For nearly half a century, the Ganga as it crosses Bihar has been slowly rearranging its own floodplain, shifting roughly twenty meters a year and reshaping the land on which millions depend. The satellite record from 1972 to 2020 shows a river that takes and gives back in near balance across the whole reach, yet whose losses gather stubbornly on the northern left bank. In that pattern lies a familiar human lesson: a great river's generosity is not evenly shared, and those who live closest to its moving edge bear the cost of its restlessness.
Ganga Channel in Bihar Shifted About 20 Meters a Year Between 1972 and 2020, Eroding North-Bank Land
Recurring riverbank erosion along north Bihar's left bank disrupts the livelihoods of riparian populations and threatens regional food security.
The river took and gave back close to the same amount.
Mark
The river moved 20 meters a year for nearly five decades. Why does that number matter?
Mimi
Because it's an average of a river that never stays put. Over 1972 to 2020, the channel shifted sideways 20.33 meters a year, and its shape moved about 9.73 square kilometers a year. The ground people farm is part of that movement.
Mark
But erosion and deposition nearly balance. Isn't that reassuring?
Mimi
At the scale of the whole reach, yes. 1,137.46 square kilometers eroded, 1,107.66 deposited. But land lost in one place isn't land gained in the same place.
Luke
Right, and the balance is a reach-wide total. The summary doesn't say who gains the deposited land or whether it is usable. I wouldn't read it as a net benefit to anyone.
Mark
What does the Moran's I of 0.91 tell us?
Mimi
That erosion is strongly clustered. It happens in particular stretches rather than randomly along the river.
Luke
Worth being exact: it shows spatial clustering, not why. The reporting doesn't give causes for where the clusters sit.
Mark
And the hotspots?
Mimi
7,584.20 square kilometers of significant hotspots against 5,276.48 of coldspots, with the worst bank loss on the left bank, in north Bihar.
Luke
Notice the hotspot figure is larger than the total eroded land. These are statistical zones, not area lost, so the two numbers shouldn't be compared directly.
Mark
What about the people affected?
Mimi
The study links the impacts to food security and local economies, and notes that flood erosion disrupts riparian livelihoods.
Luke
Those links are stated, but the material gives no counts of households displaced or acres of crops lost. That remains unknown here.
The Pulse
- The Ganga in Bihar has been moving sideways at about 20 meters every year, and the banks are giving way faster than the land can settle.
- Erosion has not spread thinly; it clusters tightly on the north Bihar left bank, where farmland and homes sit closest to the moving water.
- Researchers are mapping the river through satellite images and spatial statistics to pinpoint exactly where mitigation must be focused.
- Nearly 1,137 square kilometers were lost to erosion against 1,108 gained by deposition, so the river is balanced overall but unjust in its effects.
- The pressure on food security and local livelihoods is building, and planners must act on the hotspots before the next cycle of floods reshapes the land again.
For nearly half a century, the Ganga as it crosses Bihar has been slowly rearranging its own floodplain, shifting roughly twenty meters a year and reshaping the land on which millions depend. The satellite record from 1972 to 2020 shows a river that takes and gives back in near balance across the whole reach, yet whose losses gather stubbornly on the northern left bank. In that pattern lies a familiar human lesson: a great river's generosity is not evenly shared, and those who live closest to its moving edge bear the cost of its restlessness.
Between 1972 and 2020, the Ganga moved steadily across its Bihar floodplain, averaging about 20 meters a year. A satellite study published in 2026 traced those shifts over 48 years, showing that the river is not holding a fixed course but continuously reshaping its own outline across the land.
The stakes are high. The Ganga sustains roughly 600 million people through water, farming, ecology and culture, and yearly floods drive severe bank erosion that unsettles the lives of those living along its edges.
The researchers, led by N. Sahu, read the river through satellite water indices and spatial statistics. Their analysis found that erosion is not random: it concentrates in clear hotspots, especially on the northern left bank. Across the whole reach, erosion and deposition almost cancel out, but that balance hides the uneven burden falling on particular communities.
The finding points toward targeted action. The places where the river is most unstable are also where farmland and livelihoods are most exposed, and that is where mitigation planning for north Bihar must begin.
Between 1972 and 2020, the Ganga River as it crosses Bihar moved sideways across its floodplain at an average of 20.33 meters a year. That is the central finding of a satellite-based study of the river's channel, published in 2026 in Discover Applied Sciences under the title "Fluvial morphodynamics and channel-floodplain interactions of the river Ganga." The authors, led by N. Sahu with colleagues including A. Varun, S. K. Mallick, P. Das and M. Kumar, covered 48 years of change along the Bihar reach.
The river does not only shift in length terms. Measured by area, the channel's shape migrated at an average of 9.73 square kilometers per year over the same period. The authors describe this as active evolution of the river's planform, meaning the outline of the channel across the regional floodplain has been in continuous motion rather than holding a fixed course.
The stakes are large because of what the Ganga supports. It is one of South Asia's principal river systems and directly sustains roughly 600 million people, underpinning water supply, agriculture, ecological networks and cultural life. Floods recur every year across the basin and cause severe bank erosion. That erosion keeps reshaping the floodplain and disrupts the livelihoods of people who live along the river.
To trace the movement, the researchers worked from satellite observations. They derived two water indices, the Normalized Difference Water Index and the Modified Normalized Difference Water Index, and used them to delineate channel skeletons for different years. They then followed the lateral evolution of meanders and the migration of the channel using the Fluvial Corridor and Channel Migration Toolbox.
To find where the banks are least stable, the team applied spatial statistics, pairing the Getis-Ord Gi* statistic with Moran's I. The Moran's I value came out at 0.91, which the study reads as a strongly clustered pattern of bank erosion along the Bihar reach rather than a random scatter. Erosion, in other words, concentrates in particular places.
Over the full 1972 to 2020 span, the totals nearly cancel. Erosion removed 1,137.46 square kilometers of land, while deposition added 1,107.66 square kilometers across the river corridor. At the scale of the whole reach, the river took and gave back close to the same amount. That balance does not mean the losses fall evenly, and the mapping shows they do not.
The analysis identified 7,584.20 square kilometers of statistically significant erosion hotspots and 5,276.48 square kilometers of coldspots. The heaviest loss of bank land occurred on the river's left bank, which lies on the northern side. The study reports pronounced physical impacts across north Bihar, and it ties them to food security, local economic stability and the planning of environmental mitigation.
What the study points toward is targeting. If erosion is strongly clustered and concentrated on the left bank, mitigation planning for north Bihar has identifiable places to focus on, and those same places are where farmland and livelihoods are most exposed.
Notable Quotes
Bank erosion across the Bihar reach follows a strongly clustered distribution rather than a random pattern, with the most extensive loss on the left bank.— Sahu et al., Discover Applied Sciences (2026), paraphrased
Want the full story?
Read the original at Bioengineer.org ↗