Search

An ensemble three‐dimensional ensemble‐variational (3DEnVar) data assimilation (En3DA) approach that directly assimilates radar reflectivity was developed based on the Weather Research and Forecasting model data assimilation system. This system adopts radar reflectivity as the control variable to avoid the need for a tangent linear and adjoint of the observation operator. Flow‐dependent covariance was introduced via ensemble forecasts updated by a group of 3DEnVar. The performance of the En3DA system was examined for two selected cases of high‐impact severe tornadic supercells over China. Results for both cases indicated that the structure of the storms in terms of intensity, coverage, and associated low‐level mesocyclones were analysed more accurately when using the En3DA approach than when adopting the 3DVar method. Hydrometeor analysis showed that En3DA provided a more physically reasonable increment of hydrometeors compared to 3DVar, especially for the graupel mixing ratio. Furthermore, the En3DA forecast was better than the 3DVar forecast throughout the forecast period for both studied cases. En3DA produced smaller errors in terms of intensity and location for supercell forecasts with respect to reflectivity and reflectivity swaths. Furthermore, the quantitative forecast skill of radar reflectivity was improved using En3DA. Errors in the wind, temperature, and water vapor forecast fields produced by En3DA were also reduced compared to those of 3DVar. Diagnostics revealed that En3DA predicted an enhanced low‐level cold pool and stronger outflows in the forward‐flank downdraft and the rear‐flank downdraft regions, which are important for tornadogenesis. [ABSTRACT FROM AUTHOR]

This two‐part study presents a comprehensive analysis of (1) the vortex structure, (2) the inner core planetary boundary layer (PBL) wind profile, and (3) the overland surface winds of Typhoon Faxai(2019) during landfall observationally (Part I) and in high‐resolution Weather Research and Forecasting Model (WRF) simulations (Part II). Part II presents a framework for evaluating the wind field in the WRF simulations of Typhoon Faxai (2019) by comparing them with the observations presented in Part I. First, WRF simulations with two different surface roughness z0$$ {z}_0 $$ tables are presented. The simulation with the default z0$$ {z}_0 $$ table in WRF is shown to largely overestimate the surface wind speed. The overestimation is largely mitigated by using larger z0$$ {z}_0 $$ based on the climatology for urban and forest areas in Japan. Next, using the modified z0$$ {z}_0 $$ table, WRF simulations with three different PBL parametrizations, namely, Mellor–Yamada–Janjić, Yonsei University, and Mellor–Yamada–Nakanishi–Niino schemes are presented, and their impacts on the simulated vortex structure, inner core PBL wind profile, and overland surface winds are evaluated. In particular, the simulated surface winds are in good agreement with the observations outside the inner core of Typhoon Faxai. However, during the passage of the eyewall, the maximum surface winds are underestimated in the simulations. We show that the underestimation of the inertial stability, associated with the excessively large inner core of Typhoon Faxai in the simulations, led to the overestimation of the eyewall PBL jet height. This, in turn, caused the underestimation of the peak surface winds, as based on the empirical model of the wind reduction factor F2 km$$ {F}_{2\;\mathrm{km}} $$ proposed in Part I. [ABSTRACT FROM AUTHOR]

This two‐part study presents a detailed analysis of (a) the vortex structure, (b) the inner core planetary boundary layer (PBL) wind profile, and (c) the overland surface winds of Typhoon Faxai (2019) during landfall observationally (Part I) and in high‐resolution Weather Research and Forecasting model simulations (Part II). Part I presents a comprehensive observational analysis of (a), (b), and (c) of Typhoon Faxai (2019) based on radar dual‐Doppler analysis and spatially dense surface wind observations over the Kanto region of Japan. An empirical model for the wind reduction factor F2 km$$ {F}_{2\;\mathrm{km}} $$ is proposed and validated with surface observations and radar velocity–azimuth display analyses at Narita and Haneda airports, both of which were affected by the passage of Faxai's inner core. Two distinct PBL jets associated with the inner rain band and the eyewall were observed over Narita Airport, which was located to the right of Faxai's track. Based on the empirical model of F2 km$$ {F}_{2\;\mathrm{km}} $$, it is found that the maximum surface wind speed observed at Narita Airport was caused by the peak of F2 km$$ {F}_{2\;\mathrm{km}} $$ due to a strong PBL jet associated with the eyewall located as close as ∼$$ \sim $$0.3 km from the ground. In addition, the wind swath analysis shows that a large area far inland on the right side of Faxai's track experienced strong surface 10‐min winds due to the combination of the passage of the eyewall PBL jet and the small surface roughness length around agricultural land such as paddy fields and farm fields. [ABSTRACT FROM AUTHOR]

The accurate prediction of high‐impact weather systems using cloud‐resolving models is still a challenge among researchers. This study evaluates the consistency of the combination of the three‐dimensional variational technique within the Gridpoint Statistical Interpolation assimilation system (GSI‐3DVAR) and nudging in the same modelling system on short‐range forecasts of three heavy rainfall events from the southwest monsoon season of 2021 over the Indian subcontinent. Three experiments have been conducted (i) control (CNTL): assimilated conventional and satellite observations; (ii) radar and lightning data assimilation (RLDA): assimilated radar reflectivity and lightning proxy reflectivity data along with all observations used in CNTL; and (iii) lightning data assimilation (LDA): same as RLDA but without the assimilation of radar data; particularly done to test the impact of assimilation of only lightning data. The model‐simulated rainfall is evaluated by using the Integrated Multi‐Satellite Retrievals (IMERG) for Global Precipitation Measurement (GPM IMERG) rainfall data. The intercomparison of LDA and RLDA for event 1 highlighted that both represent the convective regions reasonably better than CNTL, but RLDA outperforms LDA and thus further assimilation experiments are done with RLDA. RLDA provided reasonably accurate forecasts compared to CNTL, which is evident in the spatial distribution of rainfall and area‐averaged three‐hourly accumulated rainfall. Verification metrics for the three selected heavy rainfall events reveal that an optimal forecast performance (especially in the first six hours of free forecast) is obtained by the simulation with assimilation of radar and lightning data during the pre‐forecast period, through correcting the position and timing of convective centres. The probability of detection (POD) values are higher for light rainfall categories than for the heavy rain categories. POD values were higher in RLDA than CNTL throughout simulation for all three events. For all these three selected events, fractions skill scores (FSS) of RLDA are always better than CNTL with different neighbourhood sizes for different threshold values throughout the forecast period. [ABSTRACT FROM AUTHOR]

