It is often identified with the Vedic Sarasvati River, but it is disputed whether all Rigvedic references to the Sarasvati should be taken to refer to this river. Many references to this river are mythical and refer to the Indian Epics and Puranas .
The present-day Sarsuti Sarasvati River originates in a submontane region (Ambala district) and joins the Ghaggar near Shatrana in PEPSU. Near Sadulgarh (Hanumangarh) the Naiwal channel, a dried out channel of the Sutlej, joins the Ghaggar. Near Suratgarh the Ghaggar is then joined by the dried up Drishadvati (Chautang) river.
The wide river bed (paleo-channel) of the Ghaggar river suggest that the river once flowed full of water, and that it formerly continued through the entire region, in the presently dry channel of the Hakra River, possibly emptying into the Rann of Kutch. It supposedly dried up due to the capture of its tributaries by the Indus and Yamuna rivers, and the loss of rainfall in much of its catchment area due to deforestation and overgrazing. This is supposed by some to have happened at the latest in 1900 BCE, but is much earlier
Puri and Verma (1998) have argued that the present-day Tons River was the ancient upper-part of the Sarasvati River, which would then had been fed with Himalayan glaciers. The terrain of this river contains pebbles of quartzite and metamorphic rocks, while the lower terraces in these valleys do not contain such rocks. However, a recent study shows that Bronze Age sediments from the glaciers of the Himalayas are missing along the Gagghar-Hakra, indicating that it did not have its sources in the high mountains.
Many settlements of the Indus Valley Civilisation have been found along the Ghaggar and Hakra rivers.
According to palaeoenvironmental scientists the desiccation of Sarasvati came about as a result of the diversion of at least two rivers that fed it, the Satluj and the Yamuna. "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 (the Sarasvati) … because its main feeders, the Satluj and Palaeo Yamuna were weaned away from it by the Indus and the Gangaa respectively”. This ended at c 1750 b.c., but it started much earlier, perhaps with the upheavals and the large flood of 1900 b.c., or more probably 2100 b.c. . P H Francfort, utilizing images from the French satellite SPOT, finds that the large river Sarasvati is pre-Harappan altogether and started drying up in the middle of the 4th millennium BC; during Harappan times only a complex irrigation-canal network was being used in the southern region of the Indus Valley. With this the date should be pushed back to c 3800 BC. R. Mughal (1997), summing up the evidence, concludes that the Bronze Age Gagghar-Hakra sometimes carried more, sometimes less water (for example from the Sutlej). The latter point agrees with a recent isotope study
The Rig Vedic hymn X, however, gives a list of names of rivers where Sarasvati is merely mentioned while Sindhu receives all the praise. It is agreed that the tenth Book of the Rig Veda is later than the others. Some think that this may indicate that the Rig Veda could be dated to a period after the first drying up of Sarasvati (c. 3500) when the river lost its preeminence. The assumption is contradicted by the appearance of horses and chariots all over the RV, which was possible only after their introduction after 2000 BCE.
The 414 archeological sites along the bed of Sarasvati dwarf the number of sites so far recorded along the entire stretch of the Indus River, which number only about three dozen. However most of the Harappan sites along the Sarasvati are found in desert country, undisturbed since the end of the Indus Civilization. This contrasts with the heavy alluvium of the Indus and other large Panjab rivers that have obscured Harappan sites, including part of Mohenjo Daro. About 80 percent of the Saravati sites are datable to the fourth or third millennium B.C.E., suggesting that the river was flowing during this period.
Along the course of the Ghaggar-Hakra river are many archaeological sites of the Indus Valley Civilization; but not further south than the middle of Bahawalpur district. It has been assumed that the Sarasvati ended there in a series of terminal lakes, and some think that its water only reached the Indus or the sea in very wet rainy seasons. However, satellite images contradict this: they do not show subterranean water in reservoirs in the dunes between the Indus and the end of the Hakra west of Fort Derawar/Marot. It may also have been affected by much of its water being taken for irrigation.
In a survey conducted by M.R. Mughal between 1974 and 1977, over 400 sites were mapped along 300 miles of the Hakra river. The majority of these sites were dated to the fourth or third millennium BCE.
S. P. Gupta however counts over 600 sites of the Indus civilization on the Hakra-Ghaggar river and its tributaries. In contrast to this, only 90 to 96 Indus Valley sites have been discovered on the Indus and its tributaries (about 36 sites on the Indus river itself.) V.N. Misra states that over 530 Harappan sites (of the more than 800 known sites, not including Late Harappan or OCP) are located on the Hakra-Ghaggar. The other sites are mainly in Kutch-Saurashtra (nearly 200 sites), Yamuna Valley (nearly 70 Late Harappan sites) and in the Indus Valley, in Baluchistan, and in the NW Frontier Province (less than 100 sites).
Most of the Mature Harappan sites are located in the middle Ghaggar-Hakra river valley, and some on the Indus and in Kutch-Saurashtra. However, just as in other contemporary cultures, such as the BMAC, settlements move up-river due to climate changes around 2000 BCE. In the late Harappan period the number of late Harappan sites in the middle Hakra channel and in the Indus valley diminishes, while it expands in the upper Ghaggar-Sutlej channels and in Saurashtra. The abandonement of many sites on the Hakra-Ghaggar between the Harappan and the Late Harappan phase was probably due to the drying up of the Hakra-Ghaggar river.
Painted Grey Ware sites (ca. 1000 BCE) have been found on the bed and not on the banks of the Ghaggar-Hakra river.
Because most of the Indus Valley sites known so far are actually located on the Hakra-Ghaggar river and its tributaries and not on the Indus river, some archaeologists, such as S.P. Gupta, have proposed to use the term "Indus Sarasvati Civilization" to refer to the Harappan culture which is named, as is common in archaeology, after the first place where the culture was discovered.
Paleobotanical information also documents the aridity that developed after the drying up of the river. (Gadgil and Thapar 1990 and references therein). The disappearance of the river may have been caused by earthquakes which may have led to the redirection of its tributaries. It has also been suggested that the loss of rainfall in much of its catchment area due to deforestation and overgrazing in what is now Pakistan may have also contributed to the drying up of the river. However, a similar phenomenon, caused by climate change, is seen at about the same period north of the Hindu Kush, in the area of the Bactria-Margiana Archaeological Complex.
At Ropar the Sutlej river suddenly flows away from the Ghaggar in a sharp turn. The beforehand narrow Ghaggar river bed itself is becoming suddenly wider at the conjunction where the Sutlej should have met the Ghaggar river. And there is a major paleochannel between the point where the Sutlej takes a sharp turn and where the Ghaggar river bed widens.
In later texts like the Mahabharata, the Rigvedic Sutudri (of unknown, non-Sanskrit etymology is called Shatudri (Shatadru/Shatadhara), which means a river with 100 flows. The Sutlej (and the Beas and Ravi) have frequently changed their courses. The Beas has also probably sometimes flown into the Sutlej further downstream from where it joins that river today. Before that Sutlej is said to have flown into Ghaggar
Scholars like Raikes (1968) and Suraj Bhan (1972, 1973, 1975, 1977) have shown that based on archaeological, geomorphic and sedimentological research the Yamuna may have flowed into the Sarasvati during Harappan times. There are several dried out river beds (paleochannels) between the Sutlej and the Yamuna, some of them two to ten kilometres wide. They are not always visible on the ground because of excessive silting and encroachment by sand of the dried out river channels. The Yamuna may have flowed into the Sarasvati river through the Chautang or the Drishadvati channel, since many Harappan sites have been discovered on these dried out river beds.