Wednesday, September 4, 2013

Geochemical Data Supports Yamuna Became Tributary Of Ganga In The Pleistocene

How long has the Yamuna been flowing along its present course in the Indian plains i.e. flowing eastwards as a tributary of the Ganga?

Exploring the temporal change in provenance encoded in the late Quaternary deposits of the Ganga Plain - Shailesh Agrawala, Prasanta Sanyal, Srinivasan Balakrishnan, Jitendra K. Dash - Sedimentary Geology April 2013

Temporal analysis of Sr isotopes in soil carbonates and Sr and Nd isotopes in silicate fractions has been carried out in a sedimentary core (Kalpi core; 50 m long) raised from the southern bank of the Yamuna river, Ganga Plain, India. The aim of the study is to constrain sediment provenance through comparison with the modern Himalayan and peninsular river systems' water and bank sediments. Sr isotopic data in soil carbonates (0.71874 to 0.71410) and Sr\Nd isotopic data in silicate (0.72865 to 0.74544 and−13.9 to−17.2, respectively) vary significantly with depth and are indicative of both Himalayan and peninsular sources for sediments in the southern Ganga Plain. The positive correlation between 87Sr/86Sr ratio and 1/Sr in soil carbonate and the negative correlation between 87Sr/86Sr and εNd in silicate confirm mixing of sediments from these sources. Variations of 87Sr/86Sr ratios in soil carbonates show that at ~80 and 45 ka the Himalaya acted as the major source of sediments in the southern part of the Ganga Plain. The gradual decrease in 87Sr/86Sr ratios after 80 and 45 ka indicates change in source to peninsular India which is also supported by limited Sr and Nd isotope data in silicates. The change in sediment provenance corresponds well with the available climatic record and is suggestive of strong climatic control in sediment supply with high supply from the Himalaya during the interglacial period and peninsular sediments during glacial period.

Background:


Presence of large dried channels in the Punjab and Haryana plains suggest that in the past the Yamuna had flowed westwards joining the Ghaggar river and was part of the Indus watershed system. Using these observations many geologists and archaeologists suggested that the Harappan civilization was being partly watered by the Yamuna. In this scenario, sometime around 2000 B.C to 1600 B.C the Yamuna shifted course and started flowing eastwards, becoming a tributary of the Ganga. At around the same time -the theory goes- the Sutlej which was also flowing into the Ghaggar shifted course away from the Ghaggar and became a tributary of the Indus. This caused large scale water stresses on the Harappans and triggered their decline and abandonment of their cities. The river Ghaggar has also been identified by many as the river Saraswati described in the Rig Ved as a mighty Himalayan river. So, a glacially connected Sutlej and Yamuna flowing into the Ghaggar until 2000 B.C or so suited those who want to establish the presence of the Vedic Aryans in the Haryana plains before 2000 B.C.

The problem was that there was no sediment provenance and sediment deposition chronology data to firmly establish the course and timing of the movement of the Yamuna and Sutlej away from the Ghaggar.

Recently, Clift et al 2012  and Giosan et al 2012 did provenance fingerprinting of ancient and recent  river sediments in the Indus plains and showed that the Sutlej and Yamuna had indeed flowed into the Ghaggar but had shifted away and started flowing along their present course in the Pleistocene, thousands of years before the Harappan civilization and the Rig Ved. The Ghaggar was not a glacially connected but a monsoonal river during Harappan times, which due to a reduced monsoon beginning around 2000 B.C. gradually dried up

Agrawala et al.'s study is specific to understanding climatic control over sediment supply and provenance and does not go into inferences about Indian civilization. This study using different geochemical techniques compliments Clift et al and Giosan et al's work. Clift et al's provenance fingerprinting and chronology suggested that the Yamuna stopped flowing into the Ghaggar sometime before 49 thousand years ago. Agrawala et al 's. work finds a Yamuna signal in the Gangetic plain as early as 80 thousand years ago.

A quick note on the use of Sr (Strontium) and Nd (Neodymium) as a way to discriminate provenance. The Sr isotope signature of the Himalayas is quite distinct from the terrain through which peninsular rivers like the Chambal flow. A large area of the Chambal drainage is covered by the mafic volcanic Deccan Traps which contain less radiogenic Sr than the Himalayan terrain made up of granitic rocks and recycled sediments. The Nd values of basalt and granite are also distinct allowing for discriminating between the river water and sediments from these terrains.

2 comments:

  1. The basic premise is that until the Yamuna joined the Ganges system, there should be NO (provenance) signal of Himalaya in the Kalpi core. In other words, Kalpi core was mainly fed by Chambal river that brought in Peninsular sediment until the time (~45 ka) when Yamuna joined the Ganges. Now as per Agrawal et al.'s, the Sr isotope ratio suggest less Himalayan and more Peninsular signal after ~45 ka which contradicts Clift et al.'s claim that Yamuna shifted to Ganges system nearly the same time. Probably Agrawal et al.'s avoided conflict by not going into drainage changes and restricted there interpretation in terms of sediment supply in response to climate change.

    ReplyDelete
  2. Agrawal et al 's data shows a Himalayan signal in the Kalpi core as early as 80 ka. This they interpret as a contribution from the river Yamuna. The signal then decreases, increases again around 45 ka and then decreases again. This is taken as an indication of the Himalayan signal becoming stronger during interglacial periods, while the glacial phases show a dominantly Penisular (Chambal) signal, since glacial phases are times of decreased weathering in the Himalaya. So, going by this data the Yamuna seems to have joined the Ganga basin much earlier than 50 ka suggested by Clift et al.

    ReplyDelete