The Geology Of Mumbai

Last Saturday I was fortunate to be given a tour of the construction site of  Mumbai Metro Line 3 near Siddhivinayak Temple in Dadar, Mumbai. There, we descended about 100 feet to the floor of an enormous pit, and then traveled south along a tunnel for a kilometer towards Worli, right up to where the Tunnel Boring Machines (TBM) were at work.

It was a fantastic experience.

Geology is not a term you would normally associate with the concerns of a bustling metropolis like Mumbai. Yet, at this enormous construction site, it is at the heart of operations. Progress very much depends on understanding the subsurface rock layers. Their thickness, strength, orientation, and water bearing capacity, pose engineering challenges that need to be understood and solved before tunneling can proceed safely. Far from just being an esoteric pursuit that delves into the earth's dusty past, at this site, every thump of the giant TBM rams home the relevance of geology in our day to day lives.

A friend asked me whether the rocks that the TBM's are encountering in Mumbai are any different from those under Pune. Metro construction has started in Pune too, but only one section about 6 km long will be underground. One of the reasons given for avoiding long underground stretches in Pune is that the rock type is very hard basalt.

Mumbai geology is somewhat different from Pune. I did not see any rock during my Metro visit since the pits and the tunnels had already been lined. But I do have a fair idea of the geological history of Mumbai area.

Like most of Maharashtra, Mumbai too is part of the Deccan Volcanic Province. This enormous area covered by mostly basalt lava formed between 68 million and 60 million years ago, from Late Cretaceous to Early Paleocene times. The bulk of the volcanism, about 80% of it, occurred between 67 million and 66 million years ago, within a time span of a few hundred thousand years. This big spurt of volcanism overlaps the mass extinction that took place at 66.03 million years ago. The main cause of this extinction is the environmental degradation resulting from a large meteorite crashing into what is now the Yucatan Peninsula of Mexico. A lively debate has now broken out on how much did Deccan volcanism contribute to the mass extinction.

At this time the Indian continent was located far south of the equator. The Mumbai region was located around 25 degrees south of the equator. The map below shows in grey the distribution of the Deccan Volcanics in context to the other major geological provinces of India. Insets show the progressive separation of India from Madagascar at 88 million years ago, and later from Seychelles at 65 million yeas ago. The black region in the right inset are the Deccan Volcanics with the smaller fragment being Seychelles.

Source: Sheth H.C. 2007

By 65 million years ago the western margin of the continent began to split apart and a chunk which became Seychelles broke and moved away from the Indian continent.  North south oriented fault systems along the western margin of India caused blocks of crust to subside westwards. The region around Mumbai would have been at sea level by around 64 million years ago. Eruptions had ceased over most of the Deccan Volcanic Province.

A map showing the major tectonic elements of the Indian western margin and the Mumbai area is shown below.

Source: Sheth H.C. 1998

In the Mumbai region though, volcanism continued for the next few million years under conditions which imparted to Mumbai its peculiar geological character. This volcanism differed from the rest of the Deccan Province. 

First, the lava composition was more 'evolved'. The Deccan Province is made mostly of basalt, which is an igneous rock rich in iron, magnesium and calcium silicate minerals. However, in the Mumbai region, besides basalt, other lava types known as rhyolites and trachytes erupted. These lavas are more silica rich and contain the mineral quartz (silica dioxide) and other sodium and calcium silicate minerals.

Secondly, since this region was at or near sea level, some of the volcanism took place under water forming characteristic pillow like lava structures. Volcanism over the rest of the Deccan Province took place in subaerial conditions above sea level. 

Thirdly, the meeting of hot lava and cold sea water caused steam explosions. This resulted in the formation of large amounts of lava rubble which when consolidated forms a rock known as volcanic breccia. Explosive volcanism also generated ash which was deposited in layers known as Tuff. 

Volcanism was also sporadic. In these interludes, in coastal embayments and lagoons, mud and silt was being deposited. Fossils of turtle, frogs, crocodiles, molluscs and various types of plant remains have been recovered from these sediments. A resumption of volcanism would bury these sediments under lava. Repeated episodes of volcanism and sediment deposition has resulted in the formation of a rock sequence made up of different lava types alternating with thinner layers of sedimentary rock (intertrappean sediments). These events took place between 64 million and 62 million years ago. 

The volcano-sedimentary environments of Mumbai are shown in the schematic below. 

Volcanism continued until around 60 million years ago. The famous Gilbert Hill in Andheri, made up of basalt columns, formed by polygonal cracking of lava as it cooled, has been dated to around 60 million years old. This makes it probably the youngest volcanic activity of the Deccan Province.

Finally, the Mumbai rock sequence differs from the rest of the province in its structural disposition. Whereas in the rest of the Deccan region the lava flows are nearly horizontal, in the Mumbai region they show a pronounced tilt (dip) to the west. This feature is known as the Panvel Flexure, as it becomes more pronounced beginning just around the town of Panvel, a few tens of kilometers east of Mumbai (see right panel of tectonic map posted earlier).  

