Rotation and penetrative convection in water cooled from below}

It has already been suggested that the outer core of the Earth might not be entirely in a convective regime [1], for instance in a region close to the CMB where the heat conducted along the adiabat could be larger than the net heat transfer and where compositional buoyancy vanishes. Radial dependence of thermal conductivity could also produce neighbouring stable and unstable regions. To what extent does convection spread into the stable region is then an important issue. Liquid water and its maximum density at 4°C provides a convenient way to observe penetrative convection. We have performed experiments where a volume of water (of square cross-section 16 cm by 16 cm and height 20 cm) initially at room temperature (around 22°C) is then suddenly cooled around 0°C at the bottom while the sides and top walls are (imperfectly) thermally insulated. Near the bottom, where temperature is below 4°C, unstable buoyancy conditions prevail and we observe convective cells. Above 4°C, quiescent stable stratification is observed. Temperature is close to 4°C within the bulk of the convective zone, so that the flux of heat extracted from the bottom is used to change the temperature from 22°C to 4°C within the successive layers of water, while the height of the convective zone increases and eventually reaches the top of the volume. Temperatures have been measured on a array of 15 PT100 probes protruding 5 mm into water along a vertical side wall. the extreme probes are 1 cm away from the lower and upper boundaries respectively. The evolution of all 15 temperatures can be seen on left-hand side of Fig. 1. Another type of information was obtained from a synthetic schlieren technique, whereby pictures of a random pattern placed behind the transparent water-filled box (perspex walls) are taken every 5 s with a digital camera. The cross-correlation of the different images provides a dynamical visualization of the gradients of refraction index within the volume of water. A snapshot is displayed on the right-hand side of Fig. 1. In a second step, we will analyze the influence of rotation on the configuration described above. The experiments will be performed in October and the results will be available for discussion during the AGU fall meeting. When transposed to the conditions of the Earth's outer core, we expect to be able to assess how stable and unstable regions can exist next to each other. [1] Braginsky, S.I., Phys. Earth Planet. Int., vol. 111, 21--34, 1999; Temperatures on the left, time derivative of index horizontal gradient on the right