V.N. Misra: ‘Late Harappan sites are concentrated on the tributaries of the [Sarasvatī] river, originating in the Siwalik Hills. They appear to be a consequence of the desiccation of the river and mass migration of the population to less dry regions near the Siwalik Hills and across the Yamuna.’

In contrast, changes taking place in the Saraswati Valley in the early second millennium were probably a major contributor to the Indus decline. In Harappan times, the Saraswati was a major river system flowing from the Siwaliks at least to Bahawalpur, where it probably ended in a substantial inland delta. The ancient Saraswati River was fed by a series of small rivers that rose in the Siwaliks, but it drew the greater part of its waters from two much larger rivers rising high in the Himalayas: the Sutlej and the Yamuna. In its heyday the Saraswati appears to have supported the densest settlement and provided the greatest arable yields of any part of the Indus realms. The Yamuna, which supplied most of the water flowing in the Drishadvati, a major tributary of the Saraswati, changed its course, probably early in the second millennium, to flow into the Ganges drainage. The remaining flow in the Drishadvati became small and seasonal: Late Harappan sites in Bahawalpur are concentrated in the portion of the Sarawati east of Yazman, which was fed by the Sutlej. At a later date the Sutlej also changed its course and was captured by the Indus. These changes brought about massive depopulation of the Saraswati Valley, which by the end of the millennium was described as a place of potsherds and ruin mounds whose inhabitants had gone away. At the same time new settlements appeared in the regions to the south and east, in the upper Ganges-Yamuna doab. Some were located on the palaeochannels that mark the eastward shift of the Yamuna. Presumably many of the Late Harappan settlers had originated in the Saraswati Valley.

American archaeologist Mark Kenoyer was able to prove in 1991 that the majority of archaeological sites of the so-called Harappan (or Dravidian) civilisation were not situated on the ancient bed of the Indus river, as first thought, but on the Saraswati. Another archaeologist , Paul-Henri Francfort, Chief of a franco-american mission (Weiss, Courty, Weterstromm, Guichard, Senior, Meadow, Curnow), which studied the Saraswati region at the beginning of the nineties, found out why the Saraswati had ‘disappeared’ : « around 2200 B.C., he writes, an immense drought reduced the whole region to aridity and famine » (Evidence for Harappan irrigation system in Haryana and Rajasthan -Eastern Anthropologist 1992). Thus around this date, most inhabitants moved away from the Saraswati to settle on the banks of the Indus and Sutlej rivers.

[the demographic eastward shift of the Harappan population during the decline of their cities, i.e. an intra-Indian movement from Indus to Ganga,] “is the only archaeologically documented west-to-east movement of human populations in South Asia before the first half of the first millennium BC”, while the archaeological record shows “no significant discontinuities” for the period when the Aryan invasion should have made its mark.

In the Indus period the Saraswati river system may have been even more productive than that of the Indus, judging by the density of settlement along its course. In the Bahawalpur region, in the western portion of the river, settlement density far exceeded that elsewhere in the Indus civilization . . . While there are some fifty sites known along the Indus, the Saraswati has almost a thousand . . . [The Yamuna] shifted its course eastward early in the second millennium, eventually reaching its current bed by the first millennium, while the Drishadvati bed retained only a small seasonal flow; this seriously decreased the volume of water carried by the Saraswati. The Sutlej gradually shifted its channel northward, eventually being captured by the Indus drainage . . . The loss of the Sutlej waters caused the Saraswati to be reduced to the series of small seasonal rivers familiar today. Surveys show a major reduction in the number and size of settlements in the Saraswati region during the second millennium.

This change ... is strongly suggestive of the dispersal of inhabitants, if not depopulation, of the Hakra flood plain during the Late Harappan. ... It seems almost certain that changing environmental conditions were profoundly affecting the long-established cultural pattern in Cholistan.