Many explanations have been given for this tilt. One theory is that it resulted from a bending of the lava flows as the crust to the west of Mumbai subsided upon cooling and due to the weight of sediment. Another explanation ties the structure to continued movement along west facing faults which initially formed  during continental breakup. A third hypothesis is that the flexure formed by tilting of the crust along an east facing listric (curved plane) fault now located under the Arabian Sea to the west of Mumbai. Such faults commonly occur along continental rift margins, where the crust in being pulled apart. This last scenario is shown below.

Source:  Sheth H.C. 1998

This tilting occurred after volcanism and sedimentation ended, later than 60 million years ago. The result is that the entire package of volcanic flows and sedimentary strata dip westwards. This Mumbai stratigraphy is shown in the cross section. 

Source: Cripps J.A. et .al 2005

After volcanism and crustal tilting, the next recorded geologic history is from much more recent times, in fact just a few thousand years old.

Early travelers and geographers describe Mumbai not as one land mass, but a collection of seven islands separated by shallow tidal inlets and marshland. This particular configuration of land and sea, is in geological time quite a recent phenomenon, forming  just about 10,000 years ago. Before that, during the Pleistocene ice age, sea level was about 100 meters lower than present. The Mumbai area and almost the entire continental shelf to the west would have been land. The earliest humans to have entered India about 70,000 years ago, following a coastal route from the Arabian Peninsula, would have walked on the now submerged land to the west of Mumbai.

During this sea level low, rivers traversing the Mumbai region would have met the sea tens of kilometers to the west. Sea level began to rise about 15,000 to 12,000 years ago at the end of the ice age. In the next few thousand years, rising sea level inundated the continental shelf and various river valleys, forming Panvel Creek, Thane Creek and Vasai Creek to name a few of the creeks in this region. These creeks are all drowned river valleys of the Pleistocene.

Sea level peaked about 3,000 to 4,000  years ago. The position of the shoreline at this time was about 2 meters higher than present. Beach rock deposited during this time is present a few hundred meters inland at Madh Island. This shelly rock is locally known as Karal. By this time Mumbai became an island locale, with topographic highs remaining as land, with low lying areas becoming marshes and shallow tidal channels.

This then is the geological inheritance of the city of Mumbai,a legacy of  volcanism and sedimentation in Paleocene times and a pronounced sea level rise during the Holocene.

The rock outcrops that tell this story have all but disappeared under the onslaught of urbanization over the last few decades. As modern Tunnel Boring Machines enter Mumbai's underworld, a few pages of this history are again being discovered.

The Mumbai Metro website, in their newsletter Metro Cube, has put up a series of ten articles titled 'What Lies Beneath The Earth' (issues February 2018 - December 2018). This series summarizes the geology beneath each of the sections of the metro route. It is an excellent resource. A perusal of this series reveals that the tunnels are mostly encountering Paleocene age hard basalt and softer breccias and tuff layers. Only at some place are sedimentary layers being intersected. This though is in contrast with the geology underneath Pune. There, only hard basalt will be found.

It is imperative that we save some of this treasure for our citizens to appreciate. Wouldn't it be wonderful if at a few of the underground metro stations, exposed rock panels and a museum like display of recovered rock cores along with a short history of Mumbai geology is displayed? It would make Mumbai's unique geology accessible to citizens and help all of us forge a more enduring connection with our natural heritage.

Stalactites And Other Calc Tufa Deposits Along Bageshwar Shama Road, Kumaon Himalaya

Traveling from Bageshwar to Shama, in Kumaon Uttarakhand, I came across a wondrous calc tufa deposit about a kilometer south of Kapkot village.

 The map below shows Bageshwar and Kapkot along Route 37. (Permanent Link).



Calc Tufa are calcium carbonate deposits which form on land in a subaerial environment. They are made up of the minerals calcite and, less commonly, aragonite. The most familiar of calcium carbonate deposits are sea floor and beach accumulations of shells and skeletons of marine organisms. Upon burial and hardening they turn into limestones. In the Proterozoic, before animals evolved the ability to biomineralize, vast thicknesses of limestones formed in the oceans by inorganic and bacterially mediated precipitation of calcium carbonate. Limestones that form in saline as well as fresh water lakes are also known.

Calc Tufa forms in the vicinity of springs, waterfalls, along river banks, caves and along hill slopes. They have a chalky texture, porous fabric and organic looking shapes. This is a result of calcium carbonate encrusting microbial, algal and moss colonies that inhabit these settings. Associated with these porous friable looking forms are more denser crystalline deposits. These are stalactites and various types of laminated and globular crusts. They are collectively called speleothems. They form generally in a cave setting by abiogenic precipitation from thin films of supersaturated water. This particular deposit containing both tufa and speleothems was along a steep hill slope with large cavities. The substrate rocks are the Mesoproterozoic age Deoban limestone and dolostones (made up of mineral dolomite). They are estimated to be around 1.5- 1.6 billion years old.

All along the exposure the rocks were shattered by prominent fracture zones. Rain water is weakly acidic. As it falls and moves through the cracks and fractures in these rocks it dissolves the minerals calcite and dolomite, becoming enriched in dissolved carbon dioxide (CO2) and calcium.  The partial pressure of CO2 (a measure of dissolved CO2 concentration) in this groundwater is more than the partial pressure of CO2 in the atmosphere. When groundwater enters a cave or emerges on a hill slope as a spring discharge, the lower partial pressure of CO2 in this open setting causes a degassing of CO2 from the groundwater. This results in the pH of the water to increase slightly, which in turn causes supersaturation of calcium carbonate in solution. Precipitation of calcium carbonate then begins on the cave walls and roof and on the hill slopes. It is possible that removal of CO2 by microbial photosynthesis may also be playing a role in triggering precipitation.

