Here is another aerial image to complement those posted earlier by Falk. This photograph shows some very cool eddies, and it is also a great illustration of how thin the surface layer of dye was to the SW of inlet. The dye-free streak on the right side of the photograph was left in the path of our dye sampling WaveRunner.
Today I tried initial analyses on “vBar” pixel instruments to measure tidal currents in the inlet. These are based on Chris Chickadel’s work and seemed potentially applicable to this situation of co-aligned waves and currents in the inlet. Below shows the locations of estimates.
The figure below shows ebb current estimates for each of the three lines. Locations of zero current are mostly where pixel resolution has degraded. The cross-inlet structure and along-inlet gradients seem interesting. Of course, I have not compared against any ground truth (but hope to).
These are initial experimental results that we will work further on. We collected these data for one transect every half hour since 04/29 and added the two other transects on 05/10. We will begin bulk analysis soon-ish.
If you are interested, here are four plots illustrating the method. The first is just an example stack and shows the mix of waves propagating into the inlet (down left) and current streaks showing ebb flow (down right). This is a moderately easy case with strong current signatures.
The second figure shows the corresponding f-k spectrum (using a convention of +/- frequency and positive wavenumber). The ridge of energy going down right corresponds to the current. The curved shape in the upper half corresponds to the gravity waves and would normally be roughly parabolic but the current turns the dispersion curve. The analysis exploits the current ridge but could also exploit the wave Doppler shift (later???). (This is my coolest plot). Streaks at high k are from aliasing and are not included in the analysis.
The third figure transforms the f-k spectrum to V-k space, a al Chickadel.
The fourth figure integrates through k to give a variance versus V curve. The peak (Gaussian fitted) is the estimated current. This case is the landward-most point in west transect, so V ~ 95 cm/s.
Today (Sunday May 13th) is instrument turnaround day and general catching up. We also have a report on the dye-bomb release we did on May 11th. In short, it was an amazing dye of observing the ocean. The RIVET community had many assets in the water measuring dye, currents, temperature, and salinity. Here we’ll focus on a few observations to give a quick overview of what happenned during the day.
In contrast to the previous releases where we continuously pumped dye for 1.5-2.5 hours, this time we poured 20 gallons of Rhodamine WT at ~0800 into the inlet 300 m down from the junction with the Inter-coastal waterway. Below you can see the dye getting poured in.
As the tide was ebbing rapidly the dye gets stretched and stirred rapidly. It is quickly well mixed (confirmed with CTD+F) casts and makes its way down the inlet. As it approaches the inlet mouth, the dye has spread all the way across the inlet and (in contrast to the point releases) the dye moves out both “old” and “new” channels.
The conditions on May 11 were very benign. Very small waves and extremely light winds during the entire release. The dye then gets rapidly ejected offshore out of the two channels and the SW near-beach channel.
By shortly after 0900 dye is clear of most of the inlet. CTD+F casts offshore of the main or “new” channel reveal a surface concentrated (< 2 m) dye field of about 20 ppb with no dye at all below. This is associated with a strong halocline. This is in part due to the weak mixing that is induced due to the light winds and lack of waves and whitecapping. See the figure below.
The fresher water of the inlet simply slides on top of the saltier (and slighly colder) oceanic water. Thus forming a “lens” within which the dye is contained. The weak winds likely contribute to the maintenance of this strong water-mass gradient. These type of vertical structure conditions will be a very significant challenge for models to reproduce as a) they may not have enough vertical coverage and more importantly b) the model-induced mixing is often far too strong to maintain such a gradient.
The tide then starts to flood at ~1000, and begins to pull water back from “old channel”
Notice the wispy dye to the SW of the inlet. This is all very much surface concentrated. Often boats and jetskis driving in it left a trail of dye-free water behind. This was confirmed by the CTD casts, which show an even thinner surface layer of dye with lower max dye. Dye stratification is again is influenced largely by a strong halocine, but here there is also a stronger thermocline than in the previous set of casts
As the tide started to seriously flood, the dye patch was broken into two parts. 1) an offshore component that spread itself out from the main channel entrance to the SW and 2) a patch that got sucked back up the inlet. This mostly occured from dye in the “old” channel. As the dye approached the “S”-turn of the inlet, it converged strongly on the ever-present scum line, just as the drifters had (see previous post). See the photo below.
This sort of “front” will also be a challenge for models to reproduce as the lateral (horizontal) eddy viscosities that they have may preclude the formation of such convergence zones in these rapidly changing regions.