Radar profiling algorithms are widely used for precipitation‐induced path‐attenuation correction. In this study, we investigate the impact of the four well‐known profiling algorithms on the efficacy of quantitative precipitation estimation retrievals. The first method, named the "simple PhiDP‐based method", distributes precipitation‐induced path attenuation, or PhiDP (differential phase shift), of a rain segment to each radar gate according to radar reflectivity (Zh). The second one is a linear version of the ZPHI method derived from an approximate linear relationship between specific attenuation (Ah) and specific differential phase (KDP). The third one is a nonlinear version of the ZPHI that respects the nonlinearity of the Ah–KDP relationship. The fourth method is inspired from the self‐consistent method of Bringi. The two‐moment normalized raindrop size distribution (DSD) with two variables, scaled drop concentration Nw$$ {N}_{\mathrm{w}} $$ and mean drop diameter Dm$$ {D}_{\mathrm{m}} $$, is used to represent natural variability of DSD. Two configurations are considered: a constant Nw$$ {N}_{\mathrm{w}} $$ over a segment and a variable Nw$$ {N}_{\mathrm{w}} $$ over a segment. The Ah, KDP, and rain intensity R$$ R $$ are estimated by the four profiling algorithms. The sensitivity of each algorithm to radar measurements uncertainty is also investigated. The evaluation of the estimates shows the following. First, for the range‐constant Nw$$ {N}_{\mathrm{w}} $$ situation, the nonlinear ZPHI and the self‐consistent method yield the best estimates if the measurements errors are not considered. However, the nonlinear ZPHI and the self‐consistent method are significantly impacted by the Zh bias and the measurements noises respectively. Second, for the range‐variable Nw$$ {N}_{\mathrm{w}} $$ situation, the self‐consistent method does not have a good performance even when radar measurements are accurate. The performance of the nonlinear ZPHI method is similar to the simple PhiDP‐based method, whereas the latter is immune to the Zh calibration error. These results suggest that the benefit of introducing Nw$$ {N}_{\mathrm{w}} $$ in quantitative precipitation estimation algorithms can be significantly reduced if Nw$$ {N}_{\mathrm{w}} $$ is not constant in reality and/or if the radar measurements are biased and noisy. [ABSTRACT FROM AUTHOR]

Riming is a key process of precipitation formation in ice-containing clouds, but quantifying riming from observations is challenging, limiting our ability to evaluate the riming process in numerical weather models. One challenge for radar observations is that riming changes both the physical properties (mass, area cross-section) and scattering properties of ice particles. These changes need to be implemented consistently as a function of riming in radar forward operators, which are required for retrievals and model evaluation in observation space. In this study, mass-size, cross-section area-size, and backscattering cross-section relations are developed as a function of the normalized rime mass for aggregates composed of various monomer types (columns, dendrites, needles, plates, and rosettes). The proposed framework allows us to simulate scattering properties of aggregated ice particles consistently as a function of riming in retrievals and radar forward operators. The parameterizations are developed from a large data set of simulated rimed aggregates of different sizes and monomer crystal types. The backscattering cross-section parameterization (the "riming-dependent parameterization") is evaluated for radar frequencies of 35.6 and 94.0 GHz and is based on the Self-Similar Rayleigh-Gans approximation (SSRGA), which is increasingly used to calculate microwave scattering of ice crystals and snowflakes. Compared with parameterizations from the literature that do not consider riming, the riming-dependent parameterization leads to significantly smaller biases in terms of backscattering cross-section. When using the particle masses and scattering properties of the individual particles simulated by the aggregation and riming model as a reference, the bias of our parameterization is below 1 dB when integrating over an exponential particle size distribution with sizes from 0.1-10 mm. [ABSTRACT FROM AUTHOR]

Although cloud-affected satellite observations are heavily used for nowcasting applications, their use in regional data assimilation is very limited despite possible benefits for convective-scale forecasts. In this article, we estimate the potential impact of assimilating cloud-affected satellite observations of visible (0.6 µm) and near thermal infrared wavelengths (6.2 and 7.3 µm) relative to the impact of assimilating radar reflectivity observations. We employed observing-system simulation experiments with a perfect-model forecast for two cases of strong convective summertime precipitation. Observations are simulated using the radiative transfer model RTTOV/MFASIS and assimilated by the ensemble adjustment Kalman filter in the Data Assimilation Research Testbed. The Weather Research and Forecasting model at 2-km grid resolution was used for forecasts. Results show that satellite observations can be nearly as beneficial as three-dimensional radar reflectivity observations. Under favorable conditions, where the prior contains no error in the stage of storm development but only in horizontal position and strength, the assimilation of visible observations leads to 88% of the radar impact. Under more difficult conditions, the impact of visible and infrared observations still reached 50 and 79%, respectively. [ABSTRACT FROM AUTHOR]

Radar‐derived refractivity provides moisture information near the surface, which plays a crucial role in convective initiation and of heavy rain. In this study, the high‐resolution Weather Research and Forecasting local ensemble transform Kalman filter data assimilation system was employed for two cases in the Southwest Monsoon Experiment to conduct two experimental sets. The first set was applied in both cases to investigate the effects of assimilating radar‐retrieved refractivity along with reflectivity and radial wind. The second set was conducted in the second case to examine the benefit of increasing the frequency of refractivity assimilation and investigate the optimal strategy to assimilate refractivity. Results of the two cases revealed that assimilating reflectivity and radial velocity modified near‐surface humidity on the basis of the flow‐dependent error correlation estimated by the ensemble, but the spatial distribution may not be fully accurate, causing underestimation of rainfall. With further refractivity assimilation, stronger convergence and more accurate low‐level moisture, temperature, and wind field corrections were obtained, leading to superior forecasts for both light and heavy rainfall during six hours. The results of the second set indicated that increasing the assimilation interval of refractivity enabled capturing the dramatic moisture variation and enhancing wind convergence, resulting in short‐term forecast improvement. The strategy that solely assimilated refractivity before the appearance of the convection system optimized the correction of environmental moisture, then accurately represented the humidity and strengthened the wind convergence when precipitation occurred. Consequently, the achieved improvement of the short‐term forecast was most noteworthy, particularly for heavy rain. [ABSTRACT FROM AUTHOR]

