Climate–tectonic interactions in the eastern Arabian Sea

Potential interactions between the solid Earth and the climate system have long been recognized by geoscientists. Such links have been the subject of both modelling and observational studies that have attempted to assess the degree and nature of any coupling. The uplift of high mountains and the closure of ocean gateways disrupt circulation patterns in the atmosphere and oceans, and thereby affect the transfer of heat from low latitudes to the poles (Borrelli et al. 2014; Korte et al. 2015; Tada et al. 2016). It has been suggested that uplift of mountains also causes long-term cooling as a result of increased chemical weathering of silicate minerals that draws down concentrations of CO₂, a greenhouse gas, in the atmosphere (Raymo & Ruddiman 1992). Rapid sedimentation may further sequester organic carbon in major depocentres where it remains, at least until it is subducted or uplifted in a new phase of orogeny (Galy et al. 2007). In turn, the atmosphere and hydrosphere influence the solid Earth, largely through the action of surface processes that cause erosion and exhumation and consequentially help to define the structure of mountain belts (Koons et al. 2003; Sinclair et al. 2005; Whipple 2009). No system is believed to typify these interactions better than the Asian monsoon and the building of the Himalaya, Tibetan Plateau and associated ranges (Prell & Kutzbach 1992; Molnar et al. 1993; Clift & Webb 2019).

This special issue of Geological Magazine is designed to showcase results of research undertaken after International Ocean Discovery Program (IODP) Expedition 355 in the spring of 2015 (Pandey et al. 2016c). This research effort involved deep coring of the sediments deposited in the Laxmi Basin of the eastern Arabian Sea by the research vessel JOIDES Resolution. The objective was to recover a record of evolving erosion, weathering and continental climate linked to the Indus Basin, as well as the oceanography of the Arabian Sea, which has long been recognized to be closely linked to the intensity of the South Asian monsoon (Kroon et al. 1991; Prell et al. 1992). A secondary objective was to better understand the tectonics of the Laxmi Basin itself and assess its role in the opening of the Indian Ocean. While some results have been published elsewhere, this special issue represents the biggest single collection of work completed by the science party, exploiting the unprecedented long-duration sedimentary records that are required to examine how climate and tectonics may have co-varied in the western Himalayas since the start of the India–Asia collision, likely at c. 60–50 Ma (Najman et al. 2010; DeCelles et al. 2014; Wu et al. 2014). The emplacement of a large mass-transport complex (MTC) into the Laxmi Basin during late Miocene time (Calvès et al. 2015; Dailey et al. 2020) means that much of the earlier history remains unknown either because of erosion by the MTC, or because of drilling difficulties caused by the rocks within the MTC (Pandey et al. 2016a, b). Nonetheless, a record of evolving climate and erosion dating from c. 10.8 Ma to the present represents a major improvement in our understanding of evolving climate, erosion and tectonics in SW Asia. Coring was undertaken at two sites. Site U1456 is located in the middle of the Laxmi Basin, whereas Site U1457 is located on the western side of the basin adjacent to Laxmi Ridge; the latter was specifically positioned to recover samples from the igneous basement, which was successful (Pandey et al. 2016c).

Reconstructions of the palaeoceanography in the western Indian Ocean represented some of the first and most important constraints on the varying intensity of the South Asian monsoon (Kroon et al. 1991; Prell et al. 1992). In the west, summer monsoon winds cause coastal upwelling of nutrient-rich deep water offshore Oman, resulting in blooms of plankton (Curry et al. 1992). A number of proxies have now been applied to the newly recovered sediment in order to understand how the palaeoceanography in the eastern Arabian Sea also varied. Satpathy et al. (2019) examined upper water column dynamics using an assemblage of records based on oxygen and carbon isotopic variations in foraminifers, as well as a Mg/Ca-based sea-surface temperature reconstruction. Their work indicates two discrete intervals of monsoon-related change in the upper water column at 2.7–1.85 Ma and 1.65–1.55 Ma. They suggest that the upper water column was more stratified between 2.7 and 1.85 Ma as a result of weak Asian monsoon circulation. The 1.65–1.55 Ma interval was instead related to strong winter monsoon winds. The study identified long-term changes in monsoon intensity not related to glacial cycles and in particular to a strengthening of the winter monsoon between 1.85 and 1.65 Ma.

