Because sources of volcano infrasound are shallow or aerial, infrasound observations can provide valuable information on eruption dynamics (Fee and Matoza, 2013). In addition to extracting eruption source parameters, higher precision source localization becomes feasible for volcano infrasound, because characteristic wavelengths of volcano infrasound are short (34–340 m for 10–1 Hz sound) compared with associated seismic wavelengths. Such high‐resolution source location makes it possible to differentiate active vents in a multivent system (Ripepe et al. , 2007; Jones et al. , 2008; Johnson et al. , 2013) and to track surface flow phenomena such as pyroclastic flow or rock avalanche (Yamasato, 1997; Ripepe and Marchetti, 2002; Moran et al. , 2008; Ripepe et al. , 2009). High precision accuracy in source location can be achieved by deploying sensors in a local range (e.g., <10 km radius). One benefit of using a local network is to reduce time‐varying propagation effects such as wind and atmospheric temperature variability. Because infrasound sensors are typically deployed on the flanks of volcanoes where topography is pronounced, interaction of infrasound with topography between source and receiver may have a significant effect on propagation physics on a local scale. For example, it was shown that sound diffraction at a volcano crater rim can significantly distort infrasound waveforms (Kim and Lees, 2011). Scattering along the source–receiver topography can cause nonisotropic radiation patterns of infrasound across the recording network (Matoza et al. , 2009; Kim et al. , 2012; Lacanna and Ripepe, 2013; Lacanna et al. , 2014). Even though the importance of near‐source topography on volcano infrasound propagation has been documented, comprehensive modeling of the problem and verification of the result by real field data are slight. In many cases, modeling was performed in a 2D domain that cannot represent realistic volcanic topography, and/or field observations were limited in azimuthal …