The sparse distribution of the observing network over the high latitudes of the Southern Hemisphere reduces the accuracy of regional atmospheric circulation (re)analysis and weather forecasting. Using operational medium‐range ensemble forecasts, we compared the forecast skill among forecast centres regarding a midlatitude cyclone close to western Australia on 17 December 2017. At forecast day 4.5, a large ensemble spread in an upper‐level trough over western Australia, which had travelled from coastal Antarctica, caused large uncertainty and error in the predicted position of the midlatitude cyclone. Using an ensemble data assimilation system and the Atmospheric General Circulation Model for the Earth Simulator and the Local Ensemble Transform Kalman Filter Experimental Ensemble Reanalysis version 2 (ALERA2), we conducted an operational forecast experiment to investigate the impact on ALERA2 and its forecast result of assimilation of 1‐hourly horizontal wind speed observational data from the PANSY radar located at the Japanese Syowa Station in Antarctica. An observing system experiment revealed that incorporation of PANSY radar data improved the reproducibility of atmospheric parameters in the troposphere and lower stratosphere as initial conditions for forecasting. To investigate the impact of PANSY radar data on forecast skill regarding the midlatitude cyclone, 63‐member ensemble forecast experiments were conducted using analysis data both with and without PANSY data as initial conditions. Comparison of the forecast results from the ensemble forecast experiments confirmed that incorporation of PANSY radar data reduced the ensemble spread and error in the upper troposphere over the Antarctic region as initial conditions, which enhanced the accuracy of the prediction of the positions of the upper‐level trough and surface cyclone over western Australia at forecast day 4.5. These experiments suggest that the PANSY radar represents an observing system suitable for continuous improvement in forecast skill of the atmospheric circulation in the Southern Hemisphere. [ABSTRACT FROM AUTHOR]

Surface, upper‐air, and radar observations are used to assess the performance of the Antarctic Mesoscale Prediction System (AMPS) in simulating the mesoscale aspects of a wind and precipitation event over the Ross Island region of Antarctica that spanned January 16–20, 2016. The observations, collected during the Atmospheric Radiation Measurement (ARM) West Antarctic Radiation Experiment (AWARE), provide a unique dataset for evaluating AMPS, especially the radar observations that facilitate a three‐dimensional depiction of winds and precipitation. Comparisons of AMPS forecast data with surface meteorology, balloon‐sounding, and profiling radar observations at and above sites near McMurdo Station reveal a mixture of similarities and differences. A generally southerly flow is evident at low levels in both the AMPS simulations and observed Doppler radial velocities. AMPS winds are comparable to those observed at the surface and aloft in terms of magnitude, direction, and timing but the strongest simulated southerly flow is displaced eastward relative to the observations. AMPS‐simulated reflectivity over the broader Ross Island region is more limited in areal extent and smaller in magnitude than observed by a scanning Doppler radar. Three episodes of surface precipitation are observed near McMurdo Station over the five‐day event with peak rates of ∼3 mm h−1 and a total accumulation of ∼22 mm. However, AMPS produces no surface precipitation at that location over the five‐day event due to a low‐level dry bias in the forecasts. The results show the first observationally based three‐dimensional understanding of meteorology in the Ross Island region. [ABSTRACT FROM AUTHOR]

An assumption of uncorrelated observation errors is commonly adopted in conventional data assimilation. For this reason, high‐resolution data are resampled with strategies such as superobbing or data thinning. These strategies diminish the advantages of high temporal and spatial resolutions that can provide essential details in convection development. However, assimilating high‐resolution data, such as radar radial wind, without considering observation error correlations can lead to overfitting and thus degrade the performance of data assimilation and forecasts. This study uses a radar ensemble data assimilation system that combines the Weather Research and Forecasting model and Local Ensemble Transform Kalman Filter (WRF‐LETKF) to assimilate radar radial wind and reflectivity data. We present a strategy to include the error correlation of the Doppler radar radial wind in the WRF‐LETKF radar assimilation system and examine its impact on the accuracy of short‐term precipitation predictions based on a heavy rainfall event on June 2, 2017 in Taiwan. For radial wind, the horizontal error correlation scale is approximately 25 km according to the innovation statistics. The introduction of observation error correlation for radar radial wind assimilation produces more small‐scale features in wind analysis corrections compared to the experiment using an independent observation assumption. Consequently, the modification of wind corrections leads to stronger convergence accompanied by higher water vapor content, which enhances local convections. This results in more accurate simulations of reflectivity and short‐term precipitation. In particular, this advantage is identified for extreme heavy rainfall thresholds at small scales according to probability quantitative precipitation forecasts and fractions skill score. [ABSTRACT FROM AUTHOR]

We provide a detailed analysis of convectively generated cold pools (CPs) over flat midlatitude land, combining ten-year high-frequency time series of measurements at several heights available from the 213-m tower observatory at Cabauw, the Netherlands, with a collocated 2D radar rainfall dataset. This combination of data allows us to relate observations of the CP’s temporal and vertical structure to the properties of each CP’s parent rain cell, which we identify by rain-cell tracking. Using a new detection method, based on the anomalies of both the vertically averaged wind and the temperature, we monitor the arrival and passing of 189 CPs during ten summers (2010–2019). The time series show a clear signature of vortex-like motion along the leading CP edge in the vertical and horizontal wind measurements. The arrival of CP gust fronts is characterized by a steep decrease in both temperature and moisture, with a recovery time of approximately two hours. We see no evidence of moisture rings on the gust front edge, and therefore no indications for thermodynamic convective triggering. From the tower data, we obtain a median CP temperature drop of Tdrop ≈ −2.9 K and a height-averaged horizontal wind anomaly of Δumax ≈4.4 m⋅s−1. Relating the individual CP’s horizontal wind anomalies and temperature drops, we confirm the validity of the theoretical density current relationship, Δumax ∝ T1∕2 drop. We further propose a simple statistical model to relate the CP strength defined by Tdrop to the environmental properties influencing the CP: rain intensity and lower boundary-layer saturation. A multivariate linear regression suggests a 1 K colder CP for a 4 mm⋅hr−1 more intense rain cell (instantaneous area-averaged rain intensity) or for a 2.5 K larger pre-CP dew-point depression. [ABSTRACT FROM AUTHOR]

Monsoon trough (MT) convection exhibits multiscale spatial variability ranging from cellular storms to large organized convective systems. Understanding the relationships between storm‐scale convection and the large‐scale environment is crucial for accurately representing the properties and effect of convection in climate models. We investigate these relationships using nine wet seasons (June–September of 2009–2017) of S‐band radar observations over a tropical location (Kolkata) in the Indian summer monsoon trough region. A Lagrangian‐based objective cell‐tracking method, Thunderstorm Identification, Tracking, Analysis, and Nowcasting (TITAN), is applied to the volumetric reflectivity data to identify and track the convective storms. ERA5 hourly reanalysis products from the European Centre for Medium‐Range Weather Forecasts (ECMWF) are utilized for the field variables representing the large‐scale environment. We focus on the relationships of radar‐derived storm properties with large‐scale dynamic, thermodynamic, and stability variables such as convergence, mid‐level humidity, convective available potential energy (CAPE), and convective inhibition (CIN). Storm‐scale convection shows a selective nature, characterized as embedded cells in their unique environment. Furthermore, storms are classified as storm cells (simple and intense) and multicellular (linear and nonlinear) based on their geometrical properties and intensity. MT convection is predominantly composed of storm cells and characterized by an environment of moist mid‐level, low‐to‐moderate CAPE, low CIN, and strong dynamic forcing, that is, vertical ascent. In dry and weak dynamic forcing (i.e., subsiding atmosphere) with high CAPE and CIN, the highest rain intensity is associated with a few linear and nonlinear multicellular storms. In the development of such major linear and nonlinear multicellular storms over the eastern flank of the MT, the mid‐troposphere acts as a dry capping inversion. It is observed that the storm‐scale convection varies with the large‐scale dynamic forcing. During weak (strong) dynamic forcing conditions, the CAPE (mid‐level humidity) relates well with the observed storm‐scale convection. [ABSTRACT FROM AUTHOR]

