LAB SIMUALTION OF PLUME SEPARATION AT A VISCOSITY BOUNDARY

Many of the earth's flood basalts have been identified as having two distinct pulses of eruptions that are chemically and isotopically similar, suggesting they came from the same source. These "double flood basalts" have been identified in association with the Ontong Java Plateau, Kerguelen Plateau, Parana Basalts, and Karo Flood Basalts. I have constructed and studied a model, as described in Bercovici and Mahoney 1994, showing plume separation at a viscosity boundary. The model uses Karo corn syrup and water mixtures to simulate a plume passing a viscosity boundary similar to the 660-kilometer discontinuity. This model shows that when a plume hits the boundary it can separate and make smaller trailing plumes. This model is relatively easy to create making it useful for demonstrations as well as further rsearch exploring plume behavior. From this model I have drawn many conclusions and hypotheses that could be used for further study. First for the separation of the plume to occur, there is a minimum size that must be released into the model. This suggests that only the larger plumes should have the double flood basalts associated with them. The size of the plume also determines the number of separations that will occur. Some of the larger flood basalts should have more than just two eruptive events. The larger the size of the plume the faster it travels. This suggests that the time between the pulses reaching the surface will get larger with each new separation. This model is also very sensitive to adjustments in viscosity. There is a threshold viscosity difference necessary for separation and increased iscosity difference can cause a greater number of separations.