The Ghaggar rises in the Shivaliks near Kalka, runs south-west past Patiala, Sangrur, Mansa and Sirsa, and dies in the sand of Rajasthan. In Pakistan the same dry channel is called the Hakra and runs through Cholistan.
It is seasonal. It floods in the monsoon and is a bed of sand the rest of the year, and it floods badly, with major events in 1993, 2010 and 2023.
Hundreds of Harappan sites lie along the Ghaggar-Hakra channel, more than along the Indus itself. Kalibangan, Rakhigarhi and Banawali are on it. That distribution is the strongest evidence that the channel once carried a large perennial river.
Whether that river is the Saraswati of the Rig Veda, when it dried, and whether it was fed by the Sutlej or the Yamuna before those rivers changed course, is one of the longest running disputes in Indian archaeology. It has become politically charged, which has not helped the evidence.
For the villages of southern Punjab the Ghaggar is not a controversy. It is the thing that takes their fields every few years.
Its feeders are the Markanda, the Tangri, the Sarsuti and the Dangri, all Shivalik choes that run only when it rains in the hills. They deliver a year's water in a few weeks and the Ghaggar has to take the whole of it at once.
It is a drain as much as a river. Sewage and effluent from Panchkula, Zirakpur, Dera Bassi, Patiala and Sangrur go into it or into the choes that feed it, and stretches of it appear on the national list of the most polluted river reaches in India.
Below Sirsa what is left reaches the Ottu weir, and what the weir lets out runs into Hanumangarh and Sri Ganganagar, where the channel is dry in most years. The last of the water is used before the sand can have it. The river has no mouth.