Monthly Sargassum Wet Biomass Estimates in the Western North Atlantic from MODIS Satellite Observations

Description

This dataset provides monthly mean Sargassum wet biomass estimates (in million metric tons) from March 2000 to February 2024, derived from MODIS observations using the methodology described in Hu et al. (2023). The data cover several subregions, including the North Sargasso Sea (NSS), South Sargasso Sea (SSS), Gulf Stream Region (GSR), Antilles Current Region (ACR), Great Atlantic Sargassum Belt (GASB), and the Northwestern Gulf of Mexico (NW_GoM). The locations of these subregions are provided in the attached READ ME file. Briefly, each image pixel is classified into one of the three types using a deep-learning computer model: Sargassum containing, Sargassum free, and invalid observation (clouds, cloud shadows, strong sun glint, etc). Then the Sargassum containing pixels are further spectrally unmixed to determine the subpixel percent (%) cover within each pixel. For a pre-gridded Sargassum map, the mean Sargassum percent cover in each 0.5-degree grid is calculated as the arithmetic average of all image pixels (both Sargassum containing and Sargassum free) in that grid collected by the satellite during that calendar month. Such a mean percent cover is converted to Sargassum wet biomass using a calibration constant of 3.34 kg per square meter of Sargassum determined from field measurements (Wang et al., 2018). Integration of the grid-specific biomass across all grids within a subregion leads to the mean Sargassum biomass for that subregion. These steps were applied to MODIS/Terra (2000 – 2024) and MODIS/Aqua (2002 – 2024) separately, with the final maps being the arithmetic average between the two. In the above steps, all MODIS data were downloaded from NASA OB.DAAC (https://oceancolor.gsfc.nasa.gov) and processed using the NASA software SeaDAS (version 8.0). The deep-learning model and spectral unmixing model as well as the method to calculate monthly means were all developed at the Optical Oceanography Lab of the University of South Florida using computer codes developed in house. The daily and weekly Sargassum maps have been made available through their Sargassum Watch System (SaWS) website: https://optics.marine.usf.edu/projects/saws.html. Hu, C., Zhang, S., Barnes, B.B., Xie, Y., Wang, M., Cannizzaro, J.P., & English, D.C. (2023). Mapping and quantifying pelagic Sargassum in the Atlantic Ocean using multi‐band medium‐resolution satellite data and deep learning. Remote Sensing of Environment, 289, 113515. https://doi.org/10.1016/j.rse.2023.113515. Wang, M., Hu, C., Cannizzaro, J., English, D., Han, X., Naar, D., Lapointe, B., Brewton, R. and Hernandez, F. (2018). Remote sensing of Sargassum biomass, nutrients, and pigments. Geophysical Research Letters, 45(22), pp.12-359. https://doi.org/10.1029/2018GL078858.

Institutions

Categories

Funding

Licence