Tuesday, March 15, 2016

Sedimentation Patterns Bay Of Bengal: How Old Is The River Ganga

...In our legends it is said that the goddess Ganga's descent from the heavens would have split the earth had Lord Shiva not tamed here torrent by tying it into his ash-smeared locks. To hear this story is to see the river in a certain way: as a heavenly braid, for instance, an immense rope of water, unfurling through a wide and thirsty plain. That there is a further twist to the tale becomes apparent only in the final stages of the river's journey - and this part of the story always comes as a surprise, because it is never told and thus never imagined. It is this : there is a point at which the braid comes undone; where Lord Shiva's matted hair is washed apart into a vast knotted tangle. Once past that point the river throws off its bindings and separates into hundreds, maybe thousands of tangled strands....

Amitav Ghosh- The Hungry Tide

A study by K.S Krishna and colleagues published recently in Current Science shows very elegantly using sesimic reflection profiles and sediment isopach maps how the Bay of Bengal has been filling up with sediment since Late Cretaceous times.

The Bay of Bengal (BoB) originated with the rifting of India from Antarctica by early Cretaceous, thus forming the Indian east coast margin. The depression over time evolved into an ocean basin with new oceanic lithosphere forming in the Bay of Bengal at sea floor spreading centres. Its conjugate oceanic crust is probably beneath the Enderby Basin at the margin of East Antarctica. In this depression, an enormous volume of sediment has been deposited from mid-late Cretaceous to recent times. The age ranges of sediment packages, their distribution, geometry and thickness tell us about the changing source regions of these sediments and the influence of the rising Himalayas and the monsoons on sedimentation history.

Scientists involved in this study used seismic  reflection profiles to construct a seismic stratigraphy of the sediment pile in the BoB. The ages ranges of the sesimic sequence were then calibrated using biostratigraphy erected from two deep sediment cores from the vicinity of the seismic lines. Thus, a Cretaceous to recent subdivision of the mega sequence into distinct depositional episodes separated by unconformities could be recognized. The seismic profiles also revealed the geometry and thicknesses of the sedimentary sequences and the topography of the basement.

Several coast perpendicular grabens (linear depressions) were clearly outlined. These are continuations of ancient Archaean and Proterozoic sutures zones and rifts now occupied by the major rivers of Peninsular India, the Cauvery, Krishna, Godavari and the Mahanadi. Sedimentary packages until the late Oligocene are thicker near the east coast and thin out into the deeper shelf areas. From early Miocene onwards the sediment packages are thicker near the Ganges Brahmaputra delta and in the deeper shelf areas to the east and north and thin towards the shallower coastal shelf. This implies a changes in direction of sediment delivery systems, with the Indian craton being the major source in the earlier phase, to the Himalayas and the Indo Burman ranges being the major source in the younger phase.

This is brought out beautifully by sediment isopach maps constructed for several time slices. Isopach maps show the thickness of sediments for a particular time slice. In this case four time slices were used: 1) Basement to Late Cretaceous 2) Late Cretaceous to Oligocene 3) Oligocene top to Late Miocene 4) Late Pleisocene to recent.

The results are shown below and they clearly show changing sediment sources and distribution pathways.

From Cretaceous to the Oligocene, thick sediment wedges coincide with the Peninsular river grabens indicating that sediments derived by erosion of the Indian craton was the major source to the BoB.

 Source: K.S Krishna et al. 2016

From late Oligocene throughout much of the Miocene the sediments are thickest in the Ganges Brahmaputra delta region and thin out southwestwards towards the east coast shelf area.

 Source: K.S Krishna et al. 2016

This indicates that river systems eroding the rising Himalayas were now the major suppliers of sediment to the BoB. Sediment thickness from late Miocene to mid Pleistocene also show enhanced sedimentation from the north and this pattern coincides with an increase in the Asian monsoon. From Pleistocene to recent times, there has been more sediment from the Godavari Krishna system to the Bob, while a strong sediment delivery system from the Ganges-Brahmaputra continues.

