The overall goal consists in understanding and overcoming the influence of weather radar returns that stem from wind energy turbines. We tackle this problem on two different levels. First, we model and simulate radar raw data. The goal is to understand the echo characteristics from wind turbine returns and to use this knowledge to construct classifiers that are able to discriminate between “clean” and “contaminated” radar echoes. If possible, raw data-based “decontamination”- routines will be applied. The second approach relies on the analysis of higher moment data. In this approach we are faced with incomplete spatial-temporal data, namely with data gaps that are caused by strong wind turbine backscatter. The goal is to develop and to apply so-called “gap-infilling” routines that deliver physically feasible higher moment recoveries for the missing parts of the data. Mathematically the first approach relies on sophisticated Fourier analysis and classification/learning theory. The second approach involves techniques from inverse problems, sparse recovery principles, partial differential equation based infilling and optical flow.