Analysis of sea surface salinity over northern Atlantic

Supervisors: Jon Saenz, Ganix Esnaola (UPV/EHU)
Ocean salinity represents the ratio of dissolved solids against the mass of seawater, usually expressed in grams per kilogram. The ocean’s salinity is key to studying the water cycle and ocean circulation, which are important to Earth’s climate. The average salinity of the world’s oceans is 35 ‰. In previous time people research of the global ocean salinity mainly by in-situ data. After the professional salinity sensing satellites were launched, the situation was improved and scientists have more methods to analyse the salinity distribution of the ocean. The purpose of the thesis is to use Aquarius salinity data from August 2011 to March 2015, to analyse the salinity characteristics of northern Atlantic and compare the satellite data to other in-situ and ocean reanalysis data. The results show that World Ocean database has the best multiplicative bias performance and scores and Levitus gets the worst. Representative error and seasonal cycle is the mainly reason of large amount of outliers in the box-plots, and for scatter plots the agreement is better for Levitus than for WOD, Levitus gets one of the best R-square and RMSE scores. In general, the observations of SSS by Aquarius provide a good observational dataset of the salinity in the North Atlantic and is complementary to other satellite data and it-situ data. For future study, EOF (Empirical Orthogonal Function) or principal component analysis need to be used as analysis methods in order to study the evolution with a reduced dimensionality.