Remote sensing as a tool for assessing landslide impacts on coastal waters at Porchat Island (SE Brazil)

Coastal waters are rarely the still, clear mirrors often seen in postcards; they are dynamic, living systems where land and sea constantly exchange materials. When rain falls on a city built on steep hills, water runs off the pavement, picking up soil and debris before spilling into the ocean. This influx of sediment changes the water's color and clarity, a process known as turbidity. In healthy ecosystems, these changes are usually gradual, but when a landslide occurs, the shift can be sudden and massive, dumping tons of earth into the sea in a single moment. Scientists have long known that such events happen, but tracking them is difficult. Traditional methods require researchers to take boats out and collect water samples, a slow process that often misses the peak of the event. Today, however, satellites orbiting overhead offer a different perspective, capturing high-resolution images of the Earth's surface every few days or even every day, allowing us to watch these underwater plumes form, spread, and fade in near real-time.

On April 29, 2025, a dramatic example of this interaction unfolded on Porchat Island, a densely populated area on the coast of São Paulo, Brazil. A water pipeline belonging to the state's sanitation company ruptured, releasing a massive volume of water onto a steep, unstable slope. The force of the water triggered a landslide, displacing an estimated 860,000 cubic meters of soil, rocks, and vegetation. This debris did not stay on land; it slid down into the adjacent coastal waters, creating a visible cloud of sediment that spread across the bay. To understand exactly how this event altered the ocean, researchers turned to satellite imagery, using two different types of sensors to monitor the water before and after the rupture. Their goal was to measure how much the water became cloudy, how deep light could penetrate, and how the sediment moved in response to the wind and waves.

The team used data from two satellite systems: Sentinel-2, which provides detailed images every five days, and PlanetScope, which offers a daily view of the same location with even finer detail. By processing these images with specialized software designed to correct for atmospheric interference, the scientists could calculate the concentration of suspended matter in the water. Before the landslide, on April 15, the water near the island contained an average of 11.15 grams of suspended matter per cubic meter. After the event, on May 2, this number had nearly doubled to 21.09 grams per cubic meter. The water also became significantly less transparent. The depth at which a standard white disk, known as a Secchi disk, could be seen dropped from 2.01 meters to just 1.12 meters. These changes were most severe right next to the island, while the neighboring bays saw much smaller shifts, confirming that the landslide was the primary source of the disturbance.

One of the most valuable aspects of this study was the ability to track the movement of the sediment cloud over time, a feat made possible by the daily images from the PlanetScope satellite. On April 30, the day after the rupture, the sediment plume covered an area of 7.50 square kilometers, a vast expanse of murky water spreading out from the island. However, the cloud did not stay large for long. By May 2, the affected area had shrunk to 2.02 square kilometers. By May 4, the plume was no longer visible in the satellite images. This rapid disappearance was not due to the sediment vanishing, but rather to the natural forces of the ocean. During this period, the wind and waves were relatively weak, and the direction of the waves shifted, pushing the sediment toward the center of the bay where local currents diluted and dispersed the material. The study suggests that the water's natural energy was sufficient to mix the sediment back into the wider ocean, clearing the surface view within just a few days.

The researchers also looked at the water's biological health, specifically measuring chlorophyll-a, a pigment found in algae that indicates the presence of plant life. They found that the levels of this pigment varied from day to day, but there was no clear spike or drop that could be directly blamed on the landslide. The data showed that the sediment cloud did not immediately trigger a massive bloom of algae or a sudden die-off, though the scientists noted that the presence of so much mud can sometimes interfere with the satellite's ability to measure algae accurately. Without direct water samples to confirm what was happening beneath the surface, the team could only interpret the satellite data as relative changes rather than absolute facts. They also noted that the heavy rains in the days leading up to the landslide may have already saturated the soil, making the ground unstable even before the pipe burst, and that the initial images taken on April 30 were partially obscured by clouds, which could have slightly distorted the measurements of the plume's size.

This event highlights a critical gap in how we monitor coastal environments. Conventional monitoring programs often rely on scheduled visits that might miss a sudden disaster entirely. In this case, the landslide was a direct result of urban infrastructure failure interacting with natural geological risks, a combination that is becoming more common in coastal cities. The study demonstrates that high-resolution remote sensing can act as a rapid response tool, allowing authorities to see exactly where the damage is, how far it spreads, and how quickly it clears. While the surface water returned to its normal appearance within a week, the study cautions that the sediment likely settled on the ocean floor, where its long-term effects on the seabed and marine life remain unknown. The work serves as a powerful reminder that while satellites can tell us a great deal about the surface of the ocean, understanding the full story of a coastal disaster often requires combining these high-tech views with on-the-ground investigation.