Testing the Performance of Radar and Lidar Vertical Wind Shear Detection at Frankfurt and Munich Airports

Abrupt changes of wind velocity can cause serious aircraft hazards. Wind shear poses a great danger during climb-out and approach operations since aircraft air speed and height are near critical values, thus rendering the aircraft susceptible to the adverse effects of wind shear. In order to detect, quantify and alert on the presence of vertical and horizontal low-level wind shear a novel combined system based on X-band Doppler polarimetric radar and 1.6 µm Doppler lidar measurements has been developed and installed at the international airports of Frankfurt and Munich. As a fact of the combination of both sensors the wind field can be observed in rain as well as in clear air conditions.

In general, wind measurements of the atmospheric boundary layer (ABL) profile at aerodromes using radiosondes are not possible. Focus of our investigations will be on inter-comparison of one year lidar and radar high-resolution wind profiles (approximately 30 m vertical resolution) of the lower atmosphere (up to 800 m vertical) at Frankfurt and Munich airports. Prior to the data processing, non-meteorological and ambiguous echoes are removed from the measurements using various echo classification techniques (radar) and modified wind standard deviation and signal-to-noise thresholds (lidar). Thereafter, aimed at vertical wind shear detection wind profiles are retrieved from the 5 minutes Doppler volume scans (radar: 11 PPIs from 1° to 60°, lidar: 5 PPIs from 1.5° to 20°) using the established volume velocity processing (VVP) method. In the last step, retrievals from both sensors are merged into a single product depending on the availability of trusted sensor data from lidar only, radar only or both lidar and radar (weighted average depending on the count of single measurements and standard deviation). Based on the difference of the horizontal wind vector between two 100 ft (about 30 m) layers the vertical wind shear vector is calculated. A wind shear advice is given automatically when the maximum vector difference exceeds 5 kn (about 4.63 m/s) per 100 ft (ICAO).

Our results show that in most cases vertical wind shear events appear in clear nights and mornings as a fact of low-level temperature inversion and low-level jets. For these events high availability of lidar measurements is important. In general, for the crucial heights up to 500 m AGL lidar data is available in about 80 % to 90 % on average. Radar wind availability strongly depends on hydrometeor reflectivity of the emitted radiation (wave length of 3.2 cm). For bad weather situations linked with precipitation and high wind speed (differences/shear) radar measurements are available and used for monitoring of the wind shear thresholds. The fraction of radar retrievals increases significantly with increasing wind speed (from 5 % at 4 m/s, to 30-40 % at 20 m/s). However, the availability of wind data depends on seasonal effects. During fog events at Munich airport sensor measurements have been available neither from lidar nor from radar.

Depending on the weather situation in light precipitation the systems' sensitivity allows wind and wind shear data to be derived simultaneously from both sensors lidar and radar. In most cases during light rain measurements of both sensors pass the quality-control process and then are merged into a single product. At Frankfurt and Munich airports about 5 % to 8 % of the measurements on average vertical wind data retrievals combine both sensors. Inter-comparisons of wind from lidar and radar are the baseline for verification/quality analyses (see following). In general, each sensor radar and lidar shows a different amount of single measurements and standard deviations. Vertical wind of both systems at Frankfurt as well as at Munich is correlated clearly. At all heights the mean difference of radar and lidar wind speed is small (bias close to 0 m/s, root mean square difference RMSD smaller than 0.5 m/s). In general, wind speed difference increase slightly with increasing wind speed to a maximum bias of 0.2 m/s on average. However, this still means a much better quality performance than given by radiosondes (bias of 0.5 m/s to 2.0 m/s on average).

According to ICAO high-frequently and high-resolution wind profile measurements could revolutionize mesoscale forecasting. In order to improve weather forecasts the quality controlled low-level wind data are foreseen to assimilate in high-resolution NWP (numerical weather prediction).