4. Vertical flows
Our models give not only an inflow of energy but they also produce an inflow of mass at a rate of
For our models M1 to M3 this gives g cm-2s-1. If this mass is accumulated inside the prominence it would grow very rapidly, its mass would be doubled wihtin s for model M1, within s for model M2 and s for model M3. Since such a rapid steady growth of the prominence as a whole is not observed prominence material has to leave the prominence at a similar rate. (Note: prominence fine structures can form and disappear on slower time scales, but the quiescent prominence as a whole will be rather stationary). Mass losses of the required magnitude could be achieved by a systematic downflow of cool material in the center of the prominence. However this downflow cannot be modelled in our 1D slab configuration. For this reason will shall give here only some order of magnitude estimates for the flow. If we assume that the prominence extends over a height h and that the vertical outflow at the bottom is , whereas there is no inflow at the top then the condition of mass conservation gives
where d is the width of the downflow region, its hydrogen density.
Such systematic downflows can provide additional energy at a rate of , as has been proposed by Heasley & Mihalas (1976) and this could lead to an additional heating of the central parts of prominences. The mean heating rate will be given by
For cm we then get
The heating by enthalpy and ionisation energy inflow from both sides amounts to . These numbers show that for the parameters chosen the gravitational energy release is twice as large as the enthalpy and ionisation energy flow. Therefore such a mechanism could be an important heat source for the central parts of prominences. There are, however, some basic problems with this scenario. Since the magnetic field in prominences is predominantly horizontal this downflow has to occur perpendicular to the field. Even for ionisation degrees as low as 0.2 the flow of neutral atoms across the field lines will be only of the order of cm s-1 (Mercier & Heyvaerts, 1977). Such flows are therefore only possible if very efficient reconnection occurs in the cool part of the prominence. An additional requirement for the reconnection mechanism is that the fields are stretched sufficiently downward to lead to the right magnetic field topology. This reconnection could then result in the required effective resistivity of the prominence plasma. But when the prominence material starts moving downward it also has to convert its kinetic energy into heat. The question how this can be achieved is also open at present. Therefore we think that this mechanism looks promising, but there are still many details which will have to be worked out.
© European Southern Observatory (ESO) 2000
Online publication: June 20, 2000