These tufa deposits occur at many places along the Bageshwar to Shama road. We finally stopped for a closer look at a largish looking deposit about a kilometer south of Kapkot. This was strictly road side geology on my part. We spent about half an hour at the deposit and so I am not presenting any detailed analysis or insights regarding this feature.

This is a complex deposit made up of varied types of tufa. We managed to photograph some beautiful calc tufa morphologies which I am posting below. My thanks to Pushkaraj Apte ( @pushkarajapte ) for contributing many of the photographs.

Lets get an idea of the size of the deposit. That's me, standing in front of the large cavern. You can see stalactites in the background.

 A peek inside the large cavity. It is about 3 meters in height and about 4 meters in width. I could have easily stood inside it. However, I did not enter it, fearing I would break some delicate mineral deposits which have formed on the floor of the cave.

Speleothems

Stalactites 1: The most striking of the formations are these stalactites. They range from thick columnar forms (1) which are more than a meter in length to smaller centimeter long thin delicate drips (2). The cave is damp. There is a thin film of water covering these columns suggesting ongoing mineral precipitation and growth of the stalactites. The floor of the cavern was also encrusted with deposits and partially covered with tufa debris.

 Stalactites 2: Along the hill slopes, exposed Deoban carbonate strata form ledges. Stalactites are growing on the undersides of these ledges. The bigger ones are about 1-2 feet in length.

 Botryoids:  At places botryoidal clusters (cave grapes) are seen. These hang from the roof (1) and accrete away from walls (2). They form by either radial or concentric growth of calcite (or aragonite) from a nucleation site. Each botryoid is about a centimeter or so in diameter.

Thin Platy Crusts: These thin (cm scale)delicate layers likely form in shallow films or pools of stagnant water on the floor of the cavity.

Flowstones?: These banded crusts  have formed on a slope from flowing water and likely represent abiogenic precipitation of calcite (flowstones). Alternatively they could be stromatolitic crusts formed by precipitation of calcite atop microbial sheaths and mats.

Calc Tufa:

Phytohermal Tufa: These are calcified moss deposits (a foot or so in height) which are formed on the floor of the cavity. They preserve the bushy morphology of the moss colonies. Calcite encrusted and eventually entirely replaced the moss colonies, turning them into fossilized organic structures.

Microhermal Tufa or Phytohermal Tufa: The thin tube like structures (few cm in length) of this calc tufa deposit suggests that it formed by mineral encrustation of filamentous algae or bacterial colonies. However, I cannot be sure. This too could be a moss colony.

Spongiform Tufa: Massive looking with dispersed holes. Such structures from by mineral encrusting organic matter (moss, microbial mats) draping the hillsides. The open spaces between the organic matter and decay of vegetation gives the deposit a sponge like texture. Some larger cavities (about 6 inches across) are lined with layered mineral deposits.

I found this broken piece along the road side next to the deposit. It is made up of small globular aggregates and columns which have accreted upon a substrate of spongiform tufa.

In this transverse section you can see clearly the calcium carbonate layers that have built up the column.

A cross section of the larger stalactites will also reveal its layered nature. Stalactites with such growth layers are of importance in reconstructing past climates. The oxygen in the calcite (CaCO3) provides the clue. Variations in the ratio of the two isotopes of oxygen (O18/O16) which are bound up in calcite are indicators of differences in the strength of rainfall. The lighter isotope (O16) is preferentially retained in the vapor phase. During phases of weak monsoons or drought, rain becomes enriched in the heavier isotope (O18). Calcite layers precipitated from this water will be enriched in the heavier isotope. In contrast, during strong monsoon phases, rain and groundwater becomes relatively enriched in the lighter isotope. As a result, calcite layers will inherit a 'lighter' oxygen isotope signal.

For the Indian subcontinent, reconstruction of the past variability of Asian monsoons going back hundreds to thousands of years, are based on precious few data points, spread rather sparsely across India. Recently, Gayatri Kathayat and colleagues published a study of Indian monsoon history over the past 5700 years based on the oxygen isotope record of cave stalactites from Sahiya in Uttarkhand, located about 200 km WNW of where we were. Judging by the size of some of the stalactites, I am guessing that deposition at this Kapkot site has been going on for a few hundred years at least. I wonder if this deposit can be a new paleo climate data source.

I did have another intriguing thought. Is the profusion of calc tufa deposits along road cuts in this region just a coincidence? Is it possible that blasting and cutting the hill side for building the road enhanced fractures and triggered collapse of blocks, resulting in the formation of caverns, and creating conditions favorable for calc tufa precipitation?   If so, then this deposit may be at most a hundred years old. Wild!

W. Bengal Bangladesh- Geologic Controls On Arsenic Distribution In Ground Water

Science writer  Priyanka Pulla has written an excellent article exploring the geologic, socio-economic and technological issues related to the widespread arsenic contamination of groundwater in W. Bengal. Sadly, the government response to this crisis has been slow.

