Monitoring and quantifying wind turbine clutter in DWD weather radar measurements

Abstract. Wind turbine clutter (WTC) is a serious threat to radar measurements from polarimetric weather radar. In Germany, wind turbines can now be built within the 5–15 km range of a weather radar in order to support and further increase the production of green energy. In order to protect the remaining 5 km radius from further wind turbine expansion, WTC is monitored and the consequences on radar data quality are quantified. There are currently no filter methods that can reliably separate wind turbine clutter from desired weather information. It is shown, that a dynamic WTC detection algorithm on the signal processor level performs well in over 80 % of the time if the rotor speed of the wind turbine is larger than 5 rpm (i.e. wind turbine is running). The dynamic WTC detection algorithm fails in situations where, presumably, the static clutter contribution from the wind turbine tower is too large. This is assessed for the first time by using wind turbine operator data from two wind turbines at a distance of about 10 km from a radar system, which are provided in real-time to the Deutscher Wetterdienst (DWD). For the DWD radar system Ummendorf, persistent WTC occupies 5 % of the area in the 5 km radius (approx. 4 km²), and 3 % in the 5–15 km radius (approx. 18 km²). WTC is found up to an elevation of 3.5° within the 5 km radius, and 1.5° with the 5–15 km radius. Using wind measurements from a synoptic weather station near the radar system Ummendorf, a WT detection probability between 60 % and 80 % can be deduced for the WTs at around the 5 km radius. We show that polarimetric radar measurements are more sensitive to WTC. Especially the areal coverage of the disturbance is larger than that observed for traditional radar reflectivity. Also, the vertical extent of WTC on polarimetric moments, illustrated by the depolarisation ratio (DR), is clearly present at 3.5° elevation, and less so in radar reflectivity. A wind park at the 5 km range from the radar system Ummendorf is used to quantify the beam blockage by WTs statistically. Though the data are noisy, a 0.5 dB blockage can be estimated for this particular wind park. This is significant, as the overall accuracy of the radar reflectivity has to be within ±1 dB, meaning 50 % of the radar reflectivity error budget is consumed by this particular wind park. Further, the radar looses sensitivity in measuring precipitation, in particular at long ranges. Static beam-blockage corrections are not applicable to WTC. We conclude, that the 5 km radius must be kept free from WT expansion.