In September 2026 the Ganga rose above its previous highest flood level at three gauges in Bihar. The published evidence points to several rivers peaking together, with large releases from barrages and reservoirs, rather than to a single distant cause.
On the morning of 5 September 2026 the Ganga at Gandhi Ghat in Patna stood at 50.53 metres, one centimetre above the previous highest flood level of 50.52 metres set on 21 August 2016. By 06:00 the next day it had reached 50.57 metres. Two other gauges also exceeded their earlier marks: Pandarak at 43.92 metres against 43.73 metres in August 2021, and Sultanganj, near Bhagalpur, at 36.81 metres against 36.75 metres on 8 September.
The word record needs care. These are the highest water levels recorded at those gauges. They do not necessarily mean the greatest discharge, the largest affected population or the worst damage in Bihar’s history.
How to read a gauge
A danger mark is the level at which flooding is expected to begin threatening land and people, and the highest flood level is the maximum recorded at a given gauge. Both are stage measurements, in metres above a fixed datum, and neither is a flow. The relationship between stage and discharge changes as a riverbed changes, and the Ganga carries heavy sediment. If the bed has risen, a given flow can now produce a higher stage than it once did, a possibility the published record does not yet allow readers to test. A cusec, the unit used for barrage releases, is a cubic foot a second, about 0.028 cubic metres a second.
The rivers that met
The Ganga at Patna is the product of many tributaries. The Ghaghara joins upstream, the Son enters at Maner and the Gandak flows in from the north opposite Patna at Hajipur. The Kosi joins the Ganga farther downstream near Katihar, and so influences the reach below Bhagalpur rather than the gauge at Patna. In early September the Gandak at Dumriaghat, the Kosi at Baltara, the Burhi Gandak at Khagaria, the Punpun, the Ghaghara at Darauli and several Ganga stations were all above danger marks.
The volumes were large. On 2 September the Son carried about 5.23 lakh cusecs, the highest of the year, roughly 14,800 cubic metres a second. The Kosi barrage at Birpur passed about 1.9 lakh cusecs, and the Gandak barrage at Valmikinagar about 1.5 lakh cusecs, falling to 1.26 lakh by 5 September. Reports said all 36 gates of the Gandak barrage were open, 56 of 69 at Indrapuri on the Son, and all 109 at Farakka.
The causes reported
Reporting by India.com listed three causes: water released from the Bansagar dam in Madhya Pradesh into the Son; heavy rain in Uttarakhand and Himachal Pradesh feeding the Yamuna and its tributaries; and glacial lake bursts in the Nepal Terai and on the Tibet–Nepal border. Rain in Bihar itself was substantial. The state received 60.8 millimetres in September up to the 5th, which was reported as 69 per cent above normal.
No published evidence supports the third cause. ICIMOD’s assessment of the 26 August collapse in Nepal found no glacial lake at its origin, and the sources reviewed contain no report of another lake outburst in the Terai or on the border in this period.
Could the August event in Nepal have contributed?
There is a real geographical link. The Trishuli, which carried the 26 August debris flood, joins the Kali Gandaki at Devghat to form the Narayani, the river that enters India as the Gandak. But the timing and size argue against a substantial contribution. The international analysis of the collapse estimated about 20 million cubic metres of excess water passing Devghat in under four hours. That works out at an average of roughly 1,400 cubic metres a second, or about 50,000 cusecs, briefly. The Gandak barrage was passing 1.26 to 1.5 lakh cusecs one to two weeks later, when that pulse had long gone.
This is an editorial calculation with rough inputs, and it does not rule out lasting effects. Sediment deposited in the channel might change how the river carries floods. But no published evidence yet shows that the Nepal event, or Himalayan cryospheric change more broadly, contributed to Bihar’s September flood, and the reported drivers are monsoon rain and reservoir and barrage releases across several basins.
Embankments, cities and people
Embankments are central to Bihar’s flood management, and their behaviour shaped the damage. Reports described more than 350 metres of embankment damaged at Bhagalpur, with a railway line threatened, and a 20-metre breach in the Mahi river embankment near Trilokchak in Saran district. Deaths from drowning were confirmed at Danapur and Maner in Patna district, and a boat carrying ten people capsized at Raghopur Diara, though all were rescued.
Patna sits on a large river with tributaries entering from several directions, so the city’s flood exposure is shaped by rivers well outside its own catchment. Bihar’s rivers are embanked over long stretches. Embankments protect the land behind them but confine the water and its sediment, and when they fail or are overtopped the flood can arrive suddenly.
The scale of human impact grew through the week. On 5 September about 10.26 lakh people were reported affected across 11 districts, 50 blocks and 201 gram panchayats, with Bhojpur worst hit at 2.84 lakh. By 8 September more than 3.4 million people were reported affected in 13 districts, with five confirmed deaths. Some outlets described four million people as displaced, which conflates affected with displaced. The relief effort deployed 858 boats, 27 state and 7 national disaster response teams and 38 community kitchens.
Reliable figures on crop damage, economic loss and migration were not available in the sources reviewed. Settling the causes will need Central Water Commission gauge and discharge series, India Meteorological Department basin rainfall, reservoir release logs and post-flood sediment surveys.
–BP Padala
