A statistical model of ionospheric signals in low-frequency SAR data

This paper focuses on deriving a realistic statistical model for ionospheric effects in low-frequency Synthetic Aperture Radar (SAR) data. The approach used to develop this statistical model is based on the assumption that, for a certain range of scales, ionospheric plasma turbulence can be considered a scale-invariant process that can be described by power-law functions or fractal statistics. Based on the parameters of a power law model, covariance functions and ionospheric variance-covariance matrices are derived. An ionospheric phase statistics simulator (IP STATS) is presented that is capable of calculating a variety of statistical descriptors and is able to predict representative ionospheric phase screens. To demonstrate its functionality and performance, the IP-STATS system is used to derive statistical models for a 10 year time series of T-band SAR data over the area of Alaska. The developed theory has potential applications in statistical modeling of SAR data, sensitivity analysis of spaceborne SAR systems, and signal simulation. IP-STATS may also be a useful tool in mission design especially in selecting favorable system center frequencies and orbit parameters.