Observations of Near-Surface Temperature and Salinity from Profiling Floats: Vertical Variability, Structure, and Connection to Deeper Properties
Anderson, Jessica Erin
MetadataShow full item record
In response to the changing climate, it is probable that there is an intensification of the oceanic hydrologic cycle underway. Due to the difficulty of observing the ocean, especially very close to the sea surface, we lack the detailed observations of ocean salinity (a proxy for the oceanic freshwater cycle) necessary to fully understand and model the complex dynamics governing the ocean’s role in the hydrologic cycle and how it might change in the future. While relatively new measurements of sea surface salinity from satellites (Aquarius/SAC-D, SMOS, and SMAP) have enhanced our understanding of the freshwater cycle over the ocean, sea surface satellite maps integrate over relatively large scales and blur the short-lived, small-scale variability (where precipitation occurs) that remains poorly constrained. This dissertation explores the magnitude, frequency, and structure of such near-surface temperature and salinity variability and how it is connected to larger-scale, subsurface ocean properties. The first two chapters of this dissertation utilize observations obtained from profiling floats equipped with auxiliary Surface Temperature and Salinity (STS) sensors. These novel instruments allow for high vertical resolution (10 cm) examination of near-surface (~0.2-30 m) temperature and salinity stratification usually not possible in the context of Argo. Observations from STS equipped Argo-type floats deployed in the tropical and subtropical Pacific, Atlantic, and Indian Oceans show that the upper 4 m of the ocean are well-mixed most of the time (87% for temperature, 97% for salinity), generally associated with wind speeds > 6 m/s. This homogeneity is interrupted by significant and often short-lived warming/cooling and freshening events. A subset of floats programmed to profile rapidly (~2.5 hours) shows a strong diurnal signal in temperature with salinity exhibiting somewhat weaker diurnal variations. The diurnal cycle magnitude is largest in areas with light winds and heavy precipitation and was found to decay rapidly with depth (50% over the top 2 m). Observations of near-surface salinity drop events associated with rainfall show an average freshening of -0.37 PSU associated with a cooling of 0.13 °C. The fresh lenses are typically 3 m thick with maximum freshening in the upper ~0.6 m. Conditions favorable for double diffusion are present at the base of the fresh lens. Fresh lenses are typically short-lived with downward mixing occurring within 6-8 hrs (equivalent diffusivity ~10-4 m2/s). A linear correlation between rain rate and salinity drop magnitude was not found across a range of wind speeds due to covariance between wind speed and rain rate. This work provides insight into near-surface vertical processes with the goal of refining the representation of upper ocean dynamics in models and putting observed discrepancies between satellite and in situ measurements in context. The final chapter of this dissertation examines mixed layer properties and subduction rates of high salinity water in an evaporation dominated region of the North Atlantic Ocean (~ 25°N, 38°W) which was heavily surveyed during the NASA-sponsored Salinity Processes in the Upper Ocean Regional Study (SPURS). High spatial resolution, objective maps of temperature, salinity, and mixed layer depth (MLD) (created from Argo, Seaglider, and mooring data) show low spatial variability during the late spring and summer months and higher spatial variability during the late winter and early spring as the mixed layer shoals. This spatial variability is not represented in previous climatologies; meaning prior estimates of annual subduction may be low. These results are put into context with updated, Argo era climatological values of annual subduction for the North Atlantic. Higher temporal and spatial resolution MLD maps combined with mooring velocities and satellite wind stresses were used to investigate both the annual mean and eddy-varying subduction rates in the SPURS region. MLD spatial variability leads to an enhanced lateral induction contribution to annual subduction rates. Eddy subduction rates are locally large when a time-varying MLD is used. Averaged over the SPURS domain however, the net eddy contribution is likely small. This work highlights the importance of capturing MLD variations when examining subduction and water mass formation rates.
- Oceanography