The dataset contains ARGO data, CMEMS model results and data emulated by Neural networks (NN) used in the study 'Dynamics of the euphotic zone in the Black Sea: The synergy of data from profiling floats, machine learning and numerical modeling' - NNalongARGO.tar.gz: archive contains Matlab binary files consisting of NN-derived BGC variables (Chlorophyll-a, Oxygen and backscatter at 700 nm) along ARGO float pathes together with ARGO data themselves (time, latitude, longitude, salinity, temperature, sigma_T and BGC variables); all variables have 1m vertical resolution; depth range is [1m 150m] ; additionally, float ogs7 data include NO3, float hzg1 data exclude Chlorophyll-a and backscatter at 700 nm; (filenames are floatname_euph_1mRes_150mALLINCLNN.mat) - NN2018_CMEMS_ARGO.tar.gz: archive contains Matlab binary files consisting of NN-derived BGC variables (Chlorophyll-a, Oxygen and backscatter at 700nm) along ARGO float pathes in 2018 with vertical resolution taken von CMEMS (13 depth levels ); NN was applied either on CMEMS physics ('C') or on ARGO physics ('A'); additionally, CMEMS BGC model data (Chlorophyll-a and Oxygen) along these pathes are included; (filenames are floatname_2018_NNARGOCMEMS.mat) - NNReconBlackSea.tar.gz: archive contains Matlab binary files consisting of basin wide NN derived BGC variables (Chlorophyll-a, Oxygen and backscatter at 700nm) for the years 2015-19 and 2010/11 (weekly mean data); additionally CMEMS data (time, latitude, longitude, salinity, temperature, sea surface height) are provided at CMEMS model upper 15 depth levels; (filenames are reconNNCMEMS_2015_2019.mat and reconNNCMEMS_2010_2011.mat, respectively)