Abstract: Unsteady pressure fluctuations in the near field of a wall constitute the primary excitation source of flow-induced vibration and noise in various military and civilian equipment. These fluctuations can be classified into hydrodynamic load and acoustic load, which, despite being highly coupled, exhibit distinct functional mechanisms and propagation characteristics. This coupling presents significant challenges for accurate identification and quantitative analysis, necessitating efficient experimental separation methods. Based on the physical characteristic that both pressure waves and acoustic waves belong to the category of generalized mechanical waves but differ in energy spectrum distribution in the wavenumber domain, this study proposes an experimental hydrodynamic-acoustic separation scheme integrating the self-developed linear surface arrays and the affiliated data processing algorithms for wavenumber-frequency spectrum and wavenumber domain integration. Experiments were conducted in the FL-11 1.8m × 1.4m low-speed wind tunnel, using a flat plate as the experimental model. The experiments were conducted with two types of linear surface arrays that featured the same element channel number but different element spacings. An external additional sound source was arranged for the experiments, and the pressure fluctuation signals under different working conditions were collected to obtain the 1D streamwise wavenumber-frequency spectrum using the self-developed direct measurement. Combined with the wavenumber-domain integration algorithm, the accurate separation of hydrodynamic-acoustic loads was achieved. Finally, the auto-spectral results of the two loads were obtained, and a comparative analysis of the separation results from the linear surface arrays with different spacings was completed. The research results indicate that the proposed method can effectively realize the accurate separation of the auto-spectra of hydrodynamic and acoustic loads, and has good effectiveness for the characteristics of the tonal frequencies, as well as the broadband spectrum and amplitude. Since the linear surface array is limited by wavenumber resolution, there is limitation in the low frequency range regarding the separation. Meanwhile, affected by the energy leakage induced by the aliasing effect, the separation accuracy of the two loads in the high-frequency band decreases. Under the condition of the same element channel number, the long-spacing surface array exhibits superior performance in the low-frequency band, whereas the short-spacing array is more suitable for separation tasks in the high-frequency band. The research findings of this paper can provide key experimental technical support for the accurate traceability and quantitative analysis of flow-induced vibration and noise related to pressure fluctuations.