Modelling the diurnal variability of SST and its vertical extent

Sea Surface Temperature (SST) is a key variable in air-sea interactions, partly controlling the oceanic uptake of CO2 and the heat exchange between the ocean and the atmosphere, amongst others. Satellite SSTs are representative of skin and sub-skin temperature, i.e. in the upper millimetres of the water column where most of the heat is absorbed and where the exchange of heat and momentum with the atmosphere occurs. During day-time and under favourable conditions of low winds and high insolation, diurnal warming of the upper layer poses challenges for validating and calibrating satellite sensors and merging SST time series. When radiometer signals, typically from satellites, are validated with in situ measurements from drifting and moored buoys a general mismatch is found, associated with the different reference depth of each type of measurement. A generally preferred approach to bridge the gap between in situ and remotely obtained measurements, is through modelling of the upper ocean temperature. Models that have been used for this purpose vary from empirical parametrisations mostly based on the wind speed and solar insolation to ocean models that solve the 1 dimensional equations for the transport of heat, momentum and salt. GOTM is a model resolving the basic hydrodynamic and thermodynamic processes related to vertical mixing in the water column, that includes most of the basic methods for calculating the turbulent fluxes. Surface heat and momentum can be either calculated or externally prescribed while the model includes a 2-band parametrisation for the penetration of light in the water column. From the analysis it has been found that the data used to initialise the model, especially the temperature profiles, along with the selected light extinction scheme hold a key role in the agreement of the modelled output with observations. To improve the surface heat budget calculation and distribution of heat in the water column, the GOTM code was modified to include an additional method for the estimation of the total outgoing long-wave radiation and a 9-band parametrisation for the light extinction, correspondingly. New parametrisations for the stability functions, i.e. the quantities in the turbulent diffusivity expressions associated with vertical mixing, have been included. Preliminary results demonstrate the successful implementation of the new parametrisations and the ability of the model to reproduce the diurnal signals seen from the in situ measurements. In addition, special focus is given to testing and validation of different model set-up combinations using experimental data from different campaigns in the Atlantic Ocean, in order to establish a model set-up that can be applied to different regions.