Abstract: The time-varying plasma sheath surrounding re-entering vehicles can severely interfere with radar detection, often leading to detection errors. To investigate the impact of the time-varying plasma sheath on radar detection in-depth, we propose a method for modeling intra-pulse radar echoes based on time-varying convolution. The core idea of the proposed approach is to apply time-varying convolution in conjunction with a real-time interpolation strategy to estimate the system response at different stages of temporal variation. The resulting time-varying reflection coefficients are then convolved with the transmitted signals to reconstruct the echo signals. System responses are obtained through electromagnetic simulations of plasma sheaths with varying electron densities. Simulation results demonstrate that, compared to traditional modeling approaches, the proposed method more precisely characterizes the dynamic modulation effects of time-varying sheaths—including amplitude fluctuations, phase shifts, and pulse-compression peak broadening—on radar echoes. This enhances both the accuracy and physical consistency of echo modeling, thereby providing a more reliable tool for radar target recognition and anti-sheath interference research.