Compressive sensing (CS) takes advantage of the spatial and spectral redundancy in hyperspectral imagery to take fewer measurements than traditional sensors. We simulate compressively sensed hyperspectral airborne images of a HyMap image of Cooke City, Montana using the Coded Aperture Snapshot Spectral Imager Dual Disperser (CASSI-DD) sensor model. Flake et al's novel algorithm (2013), which incorporates both spatial total variation (TV) and spectral smoothing, is used to reconstruct the hyperspectral image cube from the CS measurement.[1] We evaluate the effect of the number of physical measurements (nt) on atmospheric compensation, anomaly detection and target detection. The results indicate that the utility of CS is application-dependent and the optimal nt value is application-driven.