WHEN THE MOUNTAIN SENDS WATER DOWNSTREAM: A GIS PERSPECTIVE ON NEPAL’S POST-FLOOD LANDSCAPE

A map feels permanent. Roads have lines, rivers have boundaries, bridges connect two sides, and infrastructure sits exactly where the database says it should.

Then a major flood arrives and suddenly the map is only a record of how the landscape used to be.

That is one of the striking lessons from Nepal’s recent flood disaster. At the Chilime Hydropower Project along the Trishuli River, water carrying silt, rock and debris entered the underground powerhouse and access tunnel. At the same time, damaged roads and bridges made it extremely difficult for rescue teams and equipment to reach the site.

The event raises a simple but important question:

After a flood, how much of yesterday’s map can we still trust?

The before-and-after satellite images tell part of the story immediately.

Before the event, the river corridor appears relatively defined, with roads, structures and developed areas clearly visible along its edge. In the post-event image, the same landscape looks fundamentally different. The river corridor has expanded, large areas are covered by sediment and debris, and features that once appeared clearly separated from the river are now within a dramatically altered surface.

This is where Remote Sensing becomes especially powerful not because it tells us everything, but because it lets us quickly see where the landscape has changed.

A road may still appear in an existing GIS database even though part of it has disappeared. A bridge may still be standing, but its approach may no longer be usable. A riverbank may have shifted. Sediment may have buried previously accessible land. An infrastructure site that was once connected to the road network may suddenly become isolated.

At Chilime, that isolation became one of the defining challenges of the response. The hydropower management described the situation simply:

“We are isolated on all sides.”

That statement is not only about rescue logistics. It is also a spatial problem.

A flood does not need to destroy an entire road network to create major disruption. Sometimes one failed bridge, one washed-out road section, or one blocked access point can disconnect an entire facility or community. GIS-based network analysis can help identify these breaks, possible alternative routes, and which connections should be restored first.

This is why post-flood GIS should go beyond drawing the extent of inundation.

The more useful questions are often:

What moved? What disappeared? What became disconnected? What is now closer to the river than before? And where should field teams investigate first?

Remote Sensing may indicate that a riverbank has shifted or a road has been disrupted, but it cannot always tell us whether a structure is safe, how much elevation has changed, or whether a slope remains stable. That is where different geospatial tools begin to work together. Satellite imagery can rapidly screen the affected landscape. GIS can relate those changes to roads, settlements, bridges and infrastructure.
Drone photogrammetry can provide detailed orthophotos and 3D surface models.
RTK-GNSS, Total Station and digital surveying can then measure the precise geometry and elevation needed for engineering assessment.

The real question comes next:

Should everything simply be rebuilt exactly where it was?

If the river has shifted, if a slope has become unstable, or if an access corridor has repeatedly failed, reconstruction may need to do more than restore the previous condition.

A road may need realignment. A bridge approach may need redesign. A riverbank may require protection. Critical infrastructure may need a more resilient access route. Some locations may even need to be reconsidered entirely.

This is where GIS moves from documenting damage to supporting better decisions. By combining the changed landscape with information on roads, settlements, infrastructure and terrain, GIS can help identify where rebuilding is suitable, where redesign may be needed, and which locations should be prioritized first.

Nepal’s mountainous terrain is very different from Bangladesh’s low-lying deltaic landscape, but the lesson is relevant to both: after a major flood, the most useful map may not be the one showing where the water reached. It may be the map showing what the water changed. The real value of post-disaster GIS is not simply recording where the flood occurred. It is helping us understand the new landscape that planners, engineers and communities must live with and plan for next.