As the progress report of 4 February indicates, we decided to inject the tracer rather than complete the CTD line, which we will get back to soon. The weather looked as though it was going to deteriorate a few days after our return to the site and stay bad for the better part of a week. The injection system is much more difficult to recover than the CTD system – much easier to damage something – and so we decided to use the short window of good weather that remained for this operation. The injection went very well, but not perfectly. We did it in two streaks (see the picture in the previous report). The first streak went smoothly. It was about 9 miles long, and we injected 42 kg of tracer. The tracer is only very slightly soluble in water so in order to dissolve it we force it at high pressure through nozzles that are just 25 microns (0.001”) in diameter. The total flow rate must be kept low so as not to saturate the water, and so it took about 9 hours to do the first streak.

The second streak did not go so smoothly at all. First, after the difficult job of deploying the injection system, we lost power to the pumps before the injection sled was more than 200 meters deep. After lots of head scratching and testing and swapping components, we found that the brand new, freshly charged, batteries were not doing their job under load. We have not had time to diagnose what is wrong with the batteries or battery cases. We had to wait several hours to recharge the batteries we used on the first streak – pushing us further towards bad weather. Then, when we deployed the sled for the second streak, the flow rate was slower than we expected, and it took 14 hours rather than 8 hours to inject the remaining 36 kg of tracer. This pushed the recovery into a rough sea state. However, under the leadership of Brian Guest on deck, we got the system back safely and empty of its tracer.

Recovery takes a coordinated effort of several people. Brian runs the show. Two others in the science party, now well trained, snag metal loops welded on the corners of the sled with hooks loosely attached to poles and tied to lines leading back to small air-driven winches (tuggers). The poles are pulled free once the hooks are on and the lines are pulled snug with the tuggers. It takes practice to hook onto the bales with the ship and sea moving and the sled dancing around. Once the lines are attached, the crane operator must lift the sled as the winch operator lifts the wire behind which the sled streams. There is another person operating the boom to bring the wire closer to the ship, and the crane swings the sled toward the ship – but not too fast because the lines attached to the tuggers must stay taut to keep the sled from swinging wildly. Science party and ship’s crew have been doing this well - really well and swiftly on the last recovery of the injection sled.

Since the end of the injection, about 40 hours ago, we have been doing a series of CTD casts in a box pattern, 24 miles on a side, around the injection area in the hopes that they will tell us the prevailing velocity of the water at the level of the tracer. The idea is to get the pressure field on a gravitationally “level surface” from the density field and the sea surface slope given by the satellite altimeter. Again, the flow will be with the high pressure to the left and can be calculated from the pressure gradient. The problem is that we are pushing the technique to a smaller scale than might work, because internal waves and tides in the ocean that are not in balance with the pressure field (and therefore are accelerating and oscillating on time scales of hours) can obscure the prevailing, balanced, flow that will move the tracer on time scales of days. We need to know how the tracer is moving if we are to have any hope of sampling it a few days from now. We do have some other tools for this, but none that really measure the current at the depth of the tracer over the full period of interest.

Since the end of the injection, about 40 hours ago, we have been doing a series of CTD casts in a box pattern, 24 miles on a side, around the injection area in the hopes that they will tell us the prevailing velocity of the water at the level of the tracer. The idea is to get the pressure field on a gravitationally “level surface” from the density field and the sea surface slope given by the satellite altimeter. Again, the flow will be with the high pressure to the left and can be calculated from the pressure gradient. The problem is that we are pushing the technique to a smaller scale than might work, because internal waves and tides in the ocean that are not in balance with the pressure field (and therefore are accelerating and oscillating on time scales of hours) can obscure the prevailing, balanced, flow that will move the tracer on time scales of days. We need to know how the tracer is moving if we are to have any hope of sampling it a few days from now. We do have some other tools for this, but none that really measure the current at the depth of the tracer over the full period of interest.