The Antarctic Bottom Water (AABW) forms in the Southern Ocean which is the coldest and high saline water mass that travels throughout the global ocean. Characterization of the AABW is not always consistently followed considering the variety of its source region. The major known source regions of AABW are the Weddell Sea, the Cape Darnley polynya, the Ross Sea, and the Adélie Coast (Rintoul, 2007; Ohshima et al., 2013). The physical characteristics of AABW vary according to their formation region. The bottom water in the Ross Sea is characterized by a potential temperature (θ) from −0.3 to −0.6°C and salinity of more than 34.72. The coldest and freshest variety of bottom water forms along the Weddell Sea (Orsi et al., 1999). The type of AABW along the Adélie coast is characterized by low saline-cold water with high oxygen (van Wijk and Rintoul, 2014). The bottom water observed in the Australian-Antarctic Basin (AAB) and the Princess Elizabeth Trough (PET) is sourced and advected from its eastern region (Rintoul, 2007; van Wijk and Rintoul, 2014). The advection of high dense bottom water occurs by the boundary currents from the Pacific to the Indian sector of the Southern Ocean (ISSO) (Orsi et al., 1999; Aoki et al., 2008; van Wijk and Rintoul, 2014). Advection of slightly less dense AABW with a neutral density of more than 28.27 kg m^–3 also occurs from the western region that is sourced from the Weddell-Enderby Basin (Orsi et al., 1999).

Mixing of various water masses such as the dense shelf water (DSW) and Circumpolar Deep Water (CDW) influences the formation of AABW (Orsi et al., 2002). The dominant production of AABW occurs in the coastal polynyas through increased sea-ice production and salt ejection associated with various dynamic and thermodynamic processes. For example, few of the known sources are reported in Cape Darnley, Adélie Land, and Vincennes Bay (Ohshima et al., 2013; Rintoul, 2013; Kitade et al., 2014). More recently, the formation of AABW in the Ross Sea has been linked to sea-ice changes through various processes associated with different modes of climatic variability such as Southern Annular Mode (SAM) and El Niño–Southern Oscillation (ENSO) (Silvano et al., 2020).

Many investigations revealed the asymmetrical modifications of AABW due to the influence of the formation of polynyas and enhanced melting in the glaciers. The variability in the fresh water budget along the Antarctica coast influences the formation of dense water and the signals get propagated into the deep ocean through ventilation of AABW (Rintoul, 2007). The global-scale contraction of AABW is reported by Purkey and Johnson (2012). The variations in the characteristics and volume of AABW are attributed to the freshening of source waters, not to the disparities in the formation rate alone (Johnson et al., 2014; van Wijk and Rintoul, 2014). A decline of AABW production in the Cape Darnley region due to enhanced ice shelf melting appears to be linked with global climate warming. Increased oceanic heat input shall lead to the irrevocable loss of ice sheets, intensified ocean/ice shelf interaction, and may further decrease the production of AABW in the future (Williams et al., 2016). Rapid collision of iceberg glaciers and associated calving events at the bottom water source regions may lead to enhanced freshening that results in the formation and spreading of fresh and less dense AABW (Menezes et al., 2017). Enhanced advection of CDW toward the ice shelf also leads to the decline of the shelf water density and subsequently influences the formation of AABW (Guo et al., 2019). The modifications of the AABW in the AAB in a decadal time scale showed asymmetrical variations (Aoki et al., 2020b).

In the present investigation, we focus on the repetitive profiling made at four stations between 55 and 64°S along the meridional transect of 57°30’E in the ISSO (Figure 1). The observed freshening and shrinkage of the AABW layer shall be linked to the alteration in characteristics or a variation in the formation rate of the source waters. Furthermore, we discussed the freshening and subduction of the AABW layer during the past decade. We linked these modifications of AABW with the shoaling and intrusion of CDW in the Prydz Bay region, variations in the wind, sea-ice, and calving events.

