Dataset created for Multivariate Probabilistic Regression with Natural Gradient Boosting application section. Supplied for replication. Variables in the dataset: Tx, Ty, Wx, Wy: Wind Stress (Pa) (T), Wind Speed (m/s) (W) in the longitudinal (x) and latitudinal (y) directions. u_av, v_av: surface geostrophic sea water velocity (m/s) longitudinal and latitudinal respectively. lon,lat: The longitude (-180,180) and latitude (-90,90) of the drifters. t: day of year (0,366). u, v: The velocities of the drifters (m/s). Note u, v, Tx, Ty, Wx, Wy are all filtered as described in the papers supplemental information. If used the following must be acknowledged or cited appropriately. You must follows the terms of the use of the following sources: The Global Drifter Program https://www.aoml.noaa.gov/phod/gdp/interpolated/data/all.php (lon, lat, u, v) CMEMS datasets which are interpolated to the drifter locations: 1. GLOBAL OCEAN GRIDDED L4 SEA SURFACE HEIGHTS AND DERIVED VARIABLES REPROCESSED (COPERNICUS CLIMATE SERVICE) https://resources.marine.copernicus.eu/product-detail/SEALEVEL_GLO_PHY_CLIMATE_L4_REP_OBSERVATIONS_008_057/INFORMATION (u_av, v_av) 2. Global Ocean Wind L4 Reprocessed 6 hourly Observations https://resources.marine.copernicus.eu/product-detail/WIND_GLO_WIND_L4_REP_OBSERVATIONS_012_006/INFORMATION (Wx, Wy, Tx, Ty) Note this dataset will not be regularly updated. Any changes in the source products will not be brought into this dataset.