An effective filtering technique is required for rainfall rate measurement by weather radar. A Jensen–Shannon distance (JSD)-based thresholding filter is proposed to mitigate nonmeteorological signals, either in clear air or rain situations. This algorithm classifies range-Doppler bins into two classes, hydrometeors and nonhydrometeors, based on spectral polarimetric variable features. The result is a mask to be applied on the spectrograms. The variable selected here is the spectral co-polar correlation coefficient, available in dual-polarization and full polarimetric radars. The algorithm first does global thresholding by finding an optimized threshold value based on the averaged clear-air spectral polarimetric variable distribution. Next, classical filtering steps are carried out like a ground clutter notch filter around 0 ms−1, a mathematical morphology to fill gaps in the hydrometeor areas, and a removal of narrow Doppler power spectra. The second part of this article is the assessment of filtering techniques without ground truth. An assessment without ground truth is useful to select optimal algorithm configurations from a large solution space. Criteria of good filtering are defined both in the spectral and time domain. Based on those criteria, subjective and objective unsupervised evaluation metrics are derived, with a focus on the objective ones. Data, including clear air and rain collected from a full polarimetric Doppler X-band radar in the urban area, are used. With the proposed unsupervised evaluation metrics, the JSD-based thresholding filter is compared to two spectral polarimetric filters. Overall, the JSD-based filter performs very well considering both the subjective and the objective evaluation metrics.@en

Radar rainfall nowcasting stands for the prediction of rainfall amounts and intensities over the next 6 hours by means of statistical extrapolation of radar measurements. It is the principal ingredient for modern flood forecasting and early warning systems. Radar forecasts are generated by identifying and tracking rainfall cells in radar images and extrapolating them along the main direction of motion based on the assumption of Lagrangian persistence. The latter means that the rainfall cells do not physically evolve over time. Therefore the key to improving predictions is to anticipate the growth and decay of rainfall cells during the next few hours. Studies have shown that estimating growth and decay based on past radar images is nearly impossible. Therefore, in this thesis, a new approach for predicting growth and decay based on physical guidance from a numerical weather prediction model is investigated. The idea is that numerical weather prediction models are better at anticipating changes in the atmosphere than radar and thus, should also contain information about the growth and decay of rain cells. To test this hypothesis, two simple machine learning models have been trained to learn the relationship between the observed growth and decay in radar images and the output parameters from the Dutch mesoscale numerical weather prediction model HARMONIE. The models were trained on data in the summer of 2019 around the city of Rotterdam for two area sizes, $20\times 20$ km$^2$ and $60\times 60$ km$^2$. The first model is static and assumes that all predictions from HARMONIE are correct. Unfortunately, this turns out to be too optimistic as HARMONIE often places the rain cells at the wrong locations with the wrong intensities and trends. As a result, no meaningful relationship between HARMONIE outputs and growth/decay in radar could be learned. The dynamic model overcomes this issue with the help of an additional classifier. The classifier predicts whether the information from HARMONIE can be trusted or not. Then, a regression model predicts the magnitude of growth and decay (only for the trusted cases). Out of 308 analysed cases, the classifier labelled 237 HARMONIE predictions as being untrustworthy, thereby removing a lot of bad cases. However, 98 of these were false negatives, meaning that they could have been used to predict growth and decay. On the other hand, only 71 cases were labelled as containing useful information, from which 27 cases were false positives. Despite these errors, the dynamic model was able to improve the root mean square error on the predicted growth/decay by $27$\% for these 71 cases. Also, the correlation between the predicted and actual growth/decay increased from near zero to $0.488$. This is encouraging and clearly highlights the potential of this approach. Still, some important challenges remain. In particular, results show that the dynamic model often underestimates the magnitude of growth and decay. Also, when the entire validation data set is considered, including all unusable cases, the improvement with respect to the static model is only $6$\%. This is mainly caused by the poor performance of HARMONIE and the large number of unusable cases. However, performance could be improved further by building a better classifier capable of identifying all good cases with a low number of false negatives.

In this article, five velocity-based energy dissipation rate (EDR) retrieval techniques are assessed. The EDR retrieval techniques are applied to Doppler measurements from Transportable Atmospheric Radar (TARA)—a precipitation profiling radar—operating in the vertically fixed-pointing mode. A generalized formula for the Kolmogorov constant is derived, which gives potential for the application of the EDR retrieval techniques to any radar line of sight (LOS). Two case studies are discussed that contain rain events of about 2 and 18 h, respectively. The EDR values retrieved from the radar are compared to in situ EDR values from collocated sonic anemometers. For the two case studies, a correlation coefficient of 0.79 was found for the wind speed variance (WSV) EDR retrieval technique, which uses 3D wind vectors as input and has a total sampling time of 10 min. From this comparison it is concluded that the radar is able to measure EDR with a reasonable accuracy. Almost no correlation was found for the vertical wind velocity variance (VWVV) EDR retrieval technique, as it was not possible to sufficiently separate the turbulence dynamics contribution to the radar Doppler mean velocities from the velocity contribution of falling raindrops. An important cause of the discrepancies between radar and in situ EDR values is thus due to insufficient accurate estimation of vertical air velocities.@en

