“…In practice, we improve EGF convergence and signal-to-noise ratio using more elaborated processing flows (Bensen et al, 2007;Schimmel et al, 2011b;Moreau et al, 2017;Ventosa et al, 2017;Schimmel et al, 2018) organized in: 1) preprocessing, that may include instrument response correction, anomalous signal rejection, and spectral whitening, usually done with programs like ObsPy (Krischer et al, 2015), SAC (Goldstein and Snoke, 2005), or MSNoise (Lecocq et al, 2014); 2) correlation, a geometrically-normalized, 1-bit (Bensen et al, 2007) or phase (Schimmel, 1999;Schimmel et al, 2011b;) correlation of many short data sequences; and 3) stacking, a sum of correlations that may include weights and denoising (Schimmel and Gallart, 2007;Ventosa et al, 2017). Aside, we include other tools, for instance to robustly measure group velocities of surface waves extracted from noise cross-correlations (Haned et al, 2016;Nuñez et al, 2020), and to characterize elliptically polarized signals within the noise wave field and corresponding sources (Schimmel et al, 2011a;Davy et al, 2015;Carvalho et al, 2019). Ambient noise data are routinely used to extract EGFs for seismic noise monitoring and imaging studies with a wide range of applications, e.g., fault and volcano monitoring (Wegler and Sens-Schönfelder, 2007;Brenguier et al, 2008;D'Hour et al, 2016;Sánchez-Pastor et al, 2019) and imaging structures at different scales (mapping discontinuities, seismic ambient noise tomography, Shapiro and Campillo, 2004;Haned et al, 2016;Romero and Schimmel, 2018;Andrés et al, 2020, among others).…”