The time-varying behavior of ocean-clutter signals is investigated with the objective of clutter suppression for target detection. It is shown that the ocean clutter may be adequately modeled with two narrowband Bragg signals with time-varying frequencies. Low-order adaptive-prediction filters can thus be used to suppress the ocean clutter. Simple adaptive filters can also be designed to improve the signal-to-clutter ratio for low Doppler target signals. It is postulated that the time variations of the Bragg frequencies are related to the ocean wave-height spectrum. Initial results along these lines suggest that this may lead to an alternate, more robust method of estimating the sea state from HF radar data. >