We perform the flow behind a squareback Ahmed body using DNS. The dynamics of the flow is analyzed by Proper Orthogonal Decomposition to extract its main characteristics. We focus on the relationship between the base pressure and the near-wake velocity field. The dynamics of the switches due to bimodal behavior of the flow is also considered. The linear stochastic estimation technique [4] is applied to estimate the base pressure modes from the wake flow field and vive-versa. We show that the velocity-based estimator can recover dominant base pressure fluctuations and that the pressure-based estimator can detect velocity deviation and switch mode but fails to capture shedding modes.