I thought I would elaborate on the geological question -  Why are Arsenic (As) levels much higher in shallower Holocene age aquifers and lower in the deeper Pleistocene age aquifer? The answer encompasses mineralogy, climate change, sea level changes and bacteria.

The ultimate source of As are high Himalayan rocks and Indo-Burman ranges with additional contributions from the Precambrian terrains of Peninsular India and the Siwalik hills.  Minerals like biotite, magnetite, illmenite, olivine, pyroxene, amphiboles contain As. These minerals release As when they undergo weathering in catchment areas and deposits of the alluvial plains. This As is absorbed on secondary minerals like Fe hydroxides like goethite. Such Fe hydroxides are authigenic, i.e. they grow in the shallow buried sediments of the alluvial plains. Under oxidizing conditions, As is immobile, sequestered in Fe hydroxides. However,  conditions may change, and these sediments may get overlain by or be redeposited in environments rich in organic material. Certain bacterial species living on this organic material break down these Fe hydroxides, using the oxygen for their metabolism, and releasing Fe and As into the groundwater. This is known as reductive dissolution of Fe hydroxides and is the principal mechanism for As entering the groundwater in the alluvial plains of Bangladesh and West Bengal.

During the Pleistocene.. 1) the high Himalaya was glaciated. Therefore, important sources of As like the Fe-Mg rich rocks of the Indus ophiolite belt (slices of oceanic crust that existed between India and Asia which have been thrust up during continental collision) and high grade metamorphic rocks such as schists and gneisses were covered in ice and not releasing sediment. Indian cratonic areas, the Siwalik foothills  and the Indo-Burman ranges were being eroded, but overall less As was making its way on to alluvial plains. 2) Since climate was cooler and drier, there was less organic material accumulating in sediment of alluvial plains. Conditions were oxidizing and As remained sequestered in Fe hydroxide minerals. 3) Sea level was much lower then. Almost the entire continental shelf was dry land. Ganga and Brahmaputra met the sea much to the south of present shoreline. Reducing environments like delta front marshes, ponds, estuaries, existed much to the south.

Sedimentary conditions changed by 12-15 thousand years ago. Glacial melt exposed As bearing rocks in high Himalaya. As a result, more As made its way on to alluvial plains. Importantly, sea level rose and flooded the continental shelf. The Pleistocene delta front reducing environments were drowned. Shorelines shifted northwards. The climate was warmer, encouraging vegetation growth. Reducing delta front environments like swamps, coastal marshes and lakes developed on previous alluvial plain sediments.

The map below shows the position of shorelines between 7 thousand and 4 thousand years ago along with the location of wells with high levels of As. This study focuses on Bangladesh but similar conditions existed in West Bengal as well. The sea has receded 2- 3 meters to its present location since 4 thousand years ago.  The delta front and shoreline belt that existed 4-7 thousand years ago is now a densely inhabited region .

 Source: Quaternary shoreline shifting and hydrogeologic influence on the distribution of groundwater arsenic in aquifers of the Bengal Basin- M. Shamsudduha, Ashraf Uddin 2007

Notice clustering of wells with high As along the past shorelines. Here, organic rich delta marshes and swamps developed. Bacterial reduction of Fe hydroxides released As in to groundwater.

As distribution also shows correlation with topography. This map shows high As levels in groundwater coinciding with topographic lows. Such low lying areas accumulate more fine sediment and organic material. Again, this will apply also to W. Bengal.

 Source: Quaternary shoreline shifting and hydrogeologic influence on the distribution of groundwater arsenic in aquifers of the Bengal Basin- M. Shamsudduha, Ashraf Uddin 2007

So, a change in climate and shifts in sedimentary environments in response to changing sea level from Pleistocene to Holocene exerted a strong control on As distribution in the alluvial plains of Bangladesh and W. Bengal. 

Note On The Sutlej Paleochannels

The topographic relief rendition posted below shows beautifully the incised valleys of the glacially sourced Yamuna and Sutlej rivers. This rendition has been derived from the NASA Shuttle Radar Topography Mission (SRTMv3) DEM (Digital Elevation Model) with a 1 arc-second or 30m spatial resolution. The region depicted in the figure is immediately west of the Himalaya frontal ranges covering parts of Punjab and Haryana.

Source: Ajit Singh et. al. 2017 - Counter-intuitive influence of Himalayan river morphodynamics on Indus Civilisation urban settlements.

This incision began during the early Holocene, beginning about 10,000 to 8700 years ago and continuing over the next few thousand years, as proposed in an earlier study by Liviu Giosan and colleagues. A decline in monsoon strength over northwest India resulted in low sediment load carried by the rivers. Under such conditions, starved of sediment, the river starts cutting down or incising into its older deposits. Over time they carve out large valleys as the Yamuna and Sutlej have.

During the mid Holocene, from about 6000 years to 3800 years ago, the region between the Yamuna and the Indus was extensively settled and farmed by the Harappan people. The river Ghaggar flows through this region. One popular theory supported by many geologists was that during Harappan times the river Sutlej flowed into the river Ghaggar, switching to its present course only about 4000 years ago. However, Giosan and colleagues had argued that had that been the case, a large incised valley should have been carved by the Sutlej from the point it exits the Himalaya to the point it joins the Ghaggar. The absence of a wide NE-SW oriented incised valley in the interfluve between the Yamuna and the Indus indicates that the Sutlej did not flow into the Ghaggar during most of the Holocene.