Saraswat et al. (2019) looked at stable oxygen isotope and assemblage data of foraminifers spanning the last 350 ka. An increasing abundance of the foraminifer Globigerina bulloides is used to infer greater upwelling during Marine Isotope Stage (MIS) 7 (243–191 ka) and the beginning of MIS 6 (191–130 Ma) in the eastern Arabian Sea caused by a stronger South Asian monsoon. This is important because there was a significant decrease in upwelling at the same time in the western Arabian Sea, identified by the same G. bulloides proxy. Decreased upwelling and increased winter convection occurred during the remainder of MIS 6, as well as in the subsequent glacial interval (MIS 4–2), whereas conditions during interglacial MIS 5 were similar to today. The coherent response on the western and eastern sides of the Arabian Sea indicates that the northern part of the Arabian Sea at least responds as a single unit to glacial–interglacial changes. Contrast between the oxygen isotope (δ¹⁸O) values seen in the glacial versus interglacial times shows that there was a maximum salinity contrast at glacial Termination II (c. 130 ka) and III (c. 240 ka), implying a major change in monsoon strength, but a very small difference at Termination IV (c. 330 ka), suggesting little change in the monsoon strength at that time.

All of the proxy records in the Laxmi Basin are reliant on good age control and, in their study, Routledge et al. (2019) revised the age framework of the cored sediments using a combination of biostratigraphy, magnetostratigraphy and strontium isotopes. In doing so they confirmed the presence of four unconformities at both sites that disrupt the continuous record desired. Their new age model also better constrained the age of emplacement of the MTC to between 9.83 and 9.69 Ma, significantly younger than previously understood. The oldest unconformity is at the base of the MTC and is of very different duration at each site. At Site U1456 in the middle of Laxmi Basin, sediment underlying the MTC is of early–middle Miocene age, whereas at Site U1457 the sediment below the MTC is dated of early Paleocene age. The unit overlying the MTC is dominated by hemipelagic sediment and thin sandstones, and is separated by a second unconformity from the overlying unit with c. 0.5–0.9 Ma of missing section. The upper Miocene interval deposited between c. 8.5 and 5.5 Ma is dominated by mudstone with frequent thin sand beds in the lower part and a hemipelagic interval of chalk deposited between c. 8 and 6 Ma. A third longer unconformity of 1.4–1.6 Ma duration spans the Miocene–Pliocene boundary. Siliciclastic sedimentation resumed during the Pliocene – early Pleistocene period, interrupted by a short hiatus of c. 0.45 Ma duration. Above that unconformity the succession consists of a rapidly deposited series of Pleistocene sand and mud, deposited in the Laxmi Basin between c. 1.8 and 1.2 Ma. Since 1.2 Ma, hemipelagic sedimentation has dominated Laxmi Basin.

Geochemical logs have become a critical part of environmental reconstructions based on piston core material, and in particular X-ray fluorescence (XRF) core scanning data are now routinely used to illuminate variations in chemical weathering and/or provenance. These data are often calibrated with the major- and trace-element compositions of discrete samples obtained using conventional XRF or inductively coupled plasma mass spectrometry techniques. In their study, Hahn et al. (2019) compared the performance of data collected by a handheld XRF to those measured using a conventional (benchtop) XRF. They demonstrated highly significant correlations between the handheld and conventional XRF measurements of discrete samples (R ² ≥ 0.98) for all elements examined, including ratios typically used for environmental interpretations such as Al/Ca, Ti/Ca, Mn/Ca and Fe/Ca, indicating that the handheld device yielded high-quality and easily generated results appropriate for the calibration of core scanning data. The study concluded that the handheld device can play a critical role in future high-resolution geochemical studies.

The flux of eroded material from mountain sources in the Himalaya to the bottom of the Arabian Sea is controlled over long timescales by sea level as well as tectonics and climatically modulated sediment supply. In their study, Kumar et al. (2019) looked at the effect of sea level and climate on clastic sediment grain size, a proxy that is potentially affected by a number of possible processes. Data were collected from 203 samples spanning the last 200 ka, and an end-member modelling technique was used to look for coherent patterns. The work identified a period between 200 and 130 ka when there was an increase in the amount of coarse silt deposited. The study concluded that this increase was the result of reworking on the Indus Fan during the development of deep-sea canyons linked to a drop in sea level that affected sediment supply from both the Indus and Narmada rivers, which drain peninsular India. In general, lower sea level caused more incision and coarser sediment sizes. Since 130 ka, sedimentation has been more clay-dominated as a result of higher sea levels and warmer climate.

