Mitigation of Wind Turbine Echoes in Weather Radar Measurements of the DWD

Zusammenfassung—Wind energy is one of the central pillars of the energy transition in Germany. As wind turbines might dramatically influence atmospheric backscatter of weather radars, the important question arises: is it possible to reduce or to completely circumvent the negative impact of wind turbines to weather radar measurements. Within this paper we propose a two step proceeding:1)Raw data identification step: on raw data level we construct classifiers that are able to discriminate between ‘clean’ and ‘contaminated’ radar echoes. This allows an identification of spatial cells that are affected by wind turbines. These identified cells will be completely removed and recovered in a second step.2)Momemt data recovery step: on the level of higher moment data the removed spatial cells will be recovered by formulating an incomplete data context and solve an associated inverse problem that yields a ‘gap-infilling’ routine that recovers the removed spatial moment data on the basis of informations from its temporal and spatial neighborhood.We provide numerical examples and evaluations on the recovery quality vs. gap size.