 A recent study led by geologist Sanjeev Gupta (Ajit Singh et. al. 2017)  has validated this scenario using geochemical criteria. They have shown that the Sutlej river once did flow into the Ghaggar but changed course and joined the Indus in the late Pleistocene -early Holocene between 15000- 12,000 years and 8000 years ago. Additional data from Giosan and colleagues shows (SI Text) fluvial deposits of Late Pleistocene-Early Holocene age (latest being 10,000 years old) along the present day Sutlej floodplain. These staggered dates imply that a major channel of the Sutlej avulsed or changed course as early as 15000 to 12000 years ago. A smaller strand of the river continued to flow into the Ghaggar until about 8000 years ago or so.

All this means that the Harappan settlements and agriculture in this region was not sustained by a large perennial glacial fed river. Rather, the Harappans adapted their water usage strategy and farming practices to exploit a smaller and maybe an ephemeral river and more distributed water sources.

The geochemical  work by Gupta and colleagues has been rightly praised and highlighted in many media reports. What did go unnoticed and unappreciated was the relief rendition of the incised channels. They provide a very powerful visual representation of the Holocene fluvial history of this region.

The modified relief rendition below also shows the course of the abandoned Sutlej incised valley. Note that this valley is much narrower than the Sutlej and Yamuna incised valleys. Also, trace these narrower incised valleys upstream and you can see that they originate in the Siwaliks. There are no deep extensive incised valleys along the route I have marked in blue. The Sutlej would have carved a prominent incised valley roughly along the blue route had it been flowing into the Ghaggar during most of the early and mid Holocene. Its absence suggests to me that the valley annotated as the abandoned Sutlej incised valley was really carved out in the earlier part of the Holocene by the smaller Ghaggar river originating in the Siwaliks.

Modified from :  Ajit Singh et. al. 2017 - Counter-intuitive influence of Himalayan river morphodynamics on Indus Civilisation urban settlements

Aside: After Liviu Giosan's paper came out, the archaeologist Shereen Ratnagar asked me whether incised valleys are diagnostic of glacial rivers. She was puzzled because the monsoonal rivers Marakand, Ghaggar and a number of smaller streams which originate in the Siwaliks have also carved incised valleys. The answer is no, they are not. What Giosan's work was pointing out was that wide incised valleys of a particular telltale orientation were absent, thus providing a clue as to when the Sutlej changed its course.

Field Photo: Sea Cliffs And Holocene Sea Level Highstand, India West Coast

All along India's coast there are indicators that 4000-6000 years ago sea level was higher than the present level, oscillating between 1-4 meters above present high tide level at different times. Since then, the sea has gradually receded to its present level. As a result, we can observe stranded beach ridges, cemented beach rock and dunes a few hundred meters inland of the present high tide mark. And we can see erosional notches on sea cliffs marking the past high tide level.

I saw these erosional notches in the sea cliffs exposed along the coast near Harnai village in Konkan.

The satellite image shows the location of the sea cliffs.

The picture below shows a sea cliff with an erosional notch (arrow) about 1.5 meters above the high tide level. This is at the Fattegad Fort near Harnai village. Also, notice the rocky platform that has formed at the current tidal level.

This notch can be traced all along the line of sea cliffs in the area. You can see it very clearly on this cliff, a little north of the previous location.

And the picture below shows a close up of the notch. Sea level must have held steady at this level for a few hundred years to have formed such a distinct erosional feature.

Why was sea level higher in the past? It has to do with the ice age and the end of the last glacial phase. The earth has been in the grips of an ice age for the past 2.6 million years. Conditions have cyclically fluctuated between colder glacial phases and warmer interglacial periods. During glacial phases,  growth of polar ice caps results in a sea level fall. During warmer interglacial phases, polar ice caps melt and raise sea levels. The last glacial phase lasted between 115 - 12 thousand  years ago.  During this time the sea level was as much as 100 meters lower than today. Large swaths of the continental shelf was land then. The earth then moved into a warmer interglacial phase. As a result of melting polar ice, the sea has been rising steadily for the past 10-11 thousand years, flooding the continental shelf, and culminating in a sea level highstand (maximum) about 4000-6000 years ago. This maximum was about 1-4 meters above the present sea level.

There is evidence scattered all along India's west and east coast (and all over the world) of this Holocene sea level high. For example, there are tidal flat deposits about 1 meter above present sea level along the Porbundar coast in Gujarat. Shells collected from these deposits give an age of about 4000 thousand years. Exposed reefs from Mithapur in Jamnagar district in Gujarat give an age of about 2100 years. Oyster reefs exposed along Saurashtra coast about 2 meters above present sea level are about 2500-3000 years old.

