Assessment of seismic velocity changes from ambient noise single-station cross-component analysis in Altamira, Brazil

Methods capable of monitoring groundwater systems are essential to support decisions and resource management. In recent years, techniques based on the analysis of ambient seismic noise have consolidated passive seismic methods as valuable monitoring tools. This work focuses on a single-station cross-component analysis based on Ambient Noise Seismic Interferometry and Coda Wave Interferometry to quantify seismic velocity variations ( δν/ν ) between 2014 and 2016 near the city of Altamira and the Belo Monte hydroelectric complex, located in the Xingu River Basin in Brazil. The goal is to determine the extent to which seismic velocity changes reflects hydrological variations derived from piezometric readings and rainfall data. Automatic change-point detection enabled the identification of attention intervals and objective comparison with hydrological data. Velocity drops showed a higher correlation with piezometric levels than with direct rainfall metrics, and the interpretation indicated rainfall-induced pore-pressure changes as the dominant mechanism driving the δν/ν variations. The time series also exhibited a seasonal pattern consistent with regional rainfall cycles, indicating that the observed changes are more likely governed by natural hydrological processes; however, the velocity decrease overlapping reservoir impoundment could not be uniquely attributed to rainfall-driven processes or reservoir-related effects.