High-resolution model simulations and radar observations are used to investigate the onset of vortex-genesis in a tornadic narrow cold-frontal rain band (NCFR). The timing and location of vortex-genesis was strongly constrained by a developing frontal wave, which tracked northeast across the United Kingdom and Ireland on 17 October 2011. In the simulations, vortices occurred preferentially during the early stages of wave development and just down-front of the wave centre, where large increases in vertical vorticity occurred in concert with decreases in the cross-frontal confluence. Vortex-genesis ceased as the frontal wave matured, due to the onset of frontal fracture. Two distinct scales of vortex-genesis are documented: primary vortex-genesis on themeso-γ-scale, and secondary vortex-genesis on the miso-scale. We show that horizontal shearing instability is the most likely vortex-genesis mechanism, consistent with previous theoretical work on the stability of vertical vortex strips in the presence of horizontal stretching deformation. Secondary vortices occurred along the braid regions between primary vortices where the shear zone became particularly narrow and intense. In the model, these vortices developed extremely rapidly (from small perturbations to maximum vertical vorticity in 5-15 min) and the strongest exhibited near-surface vertical vorticitymaxima approaching 10-1 s-1. Vortices of both scales were associated with characteristic local perturbations in the NCFR andwe show, by comparisonwith radar reflectivity data, that primary and secondary vortices were likely present in the real NCFR. Tornado reports were associated with small NCFR perturbations like those associated with the secondary vortices in the model simulations. Analysis of the sub-structure of individual simulated vortices suggests that tornado-genesis ismost likely within a region of intense near-surface vertical vorticity stretching at the north or northwest flank of the secondary vortices. [ABSTRACT FROM AUTHOR]

The local ensemble transform Kalman filter (LETKF) suggested by Hunt et al., 2007 is a very popular method for ensemble data assimilation. It is the operational method for convective‐scale data assimilation at Deutscher Wetterdienst (DWD). At DWD, based on the LETKF, three‐dimensional volume radar observations are assimilated operationally for the operational ICON‐D2. However, one major challenge for the LETKF is the situation where observations show precipitation (reflectivity) whereas all ensemble members do not show such reflectivity at a given point in space. In this case, there is no sensitivity of the LETKF with respect to the observations, and the analysis increment based on the observed reflectivity is zero. The goal of this work is to develop a targeted covariance inflation (TCI) for the assimilation of 3D‐volume radar data based on the LETKF, adding artificial sensitivity and making the LETKF react properly to the radar observations. The basic idea of the TCI is to employ an additive covariance inflation as entrance point for the LETKF. Here, we construct perturbations to the simulated observation which are used by the core LETKF assimilation step. The perturbations are constructed such that they exhibit a correlation between humidity and reflectivity. This leads to a change in humidity in such a way that precipitation is more likely to occur. We describe and demonstrate the theoretical basis of the method. We then present a case study where targeted covariance inflation leads to a clear improvement of the LETKF and precipitation forecast. All examples are based on the German radar network and the ICON‐D2 model over Central Europe. [ABSTRACT FROM AUTHOR]

Tropospheric and lower‐stratospheric motions at mesoscales and larger are a mixture of waves and two‐dimensional (2‐D) turbulence. Determining their relative importance is necessary, since waves are capable of coordinated systematic momentum transport accompanying the wave propagation, and associated wind forcing, in ways that 2‐D turbulence is not. This can impact weather forecasting. Using a network of ten windprofiler radars in eastern Ontario and western Quebec in Canada, plus an additional one in the Arctic, the relative roles of internal gravity (buoyancy) waves and two‐dimensional turbulence are examined at temporal scales from about 3–4 hrs to several tens of hours (horizontal spatial scales of typically one or two hundred kilometres to a few thousand kilometres), with the purpose of investigating the respective roles of these two distinct characteristic fluid motions as functions of location, season and year. The emphasis is on studies of spectral slope variability, rather than absolute spectral magnitudes, giving a perspective not previously substantially presented. In particular, we have found a frequency band in which gravity‐wave Doppler shifting produces distinctly different spectral slopes than those predicted for 2‐D turbulence, and these differences are employed to distinguish the flow fields. The network used (excluding the Arctic site) covers an area of ∼106 km2 and includes a variety of different terrains. Radial velocities have been recorded on time scales of minutes for data lengths covering durations of up to 12 years. Altitude coverage is from 1 km to typically 14 km, at 500 m resolution. Results suggest a region from ∼2 to ∼5 km altitude (deeper for some radars) where waves are weaker and 2‐D turbulence appears to be generally more significant, but where occasional bursts of gravity‐wave activity can occur, while above typically 6–8 km, gravity waves increase in significance. There are distinct site‐to‐site variations. [ABSTRACT FROM AUTHOR]

A stratosphere–troposphere (ST) wind profiler radar operating at a high VHF range (205 MHz) installed at Cochin (10.03°N, 76.33°E), India, provides high‐resolution winds in time and altitude extending from 315 m to 20 km. We evaluated the horizontal and vertical winds obtained from atmospheric global reanalyses (ERA5, ERA‐Interim, MERRA‐2 and NCEP) data and the high‐resolution regional model (WRF) output winds with the ST radar (STR) wind observations at Cochin. The data for one whole year, from 1 August 2018 to 31 July 2019, for 0600 UTC and 1200 UTC were used. Wind data appraisal has been carried out at 24 levels from 315 m to 18.5 km altitude for the reanalyses datasets and 29 levels for model‐simulated winds in different seasons. Vertical profiles of horizontal winds from reanalysis and model‐simulated data somewhat catch the major changes observed from STR. The zonal winds match very well with the STR, whereas meridional winds show moderate similarity. Higher correlations in zonal winds are found during the summer monsoon, whereas meridional winds show apparent likeness during winter. The ERA5, ERA‐Interim, and model‐simulated winds have better agreement with the STR winds. Observed vertical winds from STR displays large fluctuations, while the estimated vertical winds in reanalysis are highly smoothed. Vertical profiles of vertical winds from the reanalysis datasets are not uniform. The reliability of vertical winds in reanalysis and WRF model output values are dubious compared to that of the directly observed winds from STR. The study indicates that improved quality of vertical and meridional winds in reanalysis data is needed by assimilating directly observed data from the available advanced facilities. [ABSTRACT FROM AUTHOR]