To the south, in Madh Island (Mumbai) and along Konkan coast, there are layers of hardened sand and pebbles, locally known as 'Karal', which occur 2-4 meters above present sea level. These sediments once formed a pebbly beach.  At Kelshi village in Konkan, there are fossil beach ridges a few hundred meters inland of the present high tide mark. I saw these on my recent visit. In central Kerala 4000 year old beach ridges occur 3 km inland, indicating that sea level was 3-4 meters higher then.  Along the east coast, there are 4000-6000 year old beach ridges along the Krishna-Godavari coastline. These ridges become younger towards the coast, indicating that the sea has been receding since about 4 thousand years ago.  Along the Baruva-Gandavaram coast in  Andhra Pradesh, sea cliffs have preserved a succession of erosional notches at 4.7 m, 2.3 m and 1.8 m above sea level.

All these features indicate sea level peaked about 4000-6000 thousand years ago and has been falling in fits and starts since. The exact mechanism for this sea level fall is not well understood.

Scientists have put together data form various localities to come up with a composite sea level curve for the Holocene. The curve below has been drawn up using data from Gujarat and shows sea level rising throughout early and mid Holocene. The late Holocene has seen a lowering of seas.

Source: U.B Mathur et.al. 2004

This lowering has now been reversed and the seas are rising again globally, this time induced by anthropogenic global warming as continental glaciers melt and the ocean water expands as it gets warmer. It is estimated that sea level will rise between 0.5 - 1 meter by 2100. In centuries to come, the extent of sea level rise will depend on future warming trends and the extent of melting of the Greenland and Antarctica ice sheets. If significant portions of these ice sheets melt, sea level will rise by several meters in the next few hundred to couple of thousand years.

And what about the flat rocky platforms seen in the tidal zone below the sea cliffs? How do they form? The main mechanism is cliff retreat. Waves pound at cliffs cutting a notch at the bottom of the cliff. Undercutting of the cliff eventually causes collapse of the rock face and a retreat. Over time, as the cliff retreats, a broad sea facing platform develops. This is further acted upon by weathering.  The likely process involves "water layer leveling" combined with wave erosion. Water layer leveling means the lowering and leveling of the rock surface due to physical and chemical weathering by the action of sea water. Standing pools of water and the continuous wetting and drying conditions in the intertidal zone act to weaken the rock and create a loose surface layer which is then removed by wave action, generating a flat surface.

The picture below shows a wide intertidal rocky platform from near Murud on India's west coast.

India's Konkan coast is beautiful and has interesting geology too. Do visit if you can.

References:

1) Falling Late Holocene Sea-Level Along The Indian Coast- U.B. Mathur, D.K. Pandey, Tej Bahadur 2004

2) Quaternary Sea Level Changes Along Indian Coast - S.S Mehr 1992

3) Evidence of Late Holocene shoreline progradation in the coast of Kerala, South India obtained from OSL dating of palaeo-beach ridges - Linto Alappat et.al. 2015

Which Are Older? Lakshadweep Islands Or Andaman Nicobar Islands?

A friend asked me this question:

Which formed first, Andamans or Lakshadweep?

My answer was-

Lakshadweep islands, as a system of living coral reefs, lagoons and sparkling shell sand beaches, is Holocene in age (past 12 thousand  years). These coral communities rest on earlier Pleistocene reefs. So, the history of exposed reefs and atolls is a Quaternary Period phenomenon going back several hundred thousand years. Periodic polar ice cap growth and melting drove sea level fluctuations, resulting in  episodic shallow seas and vertical coral growth and reef building. Below these Pleistocene and Holocene corals lie earlier Cenozoic carbonate sediments (Source 1, 2 ) . These sediments were deposited in a subtidal marine setting, with reefs and sand shoal type environments prevailing from time to time.

We are not sure whether vertical coral growth during deposition of these earlier carbonate sequences created coral islands. It is possible that during this long Cenozoic history, there may have been episodic appearance of islands. Coral island systems and small sand shoals, environments lasting for thousands of years, would have developed due to vigorous coral growth and a static sea level, before being submerged again as sea level rose and drowned them.

And what lies below? All this Eocene to Pleistocene  (56 million to 2.5 million years) sediment sequence has been deposited on top of a Palaeocene-Eocene  (66 million to 56 million years) volcanic basement. This basement is the northern part of the Chagos-Laccadive ridge, formed when the Indian plate rode over a hot area of the mantle known as the Reunion hotspot. Below the lava is Indian Precambrian continental crust. The foundation of the Chagos-Laccadive ridge is therefore a rifted sliver of continental crust separated from the west coast shelf margin during India's separation from Africa.

The map below summarizes the setting of the Chagos-Laccadive ridge with respect to the Indian shelf margin. 

Source: Deepwater West Coast India - Pre-Basalt and Other Mesozoic Petroleum Plays: Glyn Roberts et al. 2010

Regarding Andamans.. This island chain are the central part of the Burma-Sunda-Java subduction complex in which an accretionary prism and deep sea turbidite deposits are exposed. This means the islands are made up of marine sediment and oceanic crust of a subducting slab (oceanic Indian plate) which got scraped off and plastered on to the overriding plate (oceanic South East Asian plate).

A tectonic cross section of the Andaman subduction complex is shown below.

