Independent Sentinel-2 mapping and mobility context of the 10 September 2024 Lang Nu debris avalanche (Lao Cai, Vietnam): a reproducible open-data workflow
This paper presents a reproducible open-data workflow using Sentinel-1 and Sentinel-2 satellite imagery to independently map the September 2024 Lang Nu debris avalanche in Vietnam, confirming its 0.66 km² extent and demonstrating its exceptional mobility (2.3 times the global mean) consistent with mechanisms of saturated deposit entrainment and velocity-weakening friction.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
When heavy rain falls on steep mountains, the ground can sometimes give way, sending rocks, soil, and trees careening down into valleys below. These events, known as landslides, are among the most dangerous natural hazards on Earth, particularly in regions where the seasons bring intense storms. Scientists have long studied how far these flows travel and what makes them move so quickly. A key factor is the friction between the moving mass and the ground; if the ground is wet or if the flow picks up more material along the way, it can slide much farther than expected. Understanding these movements is vital for protecting communities living in mountainous areas, as it helps predict where a disaster might strike and how far the danger could reach.
On September 10, 2024, a catastrophic event occurred in the Lang Nu hamlet of northern Vietnam. Triggered by the intense rainfall from Typhoon Yagi, a massive avalanche of rock and debris swept down a steep valley, destroying a village and claiming sixty-seven lives. While teams on the ground and using drones later provided a detailed account of the destruction, a separate group of researchers asked a different question: could this disaster be mapped and understood just by looking at open satellite images? Their goal was not to replace the field reports but to see if freely available data from space could quickly reveal the extent of the damage and explain why the debris traveled so far.
The researchers turned to a pair of satellites that constantly orbit the Earth, capturing images of the planet's surface. One satellite, equipped with optical cameras, takes pictures in visible light, while the other uses radar to see through clouds. The team focused on a specific measure of plant health called the Normalized Difference Vegetation Index, a way of quantifying how green and healthy the vegetation is. By comparing images taken before the storm with those taken after, they looked for areas where the greenness had suddenly disappeared. A landslide strips away trees and grass, leaving a distinct scar of bare earth. The team found a clear, continuous path of vegetation loss stretching from the mountain slope down to the valley floor. This scar covered an area of about 0.66 square kilometers and extended roughly 3.6 kilometers in length.
To ensure their findings were accurate, the researchers also analyzed radar data. Unlike optical cameras, radar sends out signals and measures how much bounces back. When a landslide occurs, the smooth, wet surface of the debris reflects the radar signal differently than the rough, vegetated ground around it. The team observed a significant drop in the radar signal returning from the exact same area where they saw the vegetation loss. This independent confirmation from a different type of sensor gave them confidence that the mapped area was indeed the landslide, even though the optical images had to be taken after the storm clouds had cleared.
The study then placed this event into a broader context to understand its behavior. Using the volume of the debris reported by the field teams, the researchers calculated how efficiently the landslide moved. They found that the debris traveled about 2.3 times farther than the average landslide of a similar size would be expected to go based on global data. This exceptional mobility suggests that the flow did not just slide over the ground; it likely picked up and incorporated older, water-saturated soil and debris along its path. This process reduces the friction between the moving mass and the ground, allowing it to surge forward with great speed and distance, much like a heavy object sliding on a thin layer of water.
The researchers also examined the shape of the valley through which the debris traveled. The path began on a steep slope but flattened out significantly as it approached the village. Normally, a flow of dry rocks would stop once the slope became gentle, but this debris continued to move across a very low-gradient section for nearly a kilometer. This observation supports the idea that the flow was highly fluidized, carrying enough momentum to overcome the friction that would usually bring it to a halt. The combination of the steep upper slope and the flat valley floor created a perfect channel for the disaster to reach the settlement.
While the satellite map captured the vegetated track of the landslide, it did not show the very end of the path where the debris settled into the village, as there were no plants there to begin with. Consequently, the satellite-measured length was slightly shorter than the total distance reported by field surveys. However, the satellite data provided a rapid, independent verification of the event's scale and a clear visual record of the destruction. The researchers made their computer code and methods available to the public, creating a reproducible workflow that can be used to map future landslides in the region quickly after a storm.
This work highlights the value of using open satellite data to monitor natural hazards. It demonstrates that even without immediate access to the disaster site, scientists can quickly identify the extent of a landslide and begin to understand the physical forces that drove it. The study suggests that mapping the location of old, saturated landslide deposits in valleys could help improve future risk assessments. By knowing where these hidden layers of unstable material exist, communities might better anticipate how far a new landslide could travel, potentially saving lives in the mountainous regions of Vietnam and beyond.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.