This is the end of part 1 of the post, but I had an intriguing question..

How old is the river Ganga of the plains, flowing from the Himalayan front near Haridwar to the Bay of Bengal?  What does the data from the Bengal Basin tell us about paleo-rivers and how do geologists go about collecting and analyzing this data?

To answer this we go to the Bengal Basin, currently the site of the enormous Ganges-Brahmaputra delta, so eloquently described by Amitav Ghose in his book The Hungry Tide, and to the Gangetic plains and to the Cenozoic foreland basin sediments in Nepal and Uttarakhand in India.

Let's begin with this graphic which shows the evolution of the Himalayan orogen and the drainage systems as India collided with Asia in the Cenozoic

Source: Uddin A. and Lundberg N. 1998

You can see that geologists think that by Miocene times (beginning ~ 23 million years ago), the Himalayan orogen had enough relief to have produced a distinct drainage system along its southerly flanks. The river Brahmaputra is seen to have a Miocene history. The general agreement is that the Ganga too would have been flowing into the Bay of Bengal by at least 15 million years ago (Middle Miocene) or so.

What is the evidence for this?

Another powerful use of isopach maps of the sedimentary sequence deposited in the Bengal Basin brings out Miocene drainage networks. The figure below of the tectonic elements of the Bengal Basin is from a paper by Ashraf Uddin and Neil Lundberg. It also shows  the location of wells that provided the sediment thickness and sand/shale ratio data.

Source: Uddin A. and Lundberg N. 1999

Sediment thickness and sand/shale plots of the Middle to Late Miocene Surma Group sediments (Lower Miocene plots show a similar pattern) clearly outline two major sediment lobes representing sand rich river channel networks flowing from the N and ENE source. These have been interpreted to be delta distributory channels of a major river. The sediment mineralogy of these  Miocene sediments has also been studied.  The sediments are made up of metamorphic and sedimentary lithic grains, plagioclase and K feldspar and a variety of minerals like pyroxenes, amphiboles, epidote and tourmaline, indicating exhumation and erosion of granites, granitic gniesses and a variety of metamorphic schists that make up the Himalayan orogen to the north and the Indo-Burman ranges to the east.

Source: Uddin A. and Lundberg N. 1999

Sources and therefore paleo-drainage has also been pinpointed by 40Ar/39Ar dating of detrital white micas found in the Surma Group. 40Ar/39Ar is a thermochronometer. This is a radiometric dating method which gives the age of the closure temperature of the mineral i.e. the temperature below which there is no diffusion of the parent or daughter products into the external environment. The timing of major faulting events in the Himalayas have been calculated using this method. For example, in the Bhuban Formation of the Surma Group sediments, the 40Ar/39Ar dates of white mica show a strong 16 million year peak. This corresponds to the timing of movement of the Main Central Thrust in the Sikkim Himalayas. Movement along the Main Central Thrust generated enough heat to cause the growth of white micas in the vicinity of the fault zone. 40Ar is derived from the decay of radiogenic 40K in white mica. So, a 16 million year old 40Ar/39Ar date for white micas in the Bengal Basin sediments suggest erosion of rocks uplifted by the Main Central Thrust in the Sikkim Himalaya!

All these data taken together suggest the presence of a paleo-Brahmaputra flowing southwards from the eastern Himalaya syntaxis with additional drainage from the eroding eastern Indo Burman ranges. The sediment route brought out by the isopach maps and sand percentage maps show that this channel system in the Miocene flowed through the upper Assam valley and eastwards of the Shillong plateau and entered the Bengal Basin via the Sylhet depression. Sediments were then transported further south deeper into the Bay of Bengal. It is only in more recent times (Pleistocene) that the uplift of the Shillong plateau has forced the Brahmaputra to change course. It now takes a westerly turn and wraps itself along the northern margin of the Shillong plateau before turning south into the Bengal Basin. Fig Source: Uddin A. and Lundberg N. 1999

What about the paleo-Ganga?