Marco Madella and Dorian Fuller : ‘Archaeological research in Cholistan has led to the discovery of a large number of sites along the dry channels of the Ghaggar-Hakra river (often identified with the lost Sarasvati and Drishadvati rivers of Sanskrit traditions) ... The final desiccation of some of these channels may have had major repercussions for the Harappan Civilisation and is considered a major factor in the de-centralisation and de-urbanisation of the Late Harappan period.’

The chain of tectonic events... diverted the Satluj westward [into the Indus] and the Palaeo Yamuna eastward [into the Ganga]... This explains the ’death’ of such a mighty river [ i.e, Sarasvati] ... because its main feeders, the Satluj and Palaeo Yamuna were weaned away from it by the Indus and the Ganga respectively.

It seems that during the Indus Age the Sarasvati was a large river and that water that now flows in the Yamuna and/or Sutlej Rivers made it so. Over time these waters were withdrawn and the Sarasvati became smaller, eventually dry. The agency for these changes was the tectonic reshaping of the doab [interfluve] separating the Yamuna from the rivers of the Punjab.

The importance of the Sarasvatī in Indian historical studies has multiplied manifold since archaeological analyses of the Ghaggar-Hakra river bed, combined with detailed satellite imagery of the course of the ancient (now dried up) river, conclusively showed that it had almost dried up by the mid-second millennium BCE itself, and that, long before that, it was a mighty river, mightier than the Indus, and that an overwhelming majority of the archaeological sites of the Harappan cities are located on the banks of the Sarasvatī rather than of the Indus. This has lethal implications for the AIT, which requires an Aryan invasion around 1500 BCE after the decline of the Harappan civilization, since it shows that the Vedic Aryans, who lived ―on both banks (Rigveda VII.96.2) of a mighty Sarasvatī in full powerful flow, must have been inhabitants of the region long before 1500 BCE and in fact may be identical with the indigenous Harappans. Therefore, there is now a desperate salvage operation on, in powerful leftist and "secularist" political circles in India, to put a complete full stop to any further official research on the Sarasvatī (including archaeological and geological investigations), and to launch an all-out Goebbelsian campaign through a captive media to deny that there ever was a Vedic Sarasvatī river in existence in India: the river named in the Rigveda was either completely mythical, or it was the river in Afghanistan, but it definitely was not identical with the Ghaggar-Hakra!

The Hindu mind works in such a way that continuity of worship is more important than physical fact. When the Harappans migrated eastward towards the Gangetic region, they carried with them their memories of the Sarasvati. The myths and sanctity were transferred to Prayag.

The most exhaustive, multi-year geological study on the the possible reasons for the decline of the Harappan Civilization was published in a 2012 paper titled 'Fluvial Landscapes of the Harappan Civilization' which identified a clear cause: a prolonged drought that ultimately made monsoonal rivers go dry or become seasonal, affecting habitability along their courses. To quote: 'Hydroclimatic stress increased the vulnerability of agricultural production supporting Harappan urbanism, leading to settlement downsizing, diversification of crops and a drastic increase in settlements in the moister monsoon regions of the upper Punjab, Haryana and Uttar Pradesh.'

[The desertion of the Drishadvati and the Sutlej] is typical of the instability of the river courses in the Indus plains—but in the case of the Saraswati, the effect was not localized but devastating on a major scale. Cities, towns, and villages were abandoned, their inhabitants drifting to other regions of the Indus realms and eastward towards the Ganges, pushing back the centuries-old eastern boundaries of Indus culture and venturing into uncharted territory.

The decline of Harappan urbanism probably had many contributing factors. The shift to a concentration on kharif cultivation in the outer regions of the state may have seriously disrupted established schedules for craft production, civic flood defense, building and drain maintenance, and other publicly organized works on which the smooth running of the state depended. The reduction in the waters of the Saraswati and the response of its farmers by migrating into regions to the east tore apart the previous unity of the Harappan state, disrupting its cohesion and its ability to control the internal distribution network.