Assimilating radar reflectivity into NWP models is one of the keys to improve the accuracy of convective‐scale numerical weather prediction (NWP). There are generally two major branches of research in radar reflectivity assimilation: directly assimilating radar reflectivity, and indirectly assimilating reflectivity, i.e. assimilating hydrometeors retrieved from radar reflectivity. In this study, the indirect assimilation method based on background‐dependent hydrometeor retrievals is compared with the direct assimilation method using frequent data assimilation cycles for five real data cases. The retrieved hydrometeors in the indirect assimilation method are first verified against the hydrometeor types obtained from a polarimetric hydrometeor classification method. It is illustrated that the background‐dependent hydrometeor retrieval method can obtain reasonable model‐equivalent hydrometeors from radar reflectivity. The analysis increments for hydrometeors with both radar data assimilation methods show similar patterns but their magnitudes are different. Both quantitative and qualitative evaluations of forecasted composite reflectivities and accumulated precipitation indicate that the indirect assimilation method predicts the location and intensity of the simulated convection more accurately than the direct method. Furthermore, the indirect assimilation method is more efficient, which is valuable in real‐time applications by helping deliver forecasts quickly and thus helping forecasters make more timely warning decisions. [ABSTRACT FROM AUTHOR]

Reliable forecasting of heavy convective rainfall events continues to be one of the greatest challenges for numerical weather prediction models. Studies across the globe have indicated that Doppler weather radar (DWR) data assimilation through three‐dimensional variational methods (3D‐Var) has a considerable impact on improving the forecast skill of convective precipitation. However, the time‐averaged background error statistics (BES) employed in 3D‐Var fail to address the day‐to‐day variability of forecast errors in BES. To effectively represent the convective precipitation events through DWR assimilation, it is important to account for the day‐to‐day variability in background error's variance and correlations in the BES. In this study, the effect of assimilating DWR data using flow‐dependent BES within the variational assimilation system is analysed via the hybrid Ensemble Transform Kalman Filter (ETKF)‐3DVAR. For this, the functionality of ETKF in the hybrid ETKF‐3DVAR has been extended to assimilate both DWR radial velocity and reflectivity information. The results suggest that assimilating DWR observations through the hybrid ETKF‐3DVAR improves the forecast skills of convective precipitation as compared to the 3D‐Var. It has also been observed that updating the ensemble states through assimilation of DWR observations in the ETKF system shows an increased wind and moisture convergence. [ABSTRACT FROM AUTHOR]

We present a revised local wavelet‐based organization index (LW), which is applied to both synthetic fields and meteorological fields of precipitation and brightness temperature for different types of convective organization. The LW consists of three components that describe the scale (LWsc), quantify the intensity (LWin), and analyze the anisotropy (LWai) of the convection. It is based solely on 2D wavelet decomposition and does not require a clustering algorithm or thresholding. The great advantages of LW are that it localizes the organization in space, is a universally applicable measure of organization and can be calculated with little effort. A comparison with other organizational metrics shows that LW better describes the structure and organization of convection. We analyze the LW components in the vicinity of severe weather reports related to single‐cell storms, supercells or multi‐cell storms. Composites thereof reveal that large hail and heavy precipitation are particularly local events that lead to small‐scale and very intense precipitation. Especially in the case of hail, LWai shows the structure of a hook echo, which is expected, since hail events are usually associated with supercells. In relation to wind gusts, precipitation is rather large‐scale and linearly oriented, since strong wind gusts are often a consequence of linearly organized systems such as squall lines or derechos. The brightness temperature analysis provides similar results to those of the rain rates. Structures like the typical radar hook echo are not visible in the satellite data, but overshooting tops caused by strong updraughts are recognized as small‐scale, intense regions. The three LW components are used to classify showers, thunderstorms and precipitation/hail events with the help of a neural network, where showers are mainly classified by their scale characteristics, and precipitation/hail events by their pronounced intensity. [ABSTRACT FROM AUTHOR]

An algorithm is presented that aims to enhance the use of satellite‐based cloud‐profiling radar (CPR) data for the purpose of NWP data assimilation. It resembles the EarthCARE mission's scene construction algorithm: off‐nadir passive radiances get spectrally matched with nadir passive radiances, and the latter's collocated CPR profile gets replicated at the former's location. This process gets repeated until all passive pixels that cover an NWP domain D have a proxy column of CPR data. Only domain‐averaged profiles of CPR reflectivity ⟨ZdB⟩ and cloud fraction Ac are sought after for NWP domains measuring, nominally, 25 × 25 km. If CPR measurements intersect D, then in addition to using just the intersecting values to represent ⟨ZdB⟩ and Ac, the full array of proxy values get used. This is referred to as local estimation. If, however, CPR measurements do not intersect D, but are not too distant, this is referred to as non‐local estimation; estimates of ⟨ZdB⟩ and Ac rest entirely on proxies. Current and planned satellite‐based CPRs have nadir‐pointing narrow fields‐of‐view, so it is difficult to see how to verify the algorithm with anything other than synthetic observations. Thus, it was assessed here with simulated cloudy atmospheres and 2D and 3D distributions of associated passive radiances and CPR reflectivities. In general terms, for local domains the algorithm performs slightly better than simply averaging intersecting CPR profiles. This at least demonstrates that the proxies are not detrimental. For non‐local domains, where there are no intersecting measurements to average, the algorithm appears to perform well enough to include two or three 25 × 25 km NWP domains either side of sequences of local domains. [ABSTRACT FROM AUTHOR]

A new method has been developed to directly assimilate volume scans of radar reflectivity data with 4D‐Var in the Met Office Unified Model. The method has been demonstrated in the convective‐scale hourly‐cycling UKV forecast model. Reflectivity observations from 18 C‐band radars in the British Isles are assimilated. This article describes the method of observation processing and quality control, the observation operator, and assimilation method. The assimilation method uses a minimum threshold rainwater content in the forward operator to give sensitivity to reflectivity observations where there is no rain in the background. Furthermore, the use of the Huber norm in the observation penalty function allows the use of observations with large innovations in the assimilation. A change was made to the precipitation efficiency in the microphysics scheme of the linear perturbation forecast model to ensure stability of the scheme. A case‐study is presented which demonstrates how the inclusion of reflectivity observations enhances convergence through analysis increments to the wind field, leading to improvements to the location of convective precipitation features in the forecast. Two‐month trials for summer and winter seasons demonstrate significant improvements to rain forecasts in the nowcasting range. [ABSTRACT FROM AUTHOR]