The data from well Hazipur 1 is quite revealing. It is far away from the influence of the paleo-Brahmaputra of the Miocene (don't get confused by its proximity to the present location of the Brahmaputra). It also shows a high percentage of sand suggesting a major river channel west of the paleo-Brahmaputra. In the western Bengal Basin, seismic stratigraphy has also shown the presence of Miocene age mega channels flowing south. These channels indicate the presence of another river distributary system and strongly hint that a Miocene paleo-Ganga was reaching the Bengal Basin then. Unfortunately, no sediment mineralogy work is available that could pinpoint the source of these westerly Miocene channel sediments.

The search for the paleo-Ganga takes us northwestwards into the Himalayan foreland and alluvial plains of Nepal and Uttarakhand, India. Is there any physical sedimentological evidence of large Miocene river channels flowing southeastwards roughly parallel to the Himalayn orogen, as the Ganga does today?

Geologists have extensively studied the Cenozoic foreland basin that developed in response to the loading and flexure of the lithosphere due to the Himalayan orogeny. Essentially a big linear moat or a trough formed in front of the rising Himalayas. This moat contains sediments including the Siwalik Group, which was deposited by rivers draining the Himalayas in the Middle Miocene, beginning around 15 million years ago.

Below is a reconstruction of the depositional environment of the Siwalik Group.

Source: Kumar R. 2003

It shows that major river channels flowed perpendicular to the Himalayan thrust front and deposited sediments in alluvial floodplains and in meandering and braided stream channels. Large alluvial fans were formed due to this style of deposition. The situation is analogous to the present day river system in which south flowing transverse rivers like the Ramganga, Goriganga, Kosi and the Gandaki draining the Uttarakhand and Nepal Himalaya join an axial Ganga river flowing southeastwards. Paleocurrent analysis of Siwalik sediments show a south and south easterly current direction. This implies a transverse drainage network like the present day Ramganga and Kosi.  There is so far no evidence for a Miocene axial river (paleo-Ganga) flowing parallel to the Himalayan orogen in the Siwalik Group of sediments.

That doesn't mean that there was no Miocene Ganga, but that the exposed Siwaliks are the wrong place to look  for it. An axial river would have developed a little further south of the alluvial fan complexes prograding from the orogen front and so older axial river channels are likely buried beneath the Gangetic Pleistocene-Holocene alluvium. The oldest sediment deposited by the Ganga which is currently exposed in the Gangetic plains is about 120 thousand years old. But there is 2-3 km of older sediments ranging in age from Miocene  to mid Pleistocene underneath this recent alluvium. In effect, the Siwaliks continue in the subsurface of the Gangetic plains. This next map shows the tectonic elements of the Himalayan system and sediment thicknesses in the Gangetic plains.

 Source: Singh I B 1996

It shows that the exposed Siwaliks ranges (in map- southern boundary of frontal folded zone)  are just the deformed northernmost part of the foreland basin . All those transverse streams eroding the Himalayas and depositing sediment in the proximal part of the Siwaliks would have met an axial river flowing in the distal parts of the foreland, more to the south of the orogenic front.

Interestingly, we can use paleocurrent analysis to further constraint the timing of development of the south and southeasterly flowing Ganga drainage system using sedimentary rocks deposited in the Himalayan foreland in early Miocene times. These rocks are older than the Siwalik Group. They are the sandstones of the Dumri Formation in Nepal and they represent the renewal of deposition in the Himalayan foreland after the Tethyan ocean had completely closed and the seas had disappeared.

Sediments of the Dumri Formation were also laid down by rivers flowing down the rising Himalaya orogen. However, in contrast to the Siwaliks which represent streams flowing transverse to the rising Himalaya, the Dumri Formation may represent an axial river system flowing parallel to the orogen. Paleocurrent analysis of Dumri beds shows a south to south westerly drainage indicating rivers flowing westwards. This means that in early Miocene, around 20 million years ago, a large axial river draining the Himalayan foreland may have been flowing westwards, and ultimately feeding the Paleocene to Early Miocene age Katawaz delta system in present day Pakistan (Qayyum et. al. 1999?).