It has long been believed that there was an important climatic transition in SW Asia after 8 Ma, as initially identified from oxygen and carbon isotope studies of the foreland basin (Quade et al. 1989; Singh et al. 2011), as well as upwelling along the Oman margin (Kroon et al. 1991; Gupta et al. 2015). In their study, Clift et al. (2019a) used a combination of core scanning XRF data and colour spectral records to examine evidence for changing chemical weathering and humidity during the late Miocene Epoch. They identified increasing amounts of hematite relative to goethite, which suggests long-term drying after 7.7 Ma. Times of dry climate were associated with weaker chemical alteration as measured by K/Rb records, as well as reduced coarse clastic flux. The study also identified an interval of increased humidity between 6.3 and 5.95 Ma when there was stronger weathering and erosion. The trends in relative aridity do not coincide with those identified in palaeoceanographic records, and cannot be correlated with changes in the elevation of the mountains in the hinterland. A better correlation is seen with trends in global cooling (Zachos et al. 2001). In this respect, trends in humidity in the Indus Basin are opposed to those seen in southern China, a feature that is attributed to the contrasting topography and generally wetter climate seen in SE Asia.

Andò et al. (2019) conducted a multidisciplinary study of sediments in the Indus Fan dating from late Pliocene to early Pleistocene time. They examined the heavy mineral assemblages and showed that Indus Fan sediments are relatively rich in heavy minerals. The concentration of these phases was higher in sediments associated with channel fill rather than in overbank deposits, reflecting the fast settling rate of very dense minerals such as zircon. In contrast, the overbank deposits are richer in phyllosilicates, which have a platy morphology. This work was coupled with biomarker analysis that explored the provenance of the clay-sized fraction. This method showed that organic material at Site U1456 was largely derived from terrestrial sources, whereas at Site U1457 it was more dominated by open-marine organic material. Biomarker results also indicate low thermal maturity at both sites, with Site U1456 showing slightly higher values, although these results also indicate a somewhat higher geothermal gradient than expected based on in situ formation temperature measurements made during the expedition.

Further investigation of the thermal history of the basin was conducted using apatite fission track (AFT) data from both drilling sites (Zhou et al. 2019). Because the AFT ages are older than the depositional ages, it is clear that they were not reset after sedimentation and so reflect cooling of the source terrains as a result of erosion. Comparing the minimum cooling age of each sample with its depositional age showed that there was an acceleration in exhumation of the source bedrocks between 7.8 and 7.0 Ma. Lag times, which were c. 6.0 Ma between 8.5 and 7.8 Ma, were reduced to close to zero between 7.0 and 5.7 Ma. This period of rapid exhumation most likely reflects faster erosion in the Karakoram and to a lesser extent the Himalayas, a conclusion based on comparison with detrital zircon data (Clift et al. 2019b). The increased erosion rate in the source areas correlates with a period of climatic drying and not a stronger summer monsoon as might be expected. Following the transition, there was a resumption of slower exhumation with lag times of c. 4.5 Ma after 5.7 Ma.

The nature of vegetation in the Indus Basin is closely linked to the intensity of the summer monsoon precipitation. In their study, Suzuki et al. (2019) analysed the carbon isotope composition of fatty acids in sediments from Site U1457. They showed a shift in the δ¹³C signature of C₃₂ fatty acids towards less negative values between 10 and 6.3 Ma. In contrast, C₂₄ fatty acids did not show this change, and these shorter molecules are interpreted to represent the contribution from freshwater aquatic C3 plants. After 5.8 Ma, the δ¹³C remained relatively constant but the average chain length varied. The authors interpreted that, prior to 6.3 Ma, the fatty acid signal represents variations in the abundance of terrestrial C3 plants to aquatic (freshwater) C3 and (terrestrial) C4 plants, whereas after 5.8 Ma, the signal is dominated by C4 and aquatic C3 plants. Suzuki et al. (2019) proposed that terrestrial C3 plants were replaced by C4 plants in the drainage basin between 8.2 and 6.3 Ma, a change they infer to be driven by drying of the source area during late Miocene time.

