Depth map of part of Florida’s stretch of coral reefs, using new bathymetry technique. NASA Earth Observatory images by Wanmei Liang, using Landsat data from the U.S. Geological Survey and data from Kim, M., et al. (2024)
This week, an article on the NASA Earth Observatory caught our eye. It described a new satellite bathymetry technique developed to map shallow coastal areas using Landsat data. Satellite bathymetry, measuring water depths from space, is an area of interest for us at Pixalytics – we are currently working on mapping seagrass, which requires a knowledge of bathymetry, and Dr Lavender previously taught Hydrography – the science of measuring and describing the physical features of oceans, seas, lakes and rivers.
The bathymetry (depth of the seabed) in coastal waters often needs to be better mapped; for example, in the United States, it is estimated that only 52 per cent of coastal waters are well mapped.
Traditional – non-satellite-based – bathymetric mapping uses both ship-based sonar and aircraft-based lidar, which can be expensive and labour-intensive. In deeper water, data can also be generated from ships of opportunity as vessels will carry an echo sounder for navigation, e.g. from research vessels. To support bringing data together, SeaBed 2030 was launched at the first-ever UN Ocean Conference in New York in 2017, and its mission is to inspire ocean mapping and deliver a complete seabed map for the benefit of people and the planet – this will be a freely available ocean map. It received another boost this week at the Hydro 2024 conference with a new partnership with Hexagon – a company that maps the seabed using airborne and satellite techniques.
The earliest paper showcasing optical imagery-based satellite bathymetry techniques was by Lyzenga (1978). These optical approaches can help map the seabed when it is shallower than the depth light penetrates. This new approach, developed by scientists at the U.S. Geological Survey, suggests it can map bathymetry down to over fifty metres in clear water. The approach is described in the paper ‘Physics-Based Satellite-Derived Bathymetry (SDB) Using Landsat OLI Images’ by Kim, M., et al. (2024), and was published in the journal Remote Sensing this year in Volume 16, Issue 5. The methodology outlined uses measurements from the Operational Land Imager (OLI) and Operational Land Imager-2 (OLI-2) on the Landsat 8 and 9 satellites, respectively – although the team notes that the technique should work for any multispectral sensor with three to five spectral bands in the visible range.
The approach applies an atmospheric correction; then, a combined water optical property and seabed reflectance model is set up with an optimization technique that searches for the best solution. The researchers note that the methodology works well in clear water with a bright bottom, where depths exceeding 20 metres can be mapped. When the location has sediment, phytoplankton, or seagrass on the seafloor, calculating the depth is more challenging – although the team did develop a correction for these types of waters/seabeds. Guam, Key West in Florida, and Puerto Rico were test sites, and the image at the top of the blog shows the bathymetric map for the Florida Keys produced using this technique, with shallow channels cutting between the islands, called keys, clearly seen. The mapping of coral reefs was deliberate as these environments are vital habitats and alter over time. Therefore, they’d benefit from regular remapping.
Summary
It is exciting to see a new innovative approach to mapping coastal waters from Landsat, and there are obvious potential benefits in improved monitoring of these habitats and the ecosystems they support.