The Discrete Dipole Approximation (DDA) is widely used to simulate scattering of microwaves by snowflakes, by discretising the snowflake into small ‘dipoles’ which oscillate in response to (a) the incident wave, and (b) scattered waves from all the other dipoles in the particle. It is this coupling between all dipole pairs which makes solving the DDA system computationally expensive, and that cost grows nonlinearly as the number of crystals n within an aggregate is increased. Motivated by this, many studies have ignored the dipole coupling (the Rayleigh–Gans Approximation, RGA). However, use of RGA leads to systematic underestimation of both scattering and absorption, and an inability to predict polarimetric properties. To address this, we present a new approach (the Independent Monomer Approximation, IMA) which solves the DDA system for each crystal ‘monomer’ separately, then combines them to construct the full solution. By including intra-monomer coupling, but neglecting inter-monomer coupling, we save a factor of n in computation time over DDA. Benchmarking IMA against DDA solutions indicates that its accuracy is greatly superior to RGA, and provides ensemble scattering cross-sections which closely agree with their more expensive DDA counterparts, particularly at size parameters smaller than ∼ 5 . Addition of rime to the aggregates does not significantly degrade the results, despite the increased density. The use of IMA for radar remote sensing is evaluated, and we show that multi-wavelength and multi-polarisation parameters are successfully captured to within a few tenths of a dB for aggregates probed with frequencies between 3 and 200 GHz, in contrast to RGA where errors of up to 2.5 dB are observed. Finally, we explore the realism of the IMA solutions in greater detail by analysing internal electric fields, and discuss some broader insights that IMA provides into the physical features of aggregates that are important for microwave scattering. [ABSTRACT FROM AUTHOR]

The assimilation of radar reflectivity data is of great interest for numerical weather prediction (NWP) models at the convective scale since these observations are very dense both in space and time and they are related to microphysical prognostic variables. So far, the assimilation of reflectivity in operational frameworks has been limited to derived products, like precipitation rates and humidity profiles. This approach is also adopted at ARPAE-SIMC, the Hydro-Meteo-Climate Structure of the Regional Agency for Prevention, Environment and Energy of Emilia-Romagna region in Italy, where latent heat nudging (LHN) is employed to assimilate precipitation rates in the COSMO-2I model, the convection-permitting version of the regional model developed by the Consortium for Small-scale Modelling (COSMO) which is run operationally to provide high-resolution forecasts over Italy. However, to fully exploit the information contained in these observations, the whole reflectivity volume should be directly assimilated. Nevertheless, despite several promising studies, its implementation in an operational framework has not yet been achieved. In this study, a set-up designed to assimilate reflectivities operationally in the COSMO-2I model through a local ensemble transform Kalman filter (LETKF) has been evaluated over 37 days in 2018 and performed 303 forecasts. The comparison with the current operational set-up based on LHN reveals an average improvement in quantitative precipitation forecasts (QPFs) up to 7 hr, even if the impact on convective cases is much stronger than that observed in conditions of stratiform precipitation. Moreover, a small but positive impact is noticed for the RMSE of upper-air variables, while the impact on bias is mixed. Mixed results are also obtained when considering surface variables. In the light of the results of this study, a pre-operational parallel system in which COSMO-2I analyses are generated replacing LHN by the direct assimilation of reflectivity volumes was implemented in April 2020. [ABSTRACT FROM AUTHOR]

Observations related to cloud, such as radiances from microwave imagers, have been at the forefront of recent developments in data assimilation for numerical weather prediction (NWP). While they offer unrivalled spatial coverage, they contain limited information on the vertical structure of clouds. In contrast, active observations from profiling instruments such as cloud radar and lidar contain a wealth of information on the structure of clouds and precipitation, providing the much‐needed vertical context of clouds, but have never been assimilated directly in global NWP models. To explore the potential benefits of these profiling observations, the European Centre for Medium‐Range Weather Forecasts (ECMWF) Four‐Dimensional Variational (4D‐Var) data assimilation system has been recently adapted to allow direct assimilation of cloud profile observations from space‐borne radar and lidar instruments. In this paper, in conjunction with its companion paper, the first‐time direct assimilation of cloud radar and lidar observations into a global NWP model is demonstrated. Using CloudSat radar reflectivity and CALIPSO attenuated backscatter shows that the assimilation brings the analysis closer to these observations and has a mainly neutral affect on other assimilated observations. Some improvements in the forecast skill are also observed when verified against the experiment's own analysis, with the largest positive impact noticed for temperature at the lowest model levels and for vector wind above 500 hPa, but longer experiments are required to reach 95% statistical significance of the results. The potential improvements in the model radiation budget is explored by verifying with Clouds and the Earth's Radiation Energy System (CERES) observations. Sensitivity of the results to observation error and to the observation reduction by increased averaging is also discussed. The demonstration of statistically significant improvements to forecast skill in some metrics without any significant degredation in others shows great promise for the future use of cloud radar and lidar observations in NWP. [ABSTRACT FROM AUTHOR]

The direct assimilation of space‐borne cloud radar and lidar observations into a global numerical weather prediction model has not previously been attempted for several reasons. Firstly, the modification of a data assimilation system to handle space‐borne profiling observations is a technical challenge. Secondly, the relationship between model‐scale control variables and the relatively narrow footprint of the radar and lidar instruments were thought to be unrepresentative and too nonlinear for a variational assimilation system that is based on assumptions of linearity. However, motivation was provided when previous experiments assimilating cloud radar and lidar profiles showed, using an intermediary step to generate pseudo‐observations of temperature and humidity, that there was potential for the observations to improve both the analysis and the subsequent forecast. This article presents the developments made to facilitate the direct assimilation of cloud radar and lidar observations into the four‐dimensional variational assimilation (4D‐Var) system of the European Centre for Medium‐Range Weather Forecasts (ECMWF). A review of the observation operators shows how they have been optimised for data assimilation with an emphasis on efficiency, consistency and differentiability. A key part of this work is the specification of a new fully flow‐dependent characterisation of the observation error. By taking an error inventory approach, we account for different sources of error in a physical way. Also, to avoid degrading the analysis, careful screening, quality control and bias correction is necessary and is described herein. Finally, an initial assessment of the impact of the observations is achieved through single‐analysis tests that show that the analysis fit to the new observations is improved. In‐depth results demonstrating the impact of these observations are shown in the second part of this two‐part series. [ABSTRACT FROM AUTHOR]

