Phased microphone arrays are valuable tools for aeroacoustic measurements that can measure the directivity of multiple acoustic sources. However, when deployed in closed test-section wind tunnels, the acoustics suffer due to intense pressure fluctuations contained in the wall-bound turbulent boundary layer. Furthermore, phased microphone arrays require many sensors distributed over a large aperture to ensure good spatial resolution over a wide frequency range. Microphone arrays of such large count are not always feasible due to constraints in space and cost. This thesis describes an alternative approach for measuring single broadband acoustic sources that uses an acoustic metasurface. The metasurface is comprised of a meandering channel of quarter-wave cavities and an array of equally spaced half-wave open throughcavities. A series of tests were conducted in Virginia Tech's Anechoic Wall-Jet Tunnel where combinations of a wall-bound turbulent jet-flow and a single broadband acoustic source were used to excite the metasurface and produce acoustic surface waves. Measurements of the acoustic surface waves were performed using two methods: a pair of traversing microphones scanning the pressure field along the length of the metasurface 0.25 mm beneath its bottom face, and an array of unequally spaced microphones embedded inside the metasurface. Spectral analysis on the measurements revealed that the inclusion of multiple through-cavities leads to constructive reinforcement of select acoustic surface waves as a function of the acoustic source location. In the case of the embedded microphones, acoustic beamforming was applied in order to extract spatial information. This reinforcement was observed during measurements made with both flow and acoustic excitation, up to Wall-Jet Tunnel nozzle exit speeds of 40 𝑚/𝑠 beyond which it was no longer seen. A series of quiescent measurements made with a range of speaker locations constituted a calibration for the metasurface which was used to locate an unknown broadband acoustic source within an The Root-Mean-Square (RMS) error of 1.06 °.