Validation of Electromagnetic Wind Radar Simulator Based on LES with Scanning X-band Radar Measurements and Meteorological Data

An electromagnetic radar simulator developed by UCL in the framework of FP7 UFO project (2016) was validated using measurements obtained with scanning X-band radar developed by Thales and meteorological data provided by Meteo France. Obtained preliminary validation of the results was performed with the X-band wind profiler Curie radar (Kovalev et al. 2016). The simulator is developed for radar cross-section, wind and energy dissipation rate (EDR) retrieval, in clear air and in the presence of rain. It is based on the refractive index calculated from Large Eddy Simulations (LES) of the turbulent atmosphere in the boundary layer. The refractive index is then used for the calculation of the radar cross-section of the turbulences as well as the power received by the radar and the Doppler spectrum. The LES model was first developed by De Visscher and Winckelmans (2013) for the study of the effect of turbulence on wake vortices and further extended to take into account the humidity transport and the presence of raindrops and adapted to a turbulent atmosphere. The input parameters of the model are the turbulence intensity represented by EDR and stratification represented by the Brunt-Vaisala frequency (N). Six LES cases for different combinations of EDR and N were simulated (see the attached graph). An additional LES case was simulated to represent the conditions of the atmosphere for the days when the measurement campaign was conducted: EDR = 3.375×10^-3 m² s^-3 and N = 1.8×10^-3 Hz. The size of the LES simulation domain has been set to L=200 m in each direction, in order to have a sufficiently large domain with a reasonable resolution of 0.78 m. 60 s of continuous LES simulation with a time step of 1 s is available for all cases. The radar backscattered signal simulation is based on the method proposed by Muschinski et al. (1999) adapted for X-band. In this work, the simulator was extended for a scanning radar. Backscattering of electromagnetic waves occurs on clear air refractive index inhomogeneities with a scale equal to the Bragg wavelength, (e.g. Doviak and Zrnic, 1993) corresponding to half the radar wavelength, that is approximately 1.6 cm. So, the grid spacing of the LES is much larger than the Bragg wavelength and a full calculation of the scattering integral (Doviak and Zrnic 1993) is impossible. Muschinski uses a parametrized model of the scattering integral based on the hypothesis that the LES sub-filter turbulence obeys Kolmogorov scaling (inertial subrange) at Bragg scale and that the phases of the LES grid cell contributions to the scattering integral are statistically independent. The temporal changes of phase and amplitude of the scattered signal time series are based on the use of an LES simulation snapshot advected by the local velocity with the assumption of linear changes of all the LES parameters between two successive snapshots. A consequence of those hypotheses is that the amplitude and the phase of the coherent complex signal received by the radar are calculated as a function of the local refractive index structure constant of the atmosphere C²_n, and velocity vector field obtained from LES data (Kovalev et al. 2016). The parameters of the radar simulator as well as the mode of operation were chosen so as to correspond to X-band scanning radar used in the measurements campaign of the UFO project. The emitted power of the radar is 600 W; the carrier frequency is 9.502 GHz; the radar range cell size is 62 m and resolution in azimuth and elevation is 1.8°; the pulse repetition frequency is 4 kHz and dwell time is 16 ms which corresponds to the use of 64 consecutive pulses for Doppler spectrum calculation. The radar measurements were processed to extract the spectrum width associated with turbulence. Contributions due to the antenna beam and wind shears were removed. The transverse wind speed necessary for estimation of EDR (White et al., 1999) was estimated from the measurements of radial wind speed and compared with available meteo data. The results of the comparison show a good agreement for available data sets. A comparison between the EDR values retrieved from the simulated backscattered signal and the mean local values of LES within the corresponding radar cell is shown in the attached figure. In this figure, the local values of EDR^1/3 for four different LES cases are plotted against the values of EDR^1/3 retrieved from the radar simulator. The simulated data show an underestimation of EDR^1/3 for higher EDR levels. The reason for this is still under investigation. The SNR influences the EDR estimation: a minimum of about 15 dB SNR should be reached to have an accurate estimate of EDR from the Doppler spectrum. The results for rain will be introduced in the final version of the paper. The radar simulator will then be available for parametric analysis.
References:
Doviak, R. J., and D. S. Zrnic, 1993: Doppler radar and weather observations. 2nd ed., San Diego : Academic Press.
Kovalev, D., D. Vanhoenacker-Janvier, R. Wilson, and F. Barbaresco, 2016: Electromagnetic wind radar simulator validation using meteorological data and a zenith X-band radar. Radar Conf. 2016, London.
Muschinski, A., P. P. Sullivan, D. B. Wuertz, R. J. Hill, S. A. Cohn, D. H. Lenschow, and R. J. Doviak, 1999: First synthesis of windprofiler signals on the basis of large-eddy simulation data. Radio Sci., 34 (6), 1437–1459.
I. De Visscher, L. Bricteux, and G. Winckelmans, “Aircraft vortices in stably stratified and weakly turbulent atmospheres : simulation and modelling,” AIAA Journal, vol. 51, no. 3, pp. 551–566, Jan. 2013.
(2016) The UFO project website. [Online]. Available: http://www.ufo-wind-sensors.eu/home
White, A. B., R. J. Lataitis, and R. S. Lawrence, 1999: Space and Time Filtering of Remotely Sensed velocity Turbulence. J. Atmos. Ocean. Technol., 16 (12).