Sea-surface solar radiation in the visible wavelength (400-700nm), abbreviated photosynthetically available radiation (PAR), is an important parameter for the estimation of marine primary production and plays a key role in controlling the phytoplankton dynamics, upper ocean physics and biogeochemical processes. Because of the scarcity of continuous in-situ PAR measurements over the world oceans, remote-sensing algorithms have been developed to estimate PAR values at sea surface both spatially and temporally. However, these algorithms introduce significant uncertainties under clear/cloudy conditions and complex atmospheric/meteorological conditions. In the present study, an enhanced solar irradiance model (ESIM) is developed to estimate downwelling radiation reaching at just above the ocean surface. The ESIM uses direct and diffuse components in conjunction with a cloud-correction technique to produce the daily PAR (DPAR) values under various atmospheric conditions. The performance of ESIM was tested on MODIS-Aqua time series data and validated using concurrent in-situ observations and also systematically compared with two other algorithms. Findings showed that the PAR values estimated by ESIM agreed better with in-situ PAR observations (to within mean relative error 13.08% and mean net bias 3.77 Em-2day-1) than those of other algorithms. To demonstrate the efficiency of ESIM, a time series of GOCI observations were processed to obtain the seasonal variations for the winter (January 2019) and summer (July 2019) periods from Level-3 binned MODIS-Aqua PAR products (global gridded). In comparison with the standard PAR algorithms, the ESIM was observed to overcome the challenge of variable cloud cover across latitudes (up to 60°) and make significant improvements in the DPAR values for the global oceans regardless of the seasons and atmospheric conditions.