It is only in the Middle Miocene times (~15 mya) that an east flowing system analogous to the present day Ganga developed. The likely reason for this switch was the formation of a topographic divide due to the impingement of the rigid high-standing Aravalli Range into the Himalaya fold and thrust belt. This resulted in the Himalayan foreland being partitioned into west and east flowing river systems. (DeCelles et.al. 1998).

Further to the west, in the Himalayan foreland of NW India, evidence of an early Miocene west flowing axial river has not yet been found. The Early Miocene age (~ 20 million years) Dharamsala Group sediments of the Kangra sub-basin are interpreted to have been deposited in SW and SE flowing transverse rivers.

There is another geochemical clue that a paleo-Ganga and paleo-Brahmaputra along with a paleo-Indus were draining the Himalayas since the Miocene and that evidence lies in the changing 87Sr (Strontium) isotope and 187Os (Osmium) isotope composition of sea water. The reasoning is that the uplift of the Himalayas exposed enormous volumes of new rock which were rich in these elements. Chemical weathering of these rocks leached these elements into rivers and ultimately into the ocean. Take a look at the plot of changing 87Sr and 187Os isotope of sea water through the Cenozoic. The Sr isotope  composition of ancient sea water is estimated by measuring Sr  incorporated in CaCO3 shells of marine creatures. The Os isotope composition of sea water is estimated by measuring Os incorporated in manganese hydroxide nodules precipitated in deep sea sediments.

 Source: Myrow et al. 2015

In the plot you can clearly see an enrichment of 87Sr isotope in the early part of the Miocene. This has been interpreted to signal the exhumation and erosion of radiogenic 87Sr rich granitic rocks of the Greater Himalayan Crystalline Series uplifted by the movement along the Main Central Thrust. At around 16 million years ago, there is a flattening in the 87Sr signal indicating decreased erosion of the Greater Himalayas. This is paralleled at 16 million years ago by an increase in the 187Os content of sea water. This change has been interpreted to signal the uplift, exhumation and weathering of the Lesser Himalayas.

Neoproterozoic to Cambrian strata of the Lesser Himalayan Krol Tal Group contain  organic and phosphatic black shales which are enriched in radiogenic 187Os. This Krol-Tal belt is exposed all across the strike of the Himalayan orogen from Pakistan to the west through Himachal Pradesh and Uttarkhand in India, to Bhutan in the east. So, the changing sea water chemistry through the Cenozoic strongly hints at the erosion of these rocks by the Gangetic system rivers along with the Indus and Brahmaputra systems beginning by around 16 million years ago.

I can argue like a lawyer and point out that since physical evidence of a paleo-Ganga as an axial river is missing, the changing sea water chemistry could be due to combined contributions of the Indus and Brahmaputra. Or for that matter,  erosion anywhere in the world of Sr and Os rich rocks in the Miocene could explain the change in composition of the well mixed global ocean  So, is there any evidence that Ganga system tributaries were enriched in 187Os during the Miocene and that they were contributing this element to the ocean?

Yes there is, and this evidence comes from paleosols (ancient soils) preserved in Miocene foreland basin deposits of Nepal.

Here, geologists have found out that beginning around 15 million years ago, Mn (manganese) oxide nodules precipitated in the floodplains of river deposits are enriched in 187Os. The weathering and erosion of Krol-Tal strata released clay and silt size particles which got deposited in floodplains. These floodplain sediments were chemically weathered and released 187Os, which due to its affinity to Mn got incorporated in the growing Mn oxide nodules in the soils forming on floodplain sediments. The ground water that was carrying this 187Os also released that element in the river. These transverse rivers would have joined an axial river, the paleo-Ganga to the south, thus carrying the 187Os signal to the Bay of Bengal.