Vertically pointing radar observations combining multiple frequencies and Doppler measurements have been recently shown to contain valuable information about ice particle growth processes, such as aggregation and riming. In this study, we use a two‐months X, Ka, W‐Band Doppler radar dataset of midlatitude winter clouds to infer statistical growth signatures of ice and snow particles. The observational statistics are compared to forward‐simulated radar moments based on simulations of the campaign time period with a high‐resolution version of the ICON model and a two‐moment microphysical scheme. The statistical comparison shows very good agreement of the simulated vertical structure of radar reflectivity and surface precipitation rate. The dual‐wavelength ratios, which are closely related to the mean particle size, also show consistently a major increase at temperatures higher than –15 °C. However, at temperatures higher than –7 °C, ICON increasingly overestimates the mean particle size. The statistics of mean Doppler velocities also reveal that the model overestimates the terminal velocity of snow particles, especially at larger sizes. We discuss possible reasons for the identified discrepancies, such as an unrealistic temperature dependence of the sticking efficiency or the non‐saturation of terminal velocities at larger sizes caused by the implemented power law relations. Our study demonstrates examples of the importance of combining various radar techniques for identifying issues in simulated microphysical processes, which can otherwise be hidden due to compensating errors. [ABSTRACT FROM AUTHOR]

Hailstorms are relatively frequent in the Friuli Venezia Giulia region, northeastern Italy and, for that reason, a network of manual hailpads has been set up there since the late 1980s. On July 4, 2007, a record number of hailpads in the network were impacted by hail (115 out of about 360 total stations). To the best of the authors' knowledge, no other single hailstorm case‐study has information coming from such a large number of hailpads impacted by hail. These hailpads are analysed using automatic software, able to interpret each hail dent as an ellipse. From these data, the distribution of hailstone diameters, areal hailstone density on the pads and total kinetic energy flux are computed and analysed in the present study. A parallel analysis of radar maximum reflectivity and cloud‐to‐ground lightning shows that the evolution of the hailstorm can be divided into three different phases, the first being the most severe in terms of lightning and hail. Finally, numerical simulations performed with a non‐hydrostatic mesoscale model, implementing a prognostic module for the explicit simulation of hail (WRF‐HAILCAST), show that the simulation system provides useful information on the severity of the event in terms of maximum hailstone diameter and reflectivity shape. [ABSTRACT FROM AUTHOR]

Operational forecasting centres increasingly rely on convection‐permitting NWP simulations to assist in their forecasting of convective events. The evaluation of upgrades in the underlying NWP modelling system normally happens through routine verification using traditional metrics on two‐dimensional fields, such as gridded rainfall data. Object‐ and process‐based evaluation can identify specific physical mechanisms for model improvement, but such evaluation procedures normally require targeted and expensive field campaigns. Here, we explore the potential use of the UK operational radar network observations and its derived 3D composite product for evaluating the representation of convective storms in the Met Office Unified Model. A comparison of the 1 × 1 × 0.5 km 3D radar composites against observations made with the research‐grade radar at Chilbolton in the southern UK indicates that the 3D radar composite data can reliably be used to evaluate the morphology of convective storms. The 3D radar composite data are subsequently used to evaluate the development of convective storms in the Met Office Unified Model. Such analysis was previously unavailable due to a lack of 3D radar data of high temporal frequency. The operational nature of the UK radar data makes these 3D composites a valuable resource for future studies of the initiation, growth, development, and organisation of convective storms over the UK. [ABSTRACT FROM AUTHOR]

The impact of assimilating differential reflectivity ZDR in addition to reflectivity (ZH) and radial velocity (Vr) from a polarimetric radar on the analysis of a tornadic supercell storm using an ensemble Kalman filter (EnKF) is studied in an observing system simulation experiment (OSSE) framework assuming a perfect forecast model. A double‐moment microphysics scheme is used to allow for proper simulation of polarimetric signatures. Root‐mean‐square errors of analysed state variables are calculated and the structure and intensity of analysed fields and derived quantities are examined. Compared to the baseline experiment assimilating radial velocity and reflectivity only, the assimilation of additional ZDR further reduces the errors of all state variables. The analysed hydrometeor fields are improved in both pattern and intensity distributions. Polarimetric signatures including ZDR and KDP columns, and ZDR arc in the supercell, are much better reproduced. Sensitivity experiments are performed that exclude the updating of hydrometeor number concentrations by ZDR or of state variables not directly linked to ZDR via observation operators. The results show that if number concentrations are not updated together with the mixing ratios, most of the benefit of assimilating ZDR is lost. Among other state variables, the updating of water vapour mixing ratio qv has the largest positive impact while the impact of updating vertical wind w comes in second. The updating of horizontal wind components or temperature has a much smaller but still noticeable impact. Reliable flow‐dependent cross‐covariances among the state variables and observation prior as derived from the forecast ensemble and used in EnKF are clearly very beneficial. [ABSTRACT FROM AUTHOR]

Severe hailstorms represent some of Australia's most costly natural disasters. In particular, the densely populated east‐coast cities of Brisbane and Sydney have both been impacted by events that individually produced over AU$1 billion in insured losses. In this study, the frequency of damaging hailstorms in these regions is assessed over an eight‐year period using single‐polarisation radar observations. Reflectivity data from multiple radars are carefully calibrated and merged to produce daily grids of MESH (the maximum expected size of hail) at 1 km spatial resolution. Neighbourhood contingency tables are then used to define an optimal MESH threshold for identifying damaging hail events. This analysis makes use of a unique insurance dataset, comprising building‐scale exposure and claim data over a multi‐year period. Overall, damaging hailstorms occur on an average of 26 and 32 days per year in the Brisbane and Sydney regions, respectively. The actual cities of Brisbane and Sydney both experience about five damaging hailstorms annually. In both regions, the highest hail frequencies occur along the coastal slopes of the Great Dividing Range, with more localised maxima highlighting the potential role of mesoscale circulations in storm initiation and intensification. For Brisbane, the seasonal cycle shows a pronounced peak in December, while in Sydney hail frequencies are high throughout the warm season (November–March). This difference reflects the distinct seasonal cycles of convective instability in the two regions. Consistent with previous studies, the combination of CAPE and deep‐layer shear is found to provide reasonable discrimination of environments conducive to severe hailstorms. [ABSTRACT FROM AUTHOR]