Khim et al. (2019) reconstructed long-term variations in carbon isotopes of sedimentary organic matter at Site U1457, assuming that the sedimentary organic matter represents a mixture of marine and terrestrial materials. The δ¹³C values measured from upper Miocene to upper Pleistocene sediments were generally low and show little variation, reflecting a consistent supply of terrestrial organic matter dominated by C3 plants from the Indus basin. However, from c. 8 to 7 Ma, δ¹³C values were somewhat more depleted, suggesting increased input of C3 plant organic matter at that time. After 7 Ma, higher δ¹³C values support an increase in the contribution from C4 plants, consistent with the Suzuki et al. (2019) study. Khim et al. (2019) interpreted the change in terrestrial vegetation to reflect a shift towards enhanced aridity in the Indus catchment.

Change in the source of siliciclastic material to the submarine fan was reconstructed since 3.4 Ma by Lu et al. (2020) using a combination of neodymium and strontium isotopes, together with single-grain zircon U–Pb age data. These authors show that the sediments represent a combination of erosion from the Himalayas, Karakoram and Kohistan–Ladakh, as well as some sediment flux from the peninsula of India, including the Deccan Traps and Indian Craton. They interpreted changes in the sediment source to be associated with sea-level changes, which vary along with Asian monsoon rainfall. They also examined grain size and hematite content spanning the same time period, and found a three-stage evolution since late Pliocene time. Between 3.4 and 2.4 Ma there was strong chemical weathering despite weak monsoon precipitation. The record implies an increasing but quite variable summer monsoon between 1.8 and 1.1 Ma, followed by weaker summer rainfall and moderate chemical weathering after 1.1 Ma. Spectral analysis indicates orbital cyclicities of 100 ka (eccentricity) and 41 ka (obliquity), but not 21 ka (precession) apparent in the grain size and hematite content records. This spectral record mirrors the precipitation record of Clemens et al. (2018) from the discharge of the Yangtze River, which is tied to the East Asian monsoon. These observations suggest that the monsoon, as well as sea level and global temperatures, controlled weathering and erosion in SW Asia since late Pliocene time.

Cai et al. (2018) examined sediment deposited at Site U1456 since 3.7 Ma. They analysed the clay mineral content, and found that the samples were dominated by smectite and illite, with lower proportions of chlorite and kaolinite. Using these data they were able to define four phases of deposition that they attributed to changes in provenance. From 3.7 to 3.2 Ma the Indus River dominated, whereas from 3.2 to 2.6 Ma the Indus was supplemented by additional flux from the Deccan Traps. The Indus River again dominated between 2.6 and 1.2 Ma, with a resumption of flux from the Deccan Traps after 1.2 Ma. This study proposes that these variations were driven by changes in the intensity of the summer monsoon. When the monsoon was relatively weak the Indus River dominated, whereas strengthening of the summer rains increased sediment supply from the Deccan Traps. They concluded that clay minerals may be a useful method for tracing provenance at least since 3.7 Ma, despite the fact that these are also influenced by chemical weathering driven by climate (Thiry 2000).

A similar approach was used by Khim et al. (2018) covering the last 800 ka at Site U1456. They compared the values of strontium and neodymium isotopes of the sediment in the Indus Fan to possible river sources and interpreted the data to represent mixtures of sediment supplied from the Tapti, Narmada and Indus rivers. They propose that the Indus River was more important during glacial times, but that isotopic variations do not correspond precisely to glacial–interglacial cycles. They noted that sandy turbidite deposits tended to be associated with high ⁸⁷Sr/⁸⁶Sr ratios and low ɛ_Nd values, likely linked to erosion from the Himalayas and delivered by the Indus River. As in the Kumar et al. (2019) and Lu et al. (2020) studies, they propose that sea level plays a critical role in controlling sediment flux into the basin.

The tectonics of the Laxmi Basin was investigated by Pandey et al. (2018) using two-dimensional post-rift flexural backstripping methods (Kusznir et al. 1995). They estimated the degree of extension across the basin, assuming a low flexural rigidity and thermal subsidence following a pure shear extensional model (McKenzie 1978). They further restored the flat-topped Raman Seamount located within the basin, considered to have been emplaced shortly after the late Cretaceous extension, to sea level. Their backstripping analysis demonstrates that the Laxmi Basin has undergone an extreme stretching of pre-existing crust since Late Cretaceous time. At the same time they calculated de-compacted sedimentation rates and estimated that sedimentation rates would have peaked during early–middle Miocene time, coeval with uplift of the Himalaya and with a phase of strong activity by the South Asian summer monsoon, similar to values derived from the upper fan, adjacent to the modern delta (Clift 2006).