Scanning radars are promising sensors for atmospheric remote sensing, giving potential to retrieve parameters that characterize the local air dynamics during rain. For the observation of air motion, radars are relying on the backscatter of particles, which can, for example, be raindrops or insects. To measure wind vectors and turbulence intensities remotely during rain the radar is a common choice. This is mainly because the radar signals are not attenuated too much by the rain itself, which is the case for instruments operating at other frequencies, such as lidars. There is, however, a problem with measuring air dynamics from raindrops. Raindrops are not perfect tracers of the air motion. It may thus be necessary to make some corrections when air-dynamics parameters are estimated with a radar during the rain, and account for that raindrops are imperfect tracers of the air motion. This dissertation focuses on this problem. In addition, existing radar-based wind vector and turbulence intensity retrieval techniques are assessed for when they are applied during the rain, and they have been further developed.@en

This article presents the prospects of measurement systems for wind hazards and turbulence at airports, which have been explored in the Ultrafast Wind Sensors (UFO) project. At France’s Toulouse–Blagnac Airport, in situ, profiling, and scanning sensors have been used to collect measurements, from which wind vectors and turbulence intensities are estimated. A scanning 1.5-µm coherent Doppler lidar and a solid state X-band Doppler radar have been developed with improved update rates, spatial resolution, and coverage. In addition, Mode-S data downlinks have been collected for data analysis. Wind vector and turbulence intensity retrieval techniques are applied to demonstrate the capabilities of these measurement systems. An optimal combination of remote measurement systems is defined for all weather monitoring at airports. In this combination, lidar and radar systems are complementary for clear-air and rainy conditions, which are formulated in terms of visibility and rain rate. The added value of the measurement systems for high-resolution numerical weather prediction models is estimated by an observing system experiment, and a positive impact on the local wind forecast is demonstrated.@en

In this presentation CEWE is introduced, which is a novel open source Python module with the aim of processing summary statistics for massive data sets. It is based on the addition of raw moments and a mixed moment from which regular statistics can be calculated. The novelties of the CEWE tool in comparison with other massive data analysis tools are (1) that it is compatible with circular data; (2) there is a minimalistic worked example available that can be used as a starting for a new project and (3) there is an on-line worked example, TARA CEWE. The tool provides a fast way to explore and validate the data and gives the opportunity to discover correlations between variables. In the on-line example TARA CEWE, observables from a precipitation profiling radar are compared to an extensive list of meteorological parameters from the meteorological supersite in Cabauw, the Netherlands.@en

In this presentation a few turbulence intensity retrieval techniques are assessed by estimation of their accuracy and validated by in-situ measurements by using a polarimetric precipitation profiling radar. The turbulence intensity retrieval techniques considered include both (1) a classical approach based on analytical formulas and the assumption of independent contributions to the radar spectral width of the formula parameters (fall speed spectral width, turbulence, etc.) and as well (2) a new approach using model based estimation of radar observables that includes the inertia effect of the scatterers. The major challenge in the technique proposed is to estimate the drop size distribution from the radar measurements such that it can be used in the retrievals. Consequently the turbulence intensity can be estimated. The proposed retrieval techniques are assessed using data from the S-band profiling polarimetric radar TARA operating at 45 degrees elevation angle at the Cabauw (the Netherlands) research site. At this slant elevation agle the radar observables contain both information on fall speeds of the scatterers via the Doppler effect as well as information on axis ratios of the scatterers via polarimetric variables. The estimated turbulence intensities are compared to in-situ values from a sonic anemometer.@en

The implementation of state-of-the-art retrieval techniques of wind vectors and turbulence intensity (EDR) will be presented. They are applied to measurements from the fast scanning X-band radar during the UFO trials at Toulouse airport. It will be demonstrated which retrievals, both for wind vectors and EDR, give the most reliable results when applied in the most challenging cases, that is when there are clouds or when it is raining.@en

A new model is proposed on how to account for the inertia of scatterers in radar-based turbulence intensity retrieval techniques. Rain drop inertial parameters are derived from fundamental physical laws, which are gravity, the buoyancy force, and the drag force. The inertial distance is introduced, which is a typical distance at which a particle obtains the same wind velocity as its surroundings throughout its trajectory. For the measurement of turbulence intensity, either the Doppler spectral width or the variance of Doppler mean velocities is used. The relative scales of the inertial distance and the radar resolution volume determine whether the variance of velocities is increased or decreased for the same turbulence intensity. A decrease can be attributed to the effect that inertial particles are less responsive to the variations of wind velocities. An increase can be attributed to inertial particles that have wind velocities corresponding to an average of wind velocities over their backward trajectories, which extend outside the radar resolution volume. Simulations are done for the calculation of measured radar velocity variance, given a 3-D homogeneous isotropic turbulence field, which provides valuable insight in the correct tuning of parameters for the new model.@en