\section{The Flow Mow Experiment} The Flow Mow experiment constituted an innovative effort to avoid both the variability in advected plumes and the expense of acquiring measurements at vents on the sea floor. This dissertation presents its primary result: the most precise measurement of the flux of heat from a hydrothermal vent field accomplished to date. The experimental data sets also offer new perspectives on the plumes and currents that transport the volcanic energy into the surrounding ocean. \subsection{A Well-Mapped Study Site: the Main Endeavour Field} \subsection{Measure Convective Flux in Rising Plumes} \subsection{Use a Stable and Efficient Instrument Platform: the ABE} \subsection{Monitor Ambient Hydrography and Currents} \section(The Setting} \subsection{Topography and geology} How much does valley cross-sectional area vary along axis? How about the actual area surveyed on the N and S MEFCV walls? Did the shoaling topography in the SW corner of the CV cause the surveyed area at the S end to be significantly smaller than at the N end? \subsection{Hydrothermal activity} In addition to Quebec, there are known sources near the MEF permiter. There is a cluster of high $B$ vents near the SW corner of the MEF called Cathedral (D. Kelley, personal communication) and a lone high $B$ source $\sim$500~m NNE of the MEF recently dubbed Raven (H. P. Johnson, pers.\ comm.) and previously noted as Redd Fox (V. Bhat, pers.\ comm.). \subsection{Flow} The Endevour segment is centered (at $48^\circ~N$) latitudinally beneath the divergence of the North Pacific current which shifts seasonally between $\sim45^\circ~N$ (winter) and $\sim50^\circ~N$ (summer) \cite{Pickard+90}. While the surface current turns southward to become the California current and northward to feed the Alaska gyre, the regional flow within a few hundred meters of the segment topography ($\sim1800-2400$~m) is... [cite Roth thesis and Cannon regional papers.] Include Marv's P record, perhaps juxtaposed with P record from mooring. Also, Bill wants a PSD of the P record. And maybe a pressure prediction from the OSU model? Note that some meters were supplemented with $T$ and/or $P$ sensors. Make sure you note that both moorings were deployed and recovered by Richard Thomson of IOS/Canada. He also kindly provided calibration and processing according to standard IOS procedures (ref?). \subsubsection{Mean flows} Include table of means from current meters plus other general stats HERE (or in a different section (or Chapter?) on background conditions and setting)? PSD of v and u (or PVD's?) to show that v dominates u, generally in valley? Citations for mean + oscillatory flows in deep sea: keller, anderson, and lavelle (1975);maybe worth overlaying where HT sources are known to be now? Also mentioned french dives claiming 39~cm/s (~5mab?)... this is anecdotal, but similar to Marv's reports of Alvin being pushed around in/near MEF (get details, including sub fixes over time?). Meter C is very similar to wasp -- maybe a clue that topographic constrictions really accentuate the mean flow that is added to tidal oscillations? Meter C recorded a mean speed of 24.2~cm/s over a 50-s averaging period. Long-term mean was 8.2~cm/s. They also report high coherence between meters A and C (offset 150~m vertically and $\sim$3~km within the MAR/FAMOUS axial valley. They cite Garner (1972) regarding similar flow in Gibbs fracture zone (cross-ridge?) In the vicinity of the Rainbow vent field on the Mid-Atlantic ridge, for example, mechanical mixing at sills driven by tidal oscillations appears to maintain a northward mean flow along an axial valley that is open only at its south end \citep{thurnherr+02}. \subsubsection{Power spectra} Just refer to Rick's paper and summarize comparable articles? The displacement that occurs during $1/2$ of a pure oscillation with amplitude $v_o$ and period $\tau$ is \begin{equation} $\Delta x = \int\limits_{0}^{\tau/2} v_o sin{\frac{{2\pi}{\tau}}t} dt = v_o\tau/\pi $. \end{equation} Assuming an extreme value for $v_o=0.19$~m/s (the maximum observed hourly mean speed above the ridge crest) the displacement is 2.6~km for a diurnal ($\tau=12$~hr) oscillation and 3.5~km for an inertial ($\tau=16.1$~hr) oscillation. During the same half-periods, the mean flow will displace fluid 1-1.5~km. \subsubsection{Coherence and phase} These meters show less