 Source:  Mud volcanoes show gas hydrates potential in India's Andaman Islands-  Vignesh Ayyadurai et. al. 2015

Sediment and volcanic material and mafic igneous oceanic crust making up the Andaman chain may have started appearing above sea level from Eocene times (~50 million years ago).  Eocene sediment of the Mithakari Group contains detritus derived from earlier Late Cretaceous -Early Eocene ophiolites (slices of oceanic crust). This indicates that ophiolite blocks were thrust up and were exposed above sea level and were being eroded.  Such accretionary prism settings and forearc basins are cannibalistic, in that, the older deposits are emplaced above sea level and become a source of sediment for younger sequences. As tectonic plates continue to push against each other, these younger sequences in turn are moved upwards along thrust faults and become exposed above sea level. Certainly, by Pliocene times (5 million years to 2.5 million years ago), there would have been a large enough island chain.

I guess to the best of my knowledge the answer is that, although in the Lakshasdweep area, coral reef and atoll environments may be emerged above sea level episodically over the past tens of millions of years, as permanent land the Andamans are older.

One misconception I have encountered regarding Lakshadweep is that the Chagos-Laccadive ridge is a southerly extension of the Aravalli mountain chain.

This is not correct.

As I mentioned above, the basement of the ridge is likely Precambrain continental crust  which rifted apart from the southerly west coast margin of India. So, the continental crust making up the ridge would have been part of the Southern Granulite Terrain and western Dharwar craton (craton- earliest formed pieces of continental crust going back more than 3 billion years ago) of south India. The Aravalli craton and the Southern Granulite Terrain / Dharwar craton were two distinct cratonic blocks which collided and sutured by early -mid Proterozoic times (2.5 billion to 1 billion years ago). The Chagos-Laccadive ridge is oriented NNW-SSE parallel to the Indian west coast shelf margin and the Dharwar structural trends.  Post rifting, as the Indian western margin moved over the Renunion hot spot, volcanism covered this basement with lava, enhancing the ridge structure. The Chagos-Laccadive-Maldive ridge is a hotspot trail which marks the movement of the Indian plate above the Reunion hotspot.

One can imagine extending in an arcuate line the Aravalli mountain trend south to connect with the Chagos-Laccadive ridge.

Source:  The Central India Tectonic Zone: A geophysical perspective on continental amalgamation along a Mesoproterozoic suture-  K. Naganjaneyulu and M. Santosh 2010

But these were two different pieces of continental crust in the Archean. In the above figure the Dharwar and the Bhandara Cratons form a South Indian crustal block, while the Bundelkhand and Aravalli Fold Belt form the North Indian crustal block.  The Aravalli mountains terminate north of the Central Indian Tectonic Zone (shown by roughly east-west trending fault lines). This is the suture zone between the North Indian and South Indian crustal blocks.

5300 Year Old Iceman's Bacteria Genome Does Not Support Out Of India Theory

The genome of bacterium Helicobacter pylori found in  the stomach of the 5300 year old European mummy named the "Iceman" shows close similarity with Helicobacter pylori strains found in the gut of north Indians. This finding published in Science has been used as evidence to support the Out of India theory, which proposes that the Aryans and the Indo-European language family originated in India. One branch of it spread into Europe, diverging into various IE languages, while the branch which remained in India became the common  ancestor of Iranian and Sanskrit. A later migration into Iran founded the Iranian branch of the IE family.

Here is a tweet by Subhash Kak, one of the proponents of the Out of India theory.

He and others who use this finding of the bacterial genome to support this scenario are wrong.

Here's why.

Their scenario requires the European strain of Helicobacter pylori to have been derived from the Indian strain. That means people from India migrated  into Europe in the Neolithic-Early Bronze Age around five to six thousand years ago carrying with them the Indian bacterial strain which then evolved into the European variety found in the Iceman. This interpretation is demonstrably wrong. The analysis of the bacterial genomes clearly shows that the Indian strain shares ancestry with the European strain

" The resulting linked co-ancestry matrix (Fig. 4) showed that the ancient H. pylori genome shares high levels of ancestry with Indian hpAsia2 strains (Fig. 4, green boxes), but even higher co-ancestry with most European hpEurope strains".....

... "Furthermore, our co-ancestry results indicate that the Iceman’s strain belonged to a prehistoric European branch of hpAsia2 that is different from the modern hpAsia2 population from northern India".

In plain English what this mean is that the European strain has not evolved directly from the Indian strain.  Rather, the European strain and the Indian strain share an Asian common ancestor. This is clearly seen in the phylogeny (evolutionary relationship) presented in the supplementary materials of the  paper (page 50 of 88). See the image below.

Source: Supplementary Materials Maixner et al. 2016

The red arrow points to the common ancestor of the Iceman and Indian strains. The Iceman's strain and the Indian strain are sister lineages. The European strain is not derived from the Indian strain. The most sensible explanation of this finding is that from a common Asian source in the Anatolian / Near East region this bacteria spread into Europe and into India with the Neolithic expansion of farmers. It then diverged into the respective strains seen in the Iceman and in extant Indians. In Europe, this Asian derived strain then mixed with an African variety due to a migration in more recent times.

No evidence can be  found for a bronze age Out of India migration of people and languages from this study of the Iceman's bacterial genome.

Agriculture Changed Us

.. and I don't mean just culturally,  but biologically as well.

