Signal spectral modelling for airborne radar in the presence of windshear phenomena

A number of significant airplane accidents have resulted from windshear encounters during takeoff or landing. In these situations the radar signal received may be affected by a strong ground clutter that may make the conventional windshear detection algorithms implemented on ground-based systems unusable. Typically, the solution to this problem is to employ a clutter rejection filter and then process the filter output to derive weather information. In this study, a parametric bimodal spectral model of the raw radar signal from an airborne Doppler weather radar is proposed. The bimodal shape of the model has been defined as a superposition of clutter and windshear. The model can be used to define a windshear detection algorithm that can also directly estimate the physical weather parameters by using the estimated model parameters without using a clutter rejection pre-filter. Owing to the impossibility of testing a windshear detection system in a realistic environment, a simulated windshear data base has been developed by National Aeronautics and Space Administration (NASA) and Federal Aviation Administration (FAA) during their windshear research programme. The purpose of this study is to define a parametric bimodal spectral model and demonstrate its validity on the NASA-FAA windshear certification data set.