Geosynchronous synthetic aperture radar performance estimation and mission design

Geosynchronous Synthetic Aperture Radar (GeoSAR) missions have attracted growing interest in the scientific community thanks to their unique capability in terms of persistent coverage and temporal resolution. Despite this, no mission has flown yet and there are some uncertainties concerning the technical feasibility of such missions (in the case of high orbit inclination concept) and the achievable performance (in the case of a low inclination and low eccentricity concept). For the purpose of developing a general Performance Estimation Methodology, the need for a new clutter model arose, because the existing clutter models are not sufficient for the GeoSAR missions studied in Europe. The Billingsley clutter model and the Borealscat clutter model have been developed using an integration time which is much shorter than the typical GeoSTARe or Hydroterra ones. Moreover, these two clutter models have been developed for forests (which accounts for almost 36% of Europe’s landcover according to Eurostat) and the main landcover class, which is agriculture, is around the 39% of the total. Since wind-driven motion of vegetation is the main source of clutter power over land. A direct physics-based simulator has been developed for short vegetation using a database of wheat plant motion in a variety of wind conditions. The output of this simulator was the moving target focused signal, which has been used to create the wheat clutter model. The development of this wheat clutter model (specific for medium-high incidence angles) required a specific GeoSAR clutter simulator (developed in C++). The simulated focused signals have been fitted using clutter power models available in literature (expo- nential clutter model and power law clutter model); a comparison of the exponential model and the power law clutter power model has been performed. The power law clutter model gave a better fit to the wheat data. A Signal to Clutter Ratio (SCR) simulator has been developed using the clutter model to assess the azimuth spread of clutter power and hence to estimate SCR in representative landscapes. With the help of a Bistatic Simulator developed during this thesis, a new Single Input Multiple Output mission concept has been proposed (a single monostatic SAR and two passive bistatic receivers on other two platforms). From the study of the wheat clutter dependence on windspeed a new land scatterometer mission concept (peculiar of low azimuth speed Geosynchronous SAR missions) is suggested (to measure the windspeed at a much higher spatial and temporal resolution than current systems). The road to an End-to-End performance simulator is still long but some interesting steps have been done and the capabilities of the Performance Estimation Methodology have been shown. Further work is needed to develop similar clutter models for tall vegetation and the ocean (existing results are available for both of these), and then to combine these with realistic models of landcover and weather (including seasonal changes) to enable comprehensive end-to-end GeoSAR performance simulation. The source code of the simulators are available in the appendices together with the requirements discussion.