NOAA’s SatBathy Tool for Automating Satellite Derived Bathymetry

January 11, 2024  |  2:00-3:00 pm, ET

Presenters: Gretchen Imahori, Management and Program Analyst, Remote Sensing Division; Bryan Eder, Marine GIS Analyst, National Centers for Coastal Ocean Science (NCCOS); LT Matt Sharr, NOAA Corps Officer, Remote Sensing Division.

Our current multi-temporal Satellite-Derived Bathymetry (SDB) method incorporates a robust atmospheric correction, applies a physically-based algorithm, a multi-scene compositing method to reduce the effect of turbidity, and a switching model to improve mapping in shallow waters. Beyond defining a time range for imagery and utilizing a simplified calibration method, there is little to no manual input required to generate SDB products. Furthermore, this method has shown SDB median errors of 0.5m for depths ranging from 0-25m when validated with lidar surveys using 10m resolution imagery for the east coast of the United States.

In 2020, NOAA created a Graphical User Interface (GUI), SatBathy (now beta v2.1.6), designed to provide consistent and quick results utilizing this research. Initially, the development of the GUI was primarily performed using the Bokeh visualization Python library and required expanding on previous research to adapt the existing algorithm from MATLAB to Python and to create scripts to automate image collection and atmospheric correction. Now we have transitioned the GUI from Python’s Bokeh visualization library to the widely used and customizable Flask web framework.

While NOAA’s SatBathy tool is still in its research phase, SatBathy is a hybrid web/desktop tool which relies upon only open source technology. It utilizes 10 meter resolution satellite imagery from the Copernicus Sentinel-2 mission, and recently changed from Google Cloud Platform to now using Amazon Web Service (AWS), and ACOLITE atmospheric correction processor.

Its interactive GUI gives users the ability to create AOIs on-the-fly, query and preview imagery from different time periods, adjust for cloud coverage, review image metadata, and generate SDB products. SatBathy also is able to transform “pseudo” bathymetry (pSDB) which is the dimensionless algorithm output to final bathymetry in meters through the automation of a vertical referencing step in which a linear relationship between pSDB and ground truth data is established.

Today’s webinar will discuss the SatBathy tool, latest capabilities, short- and long-term plans and some of the on-going research to help support its evolution.