powerful, but significant spectral density peaks at the diurnal and inertial frequencies (Figure \ref{cohphase}b). The N mooring has less diurnal power and more inertial power than the S mooring (ref Mihaly and/or others here?). There is significant N/S coherence at each of these frequencies, but the diurnal oscillation is $\sim$180{$^\circ$} out of phase, while the inertial oscillation has a +20-120{$^\circ$} phase difference. \begin{figure}[htbp] \begin{center} \includegraphics[width=0.9\textwidth]{figs/power_coh_phase_NS.eps2} \end{center} \caption{(a) Power spectral density of the along-axis component of flow measured at the N and S moorings. (b) Coherence and phase between the 2 moorings. The level of 95\% confidence is indicated by the horizontal line. The three significant peaks (excluding the low frequency shoulder) correspond with the diurnal, inertial, and semidiurnal periods. Positive phase means that S leads N.} \label{cohphase} \end{figure} \subsection{Hydrography} General NE Pacific $\theta-S$ plot and interpretation with respect to water masses. Pretty good overview in Pickard+Emery. Juxtapose station Papa with Levitus? Cite Reid and Si-maximum. \cite{cannon+93}: Good context on NE Pacific $\theta$--$S$ and flow relative to the JdFR. ``Common water'' ($\sigma_{\theta}=27.76$, ) and ``Intermediate water'' ($\sigma_{\theta}=27.28, z=$800-900~m). Ref to Munk, Abyssal recipes, 1966. Also to Roemmich and McCallister (1989) on large-scale flow in the North Pacific. Maybe good to combine his mean vectors with those from Franks in 1 figure to give overview? Now compare all available ''background'' casts. Need a map of the locations: MZ stns, REVEL stns?, FM stns, Hautala stns. Then pressure-binned overlain profile plots for full water column and plume depths The Flow Mow CTD stations located near the axis of the segment show, on average, positive $\theta$ (Figure \ref{{axial_mean_sig-th}) and $S$ (Figure \ref{{axial_mean_sig-S}) anomalies relative to the background trends that are linear with respect to density ($\sigma-\theta$ and $\sigma-\theta$). This deviation from a linear background was first observed by Lupton et al. and can be used to define density-referenced (hereafter \emph{isopycnal}) anomalies in hydrothermal environments. Over the Endeavour axis, typical isopycnal temperature anomalies ($\Delta_{\rho}\theta$) are $+\sim0.03\circ~C$ and typical isopycnal salinity anomalies ($\Delta_{\rho}S$) are $+\sim0.003psu$. Thus, the overall influence of the hydrothermal system on the ocean appears to be a shift from the background to saltier, warmer conditions at all depths. \begin{figure} \begin{center} \includegraphics*[width=.45\textwidth]{figs/setting/sgth_T2.mean.vs.bkgrd.eps2}\hfill \includegraphics*[width=.45\textwidth]{figs/setting/sgth_S2.mean.vs.bkgrd.eps2} %\includegraphics*[width=.75\textwidth]{figs/setting/sgth_T2.mean.vs.bkgrd.eps2} %\includegraphics*[width=.75\textwidth]{figs/setting/sgth_S2.mean.vs.bkgrd.eps2} \caption[Background and axial CTD stations: $\sigma_\theta$ versus $\theta_0$ and $S$] {$\sigma_\theta$ versus mean $\theta_0$ and $S$ for Flow Mow background and axial CTD stations. The solid line is the mean background (both stations and both pairs of sensors), while the open circles represent 10~m depth-binned averages from stations located along the ridge axis. } \label{axial_mean_sig-th+s} %\label{axial_mean_sig-S} \end{center} \end{figure} Individual axial casts reveal intermittent returns to the background trend at variable depths, but almost exclusively above the ridge crests. The lower fluid, enclosed within the axial valley topography, almost always has positive isopycnal anomalies of $\theta$ and $S$, even within 5~m of the sea floor. The only exceptions are found within buoyant plumes. Further from the axis, however, deep casts make it clear that downstream of the volcano the positive isopycnal anomalies have been embedded within the linear background; descending through the plume depths, the positive anomalies decay back to the background trend near the depths of the axial valley floor. This downstream return to background conditions below plume equilibration depths has been used to justify the definition of an anomaly by extrapolating a slope from above the plume through intermediate depths. Plotting these same $\theta$ and $S$ data against each other (rather than density) also