Temperature and salinity data collected from the various Southern Ocean expeditions (SOE) during SOE 2010 (January 31, 2010, to March 1, 2010), SOE 2011 (February 1, 2011, to March 2, 2011), SOE 2017 (January 6 to February 28, 2017), SOE 2018 (December 8, 2017, to February 4, 2018), and SOE 2020 (January 22 to March 8, 2020) were used to characterize the AABW and its modification (Supplementary Table 1). Except for SOE 2010, the Conductivity, Temperature, Depth (CTD) instrument used during the cruises was SBE 911plus, with the temperature and conductivity accuracy of ± 0.001°C and ± 0.0001 Sm^–1, respectively. During SOE 2010, a portable SBE 19plus CTD was used to collect the temperature and salinity data. The location and CTD model used during different cruises are given in Supplementary Table 1. The CTD data were processed using the standard protocol of manufacture supplied software downcast data. The temperature (ITS90) and salinity (PSS78) data from CTD were later converted, respectively, to conservative temperature (θ)-absolute salinity (S_A) using the Gibbs-Sea-Water toolbox¹. Field calibration of CTD salinity data was carried out against the laboratory determination of salinity using a salinometer (Guideline 8400B). We have used the underway conductivity temperature depth (UCTD) (Rudnick and Klinke, 2007) during December 2017 (referred to as austral summer 2018) and the expendable conductivity temperature depth (XCTD) during January 2017 for the collection of temperature and salinity along a section from 60 to 67°S between 70 and 74°E. The XCTD profiles were quality controlled by following the guidelines in the Uchida et al. (2011). The temperature accuracy of XCTD and UCTD probes is ± 0.02°C. We have used these probes in the shallow depth regions where the range of variability is much larger (>0.5°C) than the accuracy of these probes. Additionally, we have used World Ocean Circulation Experiment (WOCE) data at 62.33°S 60°E and 64.33°S 60°E from the WOCE S041 line 2006 for our study. The instrumented seal data set (Treasure et al., 2017) of the Marine Mammals from the shelf regions off Cape Darnley for the period 2015–2017 has been used to check the variability of the CDW in recent years.

The sea-ice concentration (SIC) and area from multiple passive microwave satellite observations were acquired from the National Snow and Ice Data Center (NSIDC) (data ID G0192, version 3). The brightness temperature data for different microwave bands were converted to SIC using the National Aeronautics and Space Administration (NASA) Team algorithm with a spatial resolution of a 25 km grid. The satellite sensor characteristics, sea-ice retrieval, and validation of passive microwave sensors are well described (Fetterer et al., 2016). The time series of satellite-based sea-ice data used in this study spanned from 1979 to 2020. To understand the recent variability of AABW in the ISSO, we have considered ECMWF ERA-5 based sea-ice and wind observations in and around the polynyas located along the Weddell Sea, ISSO, Western Pacific, and the Ross Sea. The polynya locations were identified from sea-ice maps and the nomenclatures have been referred to using earlier studies (Nihashi and Ohshima, 2015; Jena et al., 2019). The different seasons are defined as follows: winter (June, July, and August) and summer (December, January, and February). The seasonal variations of sea-ice were computed by averaging the corresponding monthly mean data. The seasonal computation of the SAM index was carried out using the monthly mean observational SAM index (Marshall, 2003).

Freshening (∼0.002) and warming (∼0.04°C) were shown by AABW toward the end of the last decade in the ISSO. In this study, we encompass the modifications of AABW off the Enderby basin to the last decade and consider changes in temperature, salinity, and AABW layer thickness. By using measurements from the same location and season, we minimize the discrepancies in spatial and seasonal variability. The previous study over the same region reported a fresher (∼0.01) and warmer (0.05°C) AABW in 2010 compared to 2006 (Anilkumar et al., 2015). However, the latest observations during 2020 exhibit moderate modifications in the properties of AABW compared to the previous decade in the ISSO. The earlier studies in the AAB reported that the cause for the freshening of bottom waters in this basin may be due to the freshening of the source of the dense water (van Wijk and Rintoul, 2014). The prevalent warming, contraction of bottom waters and variations in the mixing as well as stratification detected downstream are attributed to the freshening of the sources (Purkey and Johnson, 2010; Shimada et al., 2012). During the last decade, we have observed a 50 to 120 m reduction in the layer thickness of AABW (Figure 3A). A descending rate concurrent to a previous study by Purkey and Johnson (2012) that showed a decrease of 81 m per decade for AABW with θ ≤ 0°C, was divulged in the present study.

Instead of a decline in the formation rate of Weddell Sea Bottom Water (WSBW), the robust incursion of older Warm Deep Water rather was reported by Heuzé et al. (2013). Between 1984 and 2008, the reduction in the area of cross-section of the WSBW at the Prime Meridian caused variations in the Weddell gyre wind-driven circulation instead of its formation rate (Fahrbach et al., 2011; van Wijk and Rintoul, 2014). These studies have suggested that in the AAB, the variations observed along the AABW export pathway are attributed to the formation of lighter varieties of source waters and the contraction of AABW is compensated by expansion of lower and upper CDW and surface waters. Another reason for the observed freshening of AABW could be due to the anomalous decline in sea-ice cover during recent years. Silvano et al. (2020) have reported the regaining of the properties in the salinity, density, and thickness of AABW in the Ross Sea in 2018–2019, similar to those observed in the 1990s, due to the sea ice formation caused by the rare unification of SAM and El Niño conditions between 2015 and 2018. This favors the formation of AABW.

