We utilize twelve-year ambient seismic noise (ASN) recordings to measure near-surface seismic velocity variations (dv/v) within the YNP flooding watersheds. We have observed that the annual peaks of dv/v variations have two to three months lead-time in comparison to anomalous water discharges. Our analysis indicates that the annual cycle of dv/v is highly correlated with the effective precipitation, which takes into account the snowpack loading and melting water infiltration processes. The annual peaks of dv/v are likely caused by abnormal melt water, which exceed the water storage capacity of the land. Furthermore, the best-fit poroelastic model provides further evidence that precipitation factors primarily influence the variations of the seismic field. Our analysis successfully explains the indicative changes in dv/v prior to the 2014 and 2022 YNP floods. This study demonstrates the potential of using seismic observables to monitor and assess the risk of devastating floods in the YNP.