This method can be used to explore the key issues in the dynamic processes of diffusive convection staircase structures, such as their generation, development, and disappearance. With the present experimental setup, we can clearly see the evolution of diffusive convection, which is very difficult, if it is not impossible, to observe directly in the ocean. Gather the components of the working tank.

These include copper top and bottom plates and an acrylic sidewall assembly. The top plate is electroplated and encloses a water chamber accessible by tubes. It also has places for thermistors.

The bottom plate is also electroplated and has slots for thermistors. It contains a heating pad. Use distilled water to carefully clean the copper plates and the acrylic sidewalls.

When done, assemble the tank using screws to ensure it is water-tight. This completed tank is 257 millimeters in length and height and 65 millimeters wide. At an optical table, set up a stainless-steel supporting frame.

On top of the frame, place an insulating slab. With everything in place, place the assembled tank on the slab. Insert thermistors into the top and bottom plates, and connect them to the data acquisition system.

A thermistor monitors the temperature of the plate into which it is placed. Next, move a precision vertical translation stage into place for mounting a probe. Through the top plate, place a micro scale conductivity and temperature instrument sensor into the tank.

Fix the instrument to the precision translation stage. Now, set up software for sensor data acquisition and thermistor readings. Return to the tank to adjust the position of the conductivity and temperature sensor.

Set the initial position of the sensor at its lowest point, here, 20 millimeters above the bottom of the tank. Then, set the parameters of the motion for the translation stage for the experiment. Use the shadowgraph technique to monitor the experiment.

For this, attach a piece of tracing paper on the outside of one side of the tank. On the opposite side of the tank, about five meters away, place a light source. To produce a nearly collimated beam, use a narrow beam LED as a light source.

Place a high-speed camcorder about a meter in front of the tracing paper in the beam path. With the lamp and camcorder on, adjust their positions for a clear image. Obtain two identical rectangular tanks for the working fluid.

The tanks are the same size as the working tank. Join them at their base with a clamped flexible tube, 10 centimeters in length. Place the two tanks at the same height.

Arrange for one tank to have an electric stirrer in it. Next, clamp a flexible 50-centimeter tube into a peristaltic pump. Use the tube to join the tank with the stirrer to the working tank.

Fill the tank that has no stirrer with saline solution of 60 grams per kilogram concentration. Fill the other tank with an equal volume of degassed fresh water. Once the tanks are filled, unclamp the connecting tube.

Continuously homogenize the mixing water with the electric stirrer. Control the flow rate in the working tank with the peristaltic pump. The working tank takes approximately three hours to fill and be ready for the experiment.

For an experiment, use a refrigerated circulator to set the top plate boundary conditions. At the top plate, connect the water chamber to the refrigerated circulator using eight soft plastic tubes. At the circulator, set the temperature for the top plate to the room temperature.

Now, connect the bottom plate's heating pad to a DC power supply, and set its power level. Turn on the camcorder to record the flow pattern. Begin monitoring temperature and salinity data.

Then, start vertical motion of the sensor. Finally, turn on the refrigerated circulator and the DC power supply to achieve the working fluid boundary conditions. This is an example of a shadowgraph image taken when the top plate is at the room temperature and the bottom plate is being heated.

There are three convecting layers in the image, where the fluid density is homogeneous. There are also three interface layers, where large density gradients exist. This intensity fluctuation profile has three peaks that correspond to the interfaces.

Here is the evolution of the intensity fluctuation profile in time, which reveals layer generation, development, and disappearance associated with the diffusive convection staircase. The temperature profile and the salinity profile provide other views of the evolution in time of the system. In these profiles, each continuous line represents data collected over 404 seconds.

The data sets in the temperature profile are each offset by 1.5 degrees Celsius from the one before. For the salinity profile, each data set is offset by three grams per kilogram from the previous set. This method can also be applied to simulate other phenomenon in oceans, such as global oceanic circulation, mixed layer deepening, and hydrothermal plume eruption.

If you want to try this procedure, please do remember to initially adjust the conductivity and temperature sensors to the lowest position of the translation stage to prevent the sensor for crashing on the bottom plate. With the presented results, we can further develop new parameterization of layer thickness, heat flux, and add diffusivity in diffusive convection, which would pave the way for the community of physical oceanography to study the phenomena in the oceanic application.