Results

At mooring DS, $34\%$ of the profiles contain at least one overturn, significantly more than what we find at mooring DN ($12\%$ ), and what Levine and Boyd (2005) found near the ridge top at mooring Big Boy ( $\sim 10\%$ ). The distribution of dissipation estimates follows an approximate lognormal distribution at both moorings, with more energetic overturns at DS than at DN (Figure 5.2). At DS, large overturns, corresponding to temperature differences of 30 millidegree or more, account for $15\%$ of the total dissipation, but are less frequent than the smaller overturns (Figure 5.3). These large overturns exceed $120 m$ in height (Figure 5.3b), and thus span several instruments. Inside the overturn, the vertical temperature profile displays a typical S shape (Figure 5.4). Including the detection of small overturns ( $\le 4.10^{-3^ \circ}C$ ) in our dissipation estimate cannot account for more than $10\%$ of the final estimate of dissipation (Figure 5.3c).

At mooring DN, fewer large overturns are detected , and the distribution of dissipation shows a higher percentage of dissipation contributed by smaller overturns compared to DS (Figure 5.5). Overturns at DN are in general less pronounced (Figure 5.6). The overall time-averaged dissipation at mooring DS is $1.2\times10^{-8} Wkg^{-1}$ and $1.8\times10^{-9} Wkg^{-1}$ at DN

In sections 5.3-5.6, we discuss the sensitivity of our analysis and results to the different potential sources of errors. We will show that uncertainties in our analysis are greater at DN than DS, but still cannot account for the factor of 10 difference in dissipation between the two moorings (Table 5.1). A more detailed analysis of individual mixing events in the context of the observed flow will follow in chapter 6.

Table 5.1: Summary of the overturns analysis at both moorings

Mooring	Profiles with overturns	Average Dissipation	Diffusivity
DS	$34\%$	$1.2\times10^{-8} Wkg^{-1}$	$2\times10^{-3}m^2s^{-1}$
DN	$12\%$	$1.8\times10^{-9} Wkg^{-1}$	$2.5\times 10^{-4} m^2s^{-1}$