due to the formation of deep depression near these coastal stations in June of the respective years. In a nutshell, La Niña and neutral phases of ENSO were associated with higher intensity of low-level (850 hPa) monsoon winds and rainfall for the majority of SS-RF extremes at the west coast stations during June–September.

Risk analysis of compound extremes

The return period isolines/contours of SS-RF extreme events corresponding to bivariate AND return period are shown for coastal stations (Fig. 10). At Haldia, Paradip, and Kochi the return period of a historical event exceeded 100 years. The event at Haldia occurred in September 1986 due to a depression in the Bay of Bengal that interacted with a cloud cluster over the Bay of Bengal and Burma and a northward propagating monsoon trough, leading to a deluge for three consecutive days with total rainfall exceeding 400 mm. The event at Paradip occurred (on the 3rd of June,1982) due to the landfall of an extremely severe cyclonic storm that originated from the Bay of Bengal. The rainfall and surge height corresponding to this event were 380 mm and 0.8 m, respectively. At Kochi, the compound event corresponded to floods witnessed in August 2018. Its cause could be attributed to highly anomalous rainfall due to high-frequency mixed Rossby waves in the mid-troposphere triggered by synoptic disturbances from the tropical Pacific⁵⁴. The risk of compound flooding could be quantified by the number of SS-RF extreme events exceeding 25- and 50-year return periods. It is highest at Mumbai and Haldia (6 events) and relatively lower at Paradip, Kochi, and Kandla (4 events). The most likely SS-RF extreme events exceeding 25-year return period are primarily surge dominant at the east coast stations, whereas those at the west coast stations (except Mumbai) are rainfall dominant. The rainfall dominant extremes are primarily observed during monsoon when synoptic systems of moderate intensity (depressions and monsoon LPS) drive the SS-RF extremes on the coast. On the other hand, the surge dominant extremes are primarily driven by intense LPS (Tropical cyclones) on the coast. Among SS-RF events with a return period exceeding 25 years, the highest intensity of rainfall (storm surge) was witnessed at Chennai (Haldia).

a–h correspond to the stations at Haldia, Paradip, Visakhapatnam, Chennai, Kandla, Mumbai, Marmagao, and Kochi, respectively. The black dots represent the historical events. Each contour (isoline) corresponds to a return period in years.