Wind energy is vital to sustainable energy goals, yet the aging infrastructure of many wind turbines poses challenges to its growth. Extending the operational life of turbines beyond their design limits requires a reliable structural assessment, often achieved through numerical tools such as aeroelastic modeling. However, these models face uncertainties in inputs, such as load definitions and unknown structural properties, necessitating experimental validation to improve their accuracy. This study focuses on an experimental campaign conducted on an operating onshore wind turbine to refine numerical lifetime predictions. The monitoring system consists of strain gauges and accelerometers in the blades. Rotor instrumentation systems are rare, making the resulting dataset highly unique. The measured strains were used to obtain modal parameters, which are often used as metrics of similarity between numerical models and experimental data, using the covariance-based stochastic subspace identification algorithm (SSI-COV). Three datasets representing different operational conditions are compared with respect to the identification of the first flapwise and edgewise rotor modes and the first drivetrain torsion mode. Although identification at rated RPM is hindered by the presence of rotor harmonics, good results were found with the rotor at standstill and at low RPM.