Direct Heat Flux Estimates Using a Towed Vehicle

View More View Less
  • 1 School of Earth and Ocean Sciences, University of Victoria, Victoria, British Columbia, Canada
© Get Permissions
Full access

Abstract

Direct estimates of vertical heat flux were computed using data collected with a towed vehicle that carried collocated velocity and temperature sensors. Horizontal wavenumbers from about 1 to 40 cpm were resolved, which excludes some potentially significant contribution from lower wavenumbers. The correlation coefficient of w and θ is 0.1 to 0.2 and statistically significant. The authors compared the cospectra obtained from three regions: (i) a well-defined turbulent interface at 400 m depth being overturned, (ii) sporadic layers of turbulence in the thermocline near a salinity maxima, and (iii) a thermal inversion at the base of a surface mixing layer. For all three regions the direct flux estimates are consistent with the notion of a flux Richardson number smaller than ∼0.2 and Osborn's method. The smallest wavenumber resolved was 0.5ko, where ko is the Ozmidov wavenumber, the largest wavenumber contributing to the flux was 6ko, and there was no significant counter-gradient flux in this range. The direct estimates and those of the method of Osborn and Cox agree within a factor of 2 when the smallest wavenumber resolved approaches ko, while for poorer spectral resolution the former is smaller than the latter. For the thermal inversion the directly estimated flux was upward.

Abstract

Direct estimates of vertical heat flux were computed using data collected with a towed vehicle that carried collocated velocity and temperature sensors. Horizontal wavenumbers from about 1 to 40 cpm were resolved, which excludes some potentially significant contribution from lower wavenumbers. The correlation coefficient of w and θ is 0.1 to 0.2 and statistically significant. The authors compared the cospectra obtained from three regions: (i) a well-defined turbulent interface at 400 m depth being overturned, (ii) sporadic layers of turbulence in the thermocline near a salinity maxima, and (iii) a thermal inversion at the base of a surface mixing layer. For all three regions the direct flux estimates are consistent with the notion of a flux Richardson number smaller than ∼0.2 and Osborn's method. The smallest wavenumber resolved was 0.5ko, where ko is the Ozmidov wavenumber, the largest wavenumber contributing to the flux was 6ko, and there was no significant counter-gradient flux in this range. The direct estimates and those of the method of Osborn and Cox agree within a factor of 2 when the smallest wavenumber resolved approaches ko, while for poorer spectral resolution the former is smaller than the latter. For the thermal inversion the directly estimated flux was upward.

Save