Presented at the 2020 Ocean Sciences MeetingUpper-ocean stratification, together with wind and current shear, control vertical mixing and the associated momentum, heat, freshwater, and nutrient fluxes between the thermocline and mixed layer. Near-surface stratification is particularly strong in tropical regions and is crucial for studying coupled phenomena across multiple time scales, including tropical cyclones, MJO, IOD, and ENSO. Upper-ocean stratification is mainly determined by temperature except in areas with strong precipitation, such as the Inter-Tropical Convergence Zone, or in river plumes. In-situ observations of the tropical ocean have significantly increased in the past decade yet are still too sparse to resolve ocean stratification variability in near-real-time and on small spatial scales. In this study, based on long-term observations and an ocean reanalysis dataset from 2004-2017, we first investigate the possibility of retrieving upper-ocean stratification from surface data using a simple regression method. We found that more than 90% of the mean seasonal cycle and about 30% to 70% of temperature and salinity stratification anomalies can be reconstructed using surface data from either observations or the reanalysis. We then replace the in-situ sea surface temperature (SST) and salinity (SSS) with satellite observations to create a high-resolution upper-ocean stratification dataset. While the use of satellite SST in the regression model gives similar performance compared to the use of in situ SST, satellite SSS results in significantly worse performance in the regression model. A method is created to improve the satellite salinity data and the associated inference of upper-ocean salinity stratification. The resultant stratification dataset is found to perform significantly better than a commonly used near-real-time numerical ocean model analysis and offers promise for improved real-time characterization of upper-ocean stratification.