A Google Earth Engine based package to delineate river centerline and width from Landsat Collection 1 images. The algorithm is similar to RivWidth software developed by Dr. Tamlin Pavelsky, though certain processing steps were modified so that it can be run on Google Earth Engine. Tough not officially supported, the code can be adapted to extract centerlines and widths from binary river masks derived from any image sources.

See GitHub page for software instructions. Associated publication can be found here.

A Google Earth Engine based classification algorithm to identify lake surface ice cover conditions for Landsat Collection 1 images. Note that since the publication of the original algorithm, USGS has discontinued the production of Collection 1 data. To use the SLIDE algorithm on Collection 2 data, one needs to retrain the algorithm before applying it.

See Github page for instruction. Publication here.

Lake color (dominant wavelength, nm) for 85,360 lakes worldwide, sampled at lake center for all Landsat 8 overpasses between 2014 and 2020. The data can be accessed from Zenodo and accompanying paper at here.

GROD is a manually labelled global database of human-made river obstruction/barrier database. It contains geolocations of ~33k river obstructions. These obstructions were placed into one of the five categories based on the topdown view of their visual appearance on high resolution satellite images. The data can be accessed from Zenodo and accompanying paper at here.

The collection of data is ongoing and is partnered with Dr. Kevin Rose at the Darrin Fresh Water Institute, NY. The above water radiometric measurements are being collected with the So-Rad instrument. Data collection has commenced since July 2024.