All this sound circumstantial, but those transverse rivers would have eventually had to join a larger axial river just as they do today. Where else could they have terminated?

As you might have guessed all this has been more of an intellectual exercise for me over the past few weeks. Most of the research studies don't directly address the question of a paleo-Ganga. Rather, the emphasis of Himalayan studies is on understanding the sequential uplift of lithotectonic terrains and their erosional history via its signal as preserved in the Cenozoic foreland basin and Bay of Bengal sediments. I've had to pick my way through several research studies to tease out the bits relevant to this question. A long email exchange with Ashraf Uddin, my friend from Florida State University days who has worked extensively on Bengal Basin sediments also helped.

Looking at the entirety of the data I will say that the evidence for a Miocene paleo Ganga is not as clear or direct as it is for a paleo Brahmaputra, but that is because of a lack of detailed studies. Provenance work on those Miocene mega channels in west Bangladesh and seismic profiling of the Gangetic plains should surely reveal a Miocene Ganga. Presently the Ganga breaches the Siwalik hills at Haridwar. In the past before the Siwaliks existed, beginning around 15 million years ago, there would have been a large mountain stream originating from the great heights of the Greater Himalaya, flowing through the expanse of the Lesser Himalaya and onto the plains somewhere near present day Haridwar, and then funneled by the elongate foreland trough,  taking a southeasterly course towards the Bay of Bengal.

Krishna, K., Ismaiel, M., Srinivas, K., Gopala Rao, D., Mishra, J., & Saha, D. (2016). Sediment Pathways and Emergence of Himalayan Source Material in the Bay of Bengal Current Science, 110 (3) DOI: 10.18520/cs/v110/i3/363-372


  1. Nice Suvrat. Perhaps coincidental perhaps not that the Godavari is called the 'Vriddha Ganga'. The Ganga is that which comes down as rain and deposits sediment in the Bay of Bengal[does that have an older name?]

  2. thanks Urmila- interesting! don't know about an older name for the Ganga :)

  3. Replies
    1. Wikipedia came up with this -

      In ancient Hindu scriptures, this water body is referred to as 'Mahodadhi' (Sanskrit: महोदधि, lit. great water receptacle)[5][6] while it appears as Sinus Gangeticus or Gangeticus Sinus, meaning "Gulf of the Ganges", in ancient maps.[7]

      The other Sanskrit names for Bay of Bengal are 'Vangopasagara' (Sanskrit: वङ्गोपसागर, lit. Bengal's Bay), also simply called as 'Vangasagara' (Sanskrit: वङ्गसागर, lit. Bengal Sea) and 'Purvapayodhi' (Sanskrit: पूर्वपयोधि, lit. Eastern Ocean).


  4. I read the whole account, trying to figure out the location of paleao-Ganga before Himalayan orogeny, since it is known to be antecedent. Did it exist as axial river at the northern periphery of peninsula with Chambal, Betwa, Ken etc flowing into it. Or did the closure of Tethys result in formation of Ganga. I am a biologist and hence a layman for geology. A simple answer is desired. Prakash Nautiyal

    1. thanks for reading- The Indus and Brahmaputra are antecedent. They originate on the Asian plate in Tibet and so would have existed before the India Asia collision and the rise of the Himalaya. The Ganga is not antecedent. It originates in the Greater Himalayan ranges which is uplifted Indian crust. So its origin begins after the closure of Tethys and establishment of continental conditions. A southerly drainage (paleo-Yamuna and paleo Ganga) would have developed as the Himalayas started rising.

      Before the Tethys closed, peninsular Indian rivers flowing north northeastwards similar to the Chambal and Betwa would have flowed into the Tethys sea and only later after the closure of the Tethys in the early Miocene would have been captured by an axial Yamuna-Ganga drainage system.

  5. Where was the Ganga located before Himalayan uplift, northern periphery of the Peninsula, with Chambal, Betwa etc flowing into it?