This article discusses inversions for bottom geoacoustic properties using broadband acoustic signals obtained from explosive sources. The experimental data used for the inversions are SUS charge explosions acquired on a vertical hydrophone array during the Shelf Break Primer Experiment conducted south of New England in the Middle Atlantic Bight in August 1996. The SUS signals were analyzed for their time-frequency behavior using wavelets. The group speed dispersion curves were obtained from the wavelet scalogram of the SUS signals. A genetic algorithm (GA) was used for the inversion of sound speeds in the water column and compressional wave speeds in the sediment layers. The variations in the sound speeds in the water column were represented using empirical orthogonal functions (EOFs). A range-independent normal mode routine was used to construct the replica fields corresponding to the parameters. Comparison of group speeds for modes 1 to 9 and for a range of frequencies 8 to 200 Hz was used to arrive at the best parameter fit. An efficient hybrid optimization scheme using the GA and a Levenberg-Marquardt algorithm is presented. Linear perturbation methods were also used to "fine tune" the inversions and to obtain resolution and variance estimates. Analysis was also done to compute the degree of convergence of each of the parameters by explicitly calculating the Hessian matrices numerically. A posteriori estimation of mean and covariance was also done to obtain error estimates. Group speeds for the inverted sound speed fields provide an excellent match to the experimental data. The inverted sediment compressional speed profile compares well with in situ measurements.