Temperature structure and variability in the diurnal thermocline

Salmon, James Robin (1980) Temperature structure and variability in the diurnal thermocline. University of Southampton, Doctoral Thesis.

Abstract

A field experiment in the Mediterranean Sea (Calvi Bay, Corsica) performed during calm and sunny weather on 31 July, 1975 in which 24 horizontal (n'2 km) series of vertical temperature profiles to lm depth were acquired is described in Volume Two. Temperatures were sampled with platinum resistance thermometers at 10cm depth intervals and nb0cm horizontal intervals from a small research catamaran which was towed at n'2ms. Simultaneous measurements of instantaneous thermometer depth, navigational, hydrological and meteorological parameters provided ancillary data. A problem of random drift in the calibration offset of the individual thermometers necessitated the development of a post-experiment, self-calibration technique which demonstrated significant uncertainty in relative temperature measurements. The best portions of the data were isolated for analysis. In Volume One, a one dimensional, numerical, thermodynamical model for the prediction of the diurnal behaviour of near-surface temperature profiles under calm and clear skies is described. It features a detailed, spectral, solar internal heating scheme, surface heat (and mass) exchange, molecular transport of heat and salt, convective adjustment of statically unstable layers and vertical advective heat transport. Turbulent diffusion of heat is modelled using an eddy diffusivity. Comparisons of the results of the numerical model and the field experiment showed that the model could not predict all of the temperature features observed in the data. The (averaged) data and numerical model predictions displayed the best agreement for eddy diffusivities and vertical velocities lying between 500 and 1000 tes molecular(7x10 to 13x10 5 m's-2) and -1.x10-0 to 5.x10-5 ms-' respectively. A method for the simultaneous derivation of an eddy diffusivity anda vertical velocity from a measured temperature profile was developed but not applied because of the large uncertainty in the data. The numerical model was used to investigate several potential sources of the large horizontal variability observed in the data. Spatial and temporal variations in meteorological (except wind-inducing mixing) and hydrological parameters were discounted, as were temperature distortions due to surface waves and the phenomenon of double-diffusion. Processes which seemed capable of producing the observed non-uniformities in the temperature were Langmuir circulations, localized regions of convergence or divergence, bathymetric effects, internal waves, coastal upwelling, thermal convection rolls and horizontal variations in the turbulent mixing rates due the surface wind stress. The model was modified to simulate a wind-mixed near-surface layer, maintained by the energy available from wind forcing, to demonstrate that the latter process could also account for some features of the averaged temperature profile structure.

This record has no associated files available for download.

Contributors

Download statistics