New algorithm developed by USGS promises to measure the ocean’s depth with greater accuracy than ever before

Recent advancements by the USGS EROS Center, led by Minsu Kim, developed a new algorithm using Landsat data for bathymetry, supporting initiatives like Seabed 2030. Landsat Next, launching around 2030, promises even greater accuracy and resolution than ever before.

The launch of the joint Landsat satellite program on July 23, 1972, marked a significant milestone in Earth observation. While the program’s primary focus was on land surfaces, some scientists envisioned using Landsat data to explore water bodies and even the seafloor beneath.

One of the earliest and most notable researchers to test this possibility was Jacques Cousteau. In 1975, Cousteau and his team partnered with government agencies and universities to determine if Landsat 1 and 2 could measure the depth of clear, shallow ocean waters. They discovered that these satellites were indeed capable of such measurements.

Today, scientists at the USGS Earth Resources Observation and Science (EROS) Center continue to develop techniques to enhance the usability and accuracy of Landsat data for estimating water depth near shorelines. Recent research has revealed a method that works quickly and does not require calibration or separate verification of the satellite sensor’s accuracy.

To support the international Seabed 2030 initiative, which aims to map the world’s ocean seabeds, USGS geographer Jeff Danielson and Curt Storlazzi, a coastal modeler with the USGS Pacific Coastal and Marine Science Center, proposed research on using satellite imagery to map bathymetry—water depth—in select coral reef areas. Their research, led by EROS contractor Minsu Kim, focused on using Landsat data in regions like the Florida Keys, Puerto Rico, and Guam.

“This approach and this paper really demonstrated the potential strength of Landsat to do this bathymetry water depth estimation,” Kim said. “The Landsat optics and photo detector technology are state of the art. It’s one of the best sensors in terms of radiometric optical quality. That’s why we were able to reveal its capability in the water.”

Bathymetry, essentially the elevation under the water, provides crucial information for safe ship navigation, monitoring coastline changes, modeling tide conditions and coastal flooding, and studying fish and coral habitats. The new technique from EROS involves a physics-based algorithm that calculates water depth by modeling sunlight as it travels to the seafloor and back to the satellite sensor, compensating for atmospheric particles, water surface reflection, and seafloor reflectance.

Kim and another EROS contractor, Seonkyung Park, developed software to apply this technique quickly to satellite images with three to five bands of visible light, such as those captured by the Landsat Operational Land Imager (OLI).

One significant advantage of this method is that it does not require calibration, making it useful in areas without other bathymetry verification methods like lidar. However, calibration can improve accuracy further, and fine-tuning with known data from small areas can help create seamless maps of larger regions. Kim noted, “I initially thought that the maximum depth would be more limited, but the result I got here is much better. We can penetrate clean water way beyond 20 meters. That was much deeper than I expected.”

Another study supported by Danielson for the Office of Naval Research involved using WorldView satellite stereo imagery to determine depth in Key West and southwestern Puerto Rico. This method triangulated two images of the same location taken from different angles on the same day.

The earliest depth measurements were made by throwing a weighted rope overboard, but modern methods are more sophisticated. Sonar, which emits sound and measures depth by timing the returning echo, is an effective but hazardous nearshore. Lidar, using green laser pulses, measures depth from above and is safer nearshore but limited to about 70 m (230 feet) in clear water. Satellite-derived bathymetry, while limited in depth and resolution, is less expensive and time-consuming, filling critical gaps where data is sparse or nonexistent.

Danielson anticipates that Kim’s algorithm could aid in various scientific applications, including nearshore flood modeling and filling gaps along U.S. coastal waters, which are currently only 55% mapped. He emphasized the importance of remapping changing ocean coastlines and rivers over time, noting that satellite data’s temporal aspect is highly useful for creating a time series of bathymetry changes.

The Landsat archive at EROS spans over 50 years and covers the globe, offering significant potential for studying bathymetry changes over time. Currently, Landsat 8 and Landsat 9 are in use, but the USGS and NASA are planning for the next generation, Landsat Next, set to launch around 2030-2031. Landsat Next will feature three satellites with significantly more bands and higher resolution, improving from 30 to 10 m.

“I think Landsat Next is going to be a game changer for both the USGS and for satellite-derived bathymetry,” Danielson said. This advancement will enable agencies like the National Oceanic and Atmospheric Administration to integrate Landsat data into their bathymetry efforts. The additional coastal zone bands will allow for more localized measurements of water quality, improving accuracy.

References:

¹ USGS EROS Finds New Way to Measure Ocean Depth with Landsat – USGS – June 25, 2024

² Physics-based satellite-derived bathymetry (SDB) using Landsat OLI images – USGS – February 28, 2024

Featured image credit: U.S. Navy