reveals positive temperature anomalies (Figure \ref{axial_mean_th-S}), though in this case they are salinity referenced (hereafter \emph{isohaline}). On average, the isohaline temperature anomaly ($\Delta_S\theta) is $\sim0.05\circ~C} roughly $2\times$ the isopycnal anomaly (ref sxn on correction and conversion factors). As with the isopycnal anomalies, individual axial casts show highly variable isohaline anomalies above the ridge crests, and relatively uniform conditions within the confines of the axial valley. Stations as close as a kilometer from the axis can reveal positive isohaline anomalies near the plume equilibration depths bounded above and below by background fluid. \begin{figure} \begin{center} \includegraphics*[width=.75\textwidth]{T2-S2.mean.vs.bkgrd.eps2} \caption[$\theta - S$ data from Flow Mow background and axial CTD stations] {The depth-binned average $\theta$ and $S$ from background (solid line) and axial (open circle) stations, over the depth range common to all stations. } \label{axial_mean_th-S} \end{center} \end{figure} A number of mixing processes could be responsible for the elevated $\theta-S$ conditions at the axis. The relatively rough topography of the ridge may cause enhanced vertical mixing through turbulence or boundary layers, but in the absence of a warm, salty water mass, this process can only redistribute the $\theta-S$ data along the background trend. The elevated values could be due to fluid from the west side of the ridge (opposite the Flow Mow background stations), deeper (saltier) background water warmed geothermally, shallower (warmer) background water to which salt has been added, or an end member that is both warmer and saltier than the background. These alternatives are illustrated schematically in figure \ref{th-s_diag}. Interleaving of distinct water masses across the Mid-Atlantic ridge has been observed near crest depths \cite{Thurnherr00}. The Northeast Pacific, however, has a much more uniform hydrography. Stations from multiple years and both sides of the Endeavour segment suggest that the regional hydrography is remarkably stable (Figure \ref{th-s_regional_bkg_comp.eps2}. [Cite Cannon/Reid here?] A final and illuminating way to visualize the density, temperature, and salinity fields over the Endeavour segment is in space. A depth profile of the mean density observed at background and axial stations (Figure \ref{}) immediately highlights the dynamic interaction of the hydrothermal plumes and the deep ocean. The difference between background and axial data at a particular depth defines a depth-referenced, or level-to-level, anomaly (hereafter denoted $\Delta_z$). On average, the fluid above 2000~m is denser than the surrounding ocean, while the deeper axial water is less dense than the isobathic background. Show individual $\theta-S$ scatter plots juxtaposed with density profiles: stn10 tn10_th-s-sig-zstn10_th-s-sig-z Figure \ref{stn10_th-s-sig-z} for buoyant plumes and stns 28/33 for temporal averages (explain the linear tails!) \begin{figure} \begin{center} \includegraphics*[width=.75\textwidth]{stn10.th.s.sig1.z.ps} \caption[]{ } \label{stn10_th-s-sig-z} \end{center} \end{figure} Maybe this is the place to show the $\theta-S$ trajectories from ABE flythroughs, as well? Geographic grouping of the vertical profiles provides further insight into how hydrothermal activity causes the transition from background to axial conditions. Figure \ref{profile_comp_sigt-th-s} emphasizes how the isobathic anomalies increase as one progresses from the background stations to the southern axial stations and northward up into the axial valley. Both near the bottom and at upper levels, the absolute maxima of $\Delta_z\sigma_{theta}$, $\Delta_z\theta$, and $\Delta_zS$ are measured at the SoMEF and NoMEF stations, in the vicinity of the MEF. \begin{figure} \begin{center} \includegraphics*[width=.3\textwidth]{S-N.mean_sgt2_z.eps2}\hfill \includegraphics*[width=.3\textwidth]{S-N.meanT2_z.eps2}\hfill \includegraphics*[width=.3\textwidth]{S-N.meanS2_z.eps2} % Andreas used a macro with: % \epsfig{figure=#2.eps,width=#1\unitwidth}\hfill % \epsfig{figure=#3.eps,width=#1\unitwidth} \caption[Depth profiles of mean $\sigma_{\theta}$, $\theta$, and $S$: geographic transition] {Depth profiles of mean $\sigma_{\theta}$, $\theta$, and $S$ from Flow Mow stations grouped geographically. Emphasis