One of the major events we would like to discuss in this study might have an influence on the modification of AABW observed during the present study. For example, the calving event at the Mertz Glacier Tongue (MGT) during January-February 2010, influenced the AABW formation in coastal polynyas over the Mertz Depression and Adélie Depression located east and west of the MGT (Williams et al., 2010). This reduced sea-ice production up to 20% and the outflow of DSWs to nearly 23% (Kusahara et al., 2011; Tamura et al., 2012). The MGT calving resulted in a decrease in salinity (∼0.15) of DSWs from 2008 to 2012. This high rate of freshening is compatible with 50 years of freshening observed in the Ross Sea (Shadwick et al., 2013). The effect of the MGT calving event through various oceanic processes influences the AABW on the decadal time scale (Kusahara et al., 2011).

The time span required for AABW to get transported from its formation area to the present study region is unknown. The speed of the mean abyssal currents carrying the freshening signal would be higher than 2 cm s^–1 as reported in earlier studies (Menezes et al., 2017). Strong boundary current transporting AABW westward was reported in this region (Orsi et al., 1999). The MGT calving event occurred in 2010, however, the freshening signal was observed during 2020 in the present study region (Figure 2). Considering the distance between Adélie coast (146°E; 67°S) and 64°S 57.5°E, the required traveling time of AABW was computed. We suggest a meager probability to link the MGT calving in 2010 with the freshening observed during 2020. Due to this event, unusual hastening in bottom water freshening was reported during 2016 (Menezes et al., 2017).

Williams et al. (2016) and Silvano et al. (2018) have reported the freshening of AABW due to the influence of ice shelf melting. Kusahara et al. (2011) has suggested that the effect of the calving event at the MGT influences the AABW on a decadal time scale through various oceanic processes. However, our study area is away from the abovementioned regions, and hence, actual cause for the changes observed for AABW shall be difficult to be traced. Hence, we suggest that the influence of the AABW formed in the Cape Darnley and the calving event at MGT might have also triggered the changes of AABW observed in the present study region. Nevertheless, Aoki et al. (2020a) have reported that the recently revealed bottom water off Cape Darnley is transporting an enhanced quantity of freshwater and feeding the surplus freshwater into the abyssal Weddell Sea. This suggests that the bottom water originated from Cape Darnley has influenced the changes experienced at the Weddell Sea. Hence, one of the causes for the changes observed with the AABW in our study area could be due to the influence of the AABW transported from the Cape Darnley region. Furthermore from 1974 to 2016, freshening of the bottom water closer to the present study region (65°S 61°E; 64°S 59°E and 58°S 53E) were also reported by the previous investigators (Aoki et al., 2020a).

We reported upward tilting of CDW and its onshore intrusions in the Prydz Bay, causing the freshening and subduction of AABW, possibly triggered by the weakening of easterly winds between 2016 and 2020. Our findings on modifications of AABW properties due to the penetration of warm CDW into the Prydz Bay region are supported by the results from earlier studies (Rintoul, 2007; Purkey and Johnson, 2013). The Cape Darnley polynya was identified as one of the origins of AABW contributing around 13–30% of bottom water for the Atlantic sector (Williams et al., 2016; Aoki et al., 2020a). Vibrant warming of CDW was encountered in the Cape Darnley region as portrayed in Figure 3D. This transportation of AABW occurs through the Indian sector, hence we suggest that one of the sources for the modifications of AABW observed in the present study region is likely from Cape Darnley origin.

Our study reveals that the AABW became fresher (∼0.002), warmer (0.04°C), and subducted by ∼50–120 m toward the end of the past decade in the ISSO. Various reasons discussed for the freshening and subduction of AABW are the upward tilting of CDW and its onshore intrusions in the Prydz Bay, possibly induced by the weakening of easterly winds between 2016 and 2020. The CDW experienced warming of ∼0.3°C and its onshore intrusion likely enhances melting of the ice shelves. Freshwater discharge from the enhanced melting of ice shelves induced by the onshore intrusion of warm CDW, ocean warming, and calving events likely influenced the freshening and contraction of AABW. This result postulates that more observations are required for a comprehensive understanding of the modifications of AABW properties. The latest initiation such as Deep Argo technology has facilitated bottom water sampling at a higher temporal resolution (Kobayashi, 2018). The amalgamations of sustained Eulerian and Lagrangian data collections, as well as modeling studies, are the more reliable methods for illustrating the long-term differential modifications of the AABW.

The in situ data collected during the Indian Scientific Expeditions to the Southern Ocean are available at https://data.mendeley.com/datasets/vtm8wh3933/1. Mendeley Data, V1, doi: 10.17632/vtm8wh3933.1.

NA formulated the idea and led the study and wrote the manuscript. All other authors contributed to the writing, review, and editing of the manuscripts and carried out data analysis. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.