The propagation of two ultrashort radiation pulses interacting via an intermediate level is analyzed by numerical integration of a system of equations describing coherent interaction between radiation and a three-level medium. It is shown that the propagation of the pulses depends strongly on their areas on entry to the medium and on the relationship between the dipole moments of transitions between neighboring levels. Substantial temporal narrowing was established for fairly high-power pulses (4π and 6π pulses) and this may be utilized to control their durations. Investigations are also made of the influence of relaxation processes, delay of pulses relative to one another on entry to the medium, and nonresonant dispersion during pulse propagation under the specified conditions.
The optical Hanle effect (light-shift-induced zero-field level crossing) has been investigated theoretically and experimentally. In the case of a $J=0\ensuremath{\rightarrow}1\ensuremath{\rightarrow}0$ three-level system and of a $J=0\ensuremath{\rightarrow}1$ two-level system, the resonance line shape has been obtained using a nonlinear-response-function method. Approximate expressions, valid under the conditions for the observation of light shifts, are obtained and compared with exact calculations as well as with the corresponding expression for the magnetic Hanle effect. As concerns the weak laser beam which produces the coherent excitation of atomic substates, we have calculated the linear response in the limiting case of monochromatic and of broad-band excitation. The experimental investigation has been performed for the Ba resonance line using two cw dye lasers at $\ensuremath{\lambda}=5535$ \AA{} and an atomic beam. The absorption- and dispersion-shaped level crossing resonances have been observed, for narrow- and broad-band excitation, by varying the power of the strong nonresonant laser beam. We have checked that the line shapes are only dependent upon the reduced variable $\frac{{P}_{L}}{\ensuremath{\delta}}$ (
PL; beam power; $\ensuremath{\delta}$: frequency detuning). Using the theoretical expressions, we have tried to fit the experimental curves. Provided that the Gaussian distribution of the beams is taken into account, we have obtained a good agreement between theory and experiment for the whole set of experimental data.