Seismic diagnostics of stellar convection treatment from oscillation amplitudes of P-modes

The excitation rate P of solar p-modes is computed with a model of stochastic excitation which involves constraints on the averaged properties of the solar turbulence. These constraints are obtained from a 3D simulation. Resulting values for P are found similar to 4.5 times larger than when the calculation assumes properties of turbulent convection which are derived from an 1D solar model based on Gough (1977)' s formulation of the mixing-length theory (GMLT). This difference is mainly due to the assumed values for the mean anisotropy of the velocity field in each case. Calculations based on 3D constraints bring the P maximum closer to the observational one. We also compute P for several models of intermediate mass stars (1 less than or similar to M less than or similar to 2 M-.). Differences in the values of P-max between models computed with the classical mixing-length theory and GMLT models are found large enough for main sequence stars to suggest that measurements of P in this mass range will be able to discriminate between different models of turbulent convection.