Wavelet scattering is a highly effective feature extraction method, prevalent in many other fields. This paper introduces wavelet scattering to the field of passive acoustic monitoring, employed to test its relevance to the field using a manually verified subset of an open access dataset. Additionally, we introduce an adaptive whitening method to increase detection efficacy. This approach is shown to be most performant with a spectral entropy detector enhanced by a novel thresholding technique. We demonstrate that a simple classifier trained with little data and utilizing wavelet scattering features can greatly improve the performance of the proposed spectral entropy detector. The efficacy of our method is demonstrated on Antarctic blue whale (Balaenoptera musculus intermedia) calls.