is on the transitions between background and axial stations. Axial stations are: ``Deep S," the southernmost stations, from the SE flank to the Mothra vicinity; ``SoMEF," stations 4, 6, and 33, $\sim500~m$ S of MEF; ``NoMEF," stations 3, 5, 14, and 28, $\sim500~m$ N of MEF; ``Deep N," stations just N and S of the saddle. } \label{profile_comp_sigt-th-s} \end{center} \end{figure} Given the positive $\Delta_{\rho}\theta$ and $\Delta_{\rho}S$ in figure \ref{axial_mean_sig-th+s} at all densities, a critical question is how the fluid below $\sim1975~m$ in figure \ref{profile_comp_sigt-th-s} has negative $\Delta_z\sigma_{\theta}$. The same fluid has positive $\Delta_z\theta$ and negative $\Delta_zS$. Taken all together, these observations indicate that average isopycnal, isohaline, and isotherm surfaces all dip over the axis. Isopycnal surfaces are level near 1975~m and bowed upward between 1975~m and $\sim1800~m$. Isohaline surfaces mimic the isopycnal ones, although they are level deeper, near 2075~m, and must slope more steeply than the isopycnals to generate positive $\Delta_{\rho}S$ at all depths. Isotherms dip downward at all depths (also more steeply than do isopycnals) below 1900~m and may bow slightly upwards at shallower depths. There is some indication that at the N end of the segment the isotherms begin to dip only below $\sim1975~m$. bkg-S2.transition_profiles.mean.eps2 bkg-T2.transition_profiles.mean.eps2 depth_bin_dth_hist_stn34.10m.gt2100.eps2 depth_bin_dth_hist_stn24.10m.gt2100.eps2 depth_bin_dth_hist_stn20.10m.gt2100.eps2 ALTERNATIVE PHRASING: Plotted against density, both $\theta$ and $S$ have positive anomalies at all depths. Depth profiles of $\theta$ and $S$, however, suggest that the deep layer is isobathically fresh and warm. This means that isopycnals must be sloped differently at different depths (as predicted by plume equilibration theory) --- plunging into topography below the mean equilibration depth ($z^*$) and bowing upward above. These profiles also imply that isohalines dip downward less than isopycnals below 2100~m, and bend upward more than isopycnals above $\sim$2100~m. [include Figure of background and NoSoMEF P-binned means and raw Stn 10 $\theta$-$S$, and z vs $\sigma_2$? add ref or $S$ and $\theta$ profiles akin to Lavelle technical report?] Axial profiles of $\sigma_2$ are often nearly vertical in the same depth range, evidencing a well mixed boundary layer that is $\sim$50~m thick. The fluid below $\sim$2150m is much clearer ($1/5$ backscatter intensity) than in overlying equilibrated plumes. Taken together, these observations suggest that the increase in $\Delta\theta$ within the valley is caused by an addition of heat to fluid as it flows into the axial valley. This could occur through sea floor geothermal heating of relatively fresh water, injection of warm, fresh vent fluid, or downward mixing of overlying plume fluid. Subsequent mechanical mixing may help to homogenize the added heat, resulting in a reduction of $\Delta\theta$ variance with increasing time and/or distance from a source. END ALTERNATIVE PHRASING The implication for vertical calculation is that south is consistently cooler than north. Can we add evidence that the near-bottom layer (especially below 2150m) always is clear and has a T-profile toe at 1.70$^\circ{C}$ , whereas the toe on north side (and E side?) has toe at 1.72? What is the toe, anyways? \subsection{Other measurements} analyze cm (T,U) time series The near-bottom current meters deployed in 1995 and 2000 were equipped with high-precision thermisters. These sensors provide long and continuous records of water temperature variability within the axial valley near the MEF. Table summarizing expanded temperature data from current meter moorings A) 1000 m N of MEF, WASP mooring Jun?-Sep? 2000 B) 200 m NE of MEF, RCM5 mooring Jun?-July? 1995 C) 1200 m S of MEF, RCM5 mooring Jun-Sep? 2000 Elevation Depth (m) Mean(Te) max(Te) min(Te) std(Te) A) 15 mab 2168 B) 25 mab 2175 C) 50 mab 2168 The observed variance in expanded temperature could be due to 1) Vertical displacement of isotherms by non-hydrothermal forcing 2) Advection of a temperature field that is heterogeneous due to non-uniform (in X or t) input of hydrothermal heat 3) Advection of a temperature field that is homogeneous at the depth of the thermisters, perhaps due to mechanical, near-bottom mixing