Carl Zimmer in the New York Times summarizes recent results from a wide ranging study which incorporates the genetics of extant as well as ancient Europeans. The study found evidence for several instance of natural selection altering height, digestion, skin color and our immune system.

from the article:

Previous studies had suggested that Europeans became better able to digest milk once they began raising cattle. Dr. Reich and his colleagues confirmed that LCT, a gene that aids milk digestion, did experience intense natural selection, rapidly becoming more common in ancient Europeans. But it didn’t happen when farming began in Europe, as had been supposed. The earliest sign of this change, it turns out, dates back only 4,000 years.

While agriculture brought benefits like a new supply of protein in milk, it also created risks. Early European farmers who depended mainly on wheat and other crops risked getting low doses of important nutrients.

So a gene called SLC22A4 proved advantageous as soon as Europeans started to farm, Dr. Reich and his colleagues found. It encodes a protein on the surface of cells that draws in an amino acid called ergothioneine. Wheat and other crops have low levels of ergothioneine, and the new variant increases its absorption. That would have increased the chances of survival among the farmers who had the gene.

People who are followers of the going back to a hunter gatherer Paleolithic diet fad, take note. Our digestive arsenal and our micro-biomes have responded to a different agricultural food combination. Human evolution did not freeze with the coming of the ice age. Agriculture and animal  domestication have pushed evolution into changing our DNA.

For a more detailed treatment on human evolution over the Holocene, I recommend strongly Gregory Cochran and Henry Harpending's book - The 10,000 Year Explosion: How Civilization Accelerated Human Evolution.

Harappa DNA- What Could It Tell Us About Holocene Peopling Of India

Hindustan Times carried a report a few days back on the recovery of DNA from Harappa age skeletons at Rakhigarhi village in Haryana.

What could ancient DNA tell us about the Holocene population composition of  India?

Background:

Recent genetic studies of  Indian populations  shows that Indians are a admixture of two ancient populations, the Ancestral South Indians (ASI) and  the Ancestral  North Indians (ANI). The general understanding is  that ASI has been resident  in India  since the Pleistocene, while ANI ancestry -which is related to Central and West Eurasians- was introduced in India  at various times during the Holocene. ANI and ASI are deeply divergent populations having separated from each other as early as 30 thousand to 40 thousand years ago.

ANI ancestry in Indian populations decreases along a north to south cline and from upper caste to lower caste.  Indo-European speakers have a larger component of  ANI ancestry than Dravidian speakers with North Indian upper castes showing the highest ANI ancestry.

Scenarios:

Lets assume that a representative sample of Harappa society is eventually collected. What could Harappan DNA tell us?

1) There is an absence of ANI in Harappa DNA. Harappans are unmixed ASI. This would indicate that Harappans were not Vedic Aryans. It will also have implications on how farming was introduced to the Indus valley.

2) Harappans have some ANI ancestry i.e they are a mix of ANI and ASI . This would not automatically mean that the ANI ancestry was contributed by Vedic Aryans. ANI is likely a fairly diverse group i.e. different groups of ANI after separating from West Eurasians may have  migrated into South Asia at different times in the Holocene. There is a possibility that ANI ancestry in Harappans reflects the migration of farmers (Dravidian speakers?)  from West Eurasia in the earlier part of Holocene. Moorjani et al's study indicates waves of admixture of ANI and ASI,  with middle and upper castes showing multiple layers of ANI ancestry and northern Indo European language groups shows younger admixtures dates than southern Dravidian speaking groups. These have been dated to a late and post Harappan period, although the authors say that their methods may have missed earlier admixture events. I am predicting that any ANI component in Harappans will be taken by many people as confirmation that the Vedic Aryans built the Harappan civilization.

3) Harappans have some ANI ancestry with markers suggestive of Indo-Aryan people ; One example could be the proposed West Eurasian origin -13910 C>T mutation for lactase persistence which in India  shows a northwest to southeast declining pattern. This would favor the scenario that the Vedic Aryans were a part of the Harappan civilization.  And there could be other markers typical to Indo-Aryans. Needless to say, such a finding will upset linguistic reconstructions of Indo-Aryan origins (proto-Sanskrit) thought to be not earlier than 2000 B.C. 

4) Harappans are entirely ANI. This would mean that ANI co-existed alongside ASI in the Indian subcontinent but remained genetically distinct for thousands for years until admixture in late/post Harappan times.

5) Update: November 21 2015- [ Harappans are an ASI-Austro Asiatic mix, likely speaking a Munda related language. This is a wild card entry and I am basing it on a linguistic hypothesis that there are loans words indicative of a northwest India geography in the early parts of the Rig-Ved that have phonetic similarities to Munda languages. This language substrate has been termed "Para-Munda" as it occurs really on the western most fringe of the occurrence of Munda language distribution in India, and based on its linguistic properties seems to be an early branch of the Austro-Asiatic language family.  This suggests that Indo-Aryans came in contact with resident Austro Asiatic language speakers in the Greater Punjab and Gangetic plains. Genetic work on Austro Asiatic language communities suggest a somewhat later entry (~ 2300 B.C) into East India from a Laos homeland, but Harappans just might represent an early wave of migrants from the east.]

We may get clear cut answers only when we can resolve with confidence the different layers of ANI ancestry.

I'm leaning towards Scenario (2).