In this report we present an acoustic detection study of real-world sonar data, derived from a closely controlled sea trial in the Baltic Sea. Specifically, we discuss the estimation of the parameters of a nonlinear delay differential equation from this data. Using the estimated model parameters as detection features we design signal detectors by implementing a Mahalanobis distance-based decision criteria to perform rigorous hypothesis testing and characterize detection performance. We compare the performance of the dynamical detector with a frequency band matched energy detector and demonstrate that the former provides increased detection performance. Since the dynamical detector is insensitive to signal amplitude, we also design a hybrid detector which combine the dynamical features with an energy feature, and estimate the performance of this detector. We generally find a performance gain for the dynamical detector of 2-6 dB over the energy detector. In the case of the hybrid detector we find similar gains as for the dynamical detector. In no cases does the hybrid detector perform worse than the energy detector or the dynamical detector. The hybrid detector does provide better performance in one case. Hence, in a practical system, the hybrid detector would provide the best performance for all possible signal classes encountered.