The aim of this study is to provide an evaluation of the impact of two largely used data assimilation techniques, namely three‐ and four‐dimensional variational data assimilation systems (3D‐Var and 4D‐Var), on the forecasting of heavy precipitation events using the Weather Research and Forecasting (WRF) model. For this purpose, two flash flood events in central Italy are analysed. The first occurred on September 14, 2012 during an Intensive Observation Period of the Hydrological cycle in the Mediterranean experiment (HyMeX) campaign, while the other occurred on May 3, 2018. Radial velocity and reflectivity acquired by C‐band weather radars at Mt. Midia (central Italy) and San Pietro Capofiume (northern Italy), as well as conventional observations (SYNOP and TEMP), are assimilated into the WRF model to simulate these damaging flash flood events. In order to evaluate the impact of the 3D‐Var and 4D‐Var assimilation systems on the estimation of short‐term quantitative precipitation forecasts, several experiments are carried out using conventional observations with and without radar data. Rainfall evaluation is performed by means of point‐by‐point and filtering methodologies. The results point to a positive impact of the 4D‐Var technique compared to results without assimilation and with 3D‐Var experiments. More specifically, the 4D‐Var system produces an increase of up to 22% in terms of the Fractions Skill Score compared to 3D‐Var for the first flash flood event, while an increase of about 5% is achieved for the second event. The use of a warm start initialization results in a considerable reduction in the spin‐up time and a significant improvement in the rainfall forecast, suggesting that the initial precipitation spin‐up problem still occurs when using 4D‐Var. [ABSTRACT FROM AUTHOR]

High‐intensity precipitation represents a threat for several regions of the world because of the related risk of natural disasters (e.g. floods and landslides). This work focuses on low‐level precipitation enhancement that occurs in the cloud warm layer and has been observed in relation to collision‐coalescence (CC) leading to flash floods and extreme rainfall events in tropical and temperate latitudes. Specifically, signatures of precipitation enhancement (referred to as CC‐dominant precipitation) are investigated in the observations from the Global Precipitation Measurement (GPM) core mission Dual‐frequency Precipitation Radar (DPR) over the central/eastern Contiguous United States (CONUS) during June 2014–May 2018. A classification scheme for CC‐dominant precipitation, developed for dual‐polarization S‐band radar measurements and applied in a previous work to X‐band radar observations in complex terrain, is used as a benchmark. The scheme is here applied to the GPM ground validation dataset that matches ground‐based radar observations across CONUS to space‐borne DPR retrievals. The occurrence of CC‐dominant precipitation is documented and the corresponding signatures of CC‐dominant precipitation at Ku‐ and Ka‐band are studied. CC‐dominant profiles show distinguishing features when compared to profiles not dominated by CC, e.g. characteristic vertical slopes of reflectivity at Ku‐ and Ka‐band in the liquid layer, lower freezing‐level height, and shallower ice layer, which are linked to environmental conditions driving the peculiar CC microphysics. This work aims at improving satellite quantitative precipitation estimation, particularly GPM retrievals, by targeting CC development in precipitation columns. [ABSTRACT FROM AUTHOR]

Deep moist convection shows a tendency to organize into mesoscale structures. To be able to understand the potential effect of convective organization on the climate, one needs first to characterize organization. In this study, we systematically characterize the organizational state of convection over Germany based on two years of cloud‐top observations derived from the Meteosat Second Generation satellite and of precipitation cores detected by the German C‐band radar network. The organizational state of convection is characterized by commonly employed organization indices, which are mostly based on the object numbers, sizes and nearest‐neighbour distances. According to the organization index Iorg, cloud tops and precipitation cores are found to be in an organized state for 69% and 92% of the time, respectively. There is an increase in rainfall when the number of objects and their sizes increase, independently of the organizational state. Case‐studies of specific days suggest that convectively organized states correspond to either local multi‐cell clusters, with less numerous, larger objects close to each other, or to scattered clusters, with more numerous, smaller organized objects spread out over the domain. For those days, simulations are performed with the large‐eddy model ICON with grid spacings of 625, 312 and 156 m. Although the model underestimates rainfall and shows a too large cold cloud coverage, the organizational state is reasonably well represented without significant differences between the grid spacings. [ABSTRACT FROM AUTHOR]

Convective initiation is a challenge for convection‐permitting models due to its sensitivity to sub‐km processes. We evaluate the representation of convective storms and their initiation over South Africa during four summer months in Met Office Unified Model simulations at a 1.5‐km horizontal grid length. Storm size distributions from the model compare well with radar observations, but rainfall in the model is predominantly produced by large storms (50 km in diameter or larger) in the evening, whereas radar observations show that most rainfall occurs throughout the afternoon, from storms 10–50 km in diameter. In all months, the modelled maximum number of storm initiations occurs at least 2 hr prior to the radar‐observed maximum. However, the diurnal cycle of rainfall between the model and observations compares well, suggesting that the numerous storm initiations in the simulations do not produce much rainfall. Modelled storms are generally less intense than those in radar observations, especially in early summer. In February, when tropical influences dominate, the simulated storms are of similar intensity to observed storms. Simulated storms tend to reach their peak intensity in the first 15 min after initiation, then gradually become less intense as they grow. In radar observations, storms reach their peak intensity 15 min into their life cycle, stay intense as they grow larger, then gradually weaken after they have reached their maximum diameter. Two November case studies of severe convection are analysed in detail. A higher resolution grid length initiates convection slightly earlier (300 m as opposed to 1.5 km) with the same scientific settings. Two 1.5 km simulations that apply more subgrid mixing have delayed convective initiation. Analysis of soundings indicates little difference in the convective indices, suggesting that differences in convection may be attributed to the choice of subgrid mixing parameters. [ABSTRACT FROM AUTHOR]

Turbulent mixing processes are important in determining the evolution of convective clouds, and the production of convective precipitation. However, the exact nature of these impacts remains uncertain due to limited observations. Model simulations show that assumptions made in parametrizing turbulence can have a marked effect on the characteristics of simulated clouds. This leads to significant uncertainty in forecasts from convection‐permitting numerical weather prediction (NWP) models. This contribution presents a comprehensive method to retrieve turbulence using Doppler weather radar to investigate turbulence in observed clouds. This method involves isolating the turbulent component of the Doppler velocity spectrum width, expressing turbulence intensity as an eddy dissipation rate, ϵ. By applying this method throughout large datasets of observations collected over the southern United Kingdom using the (0.28° beam‐width) Chilbolton Advanced Meteorological Radar (CAMRa), statistics of convective cloud turbulence are presented. Two contrasting case days are examined: a shallow "shower" case, and a "deep convection" case, exhibiting stronger and deeper updraughts. In our observations, ϵ generally ranges from 10−3 to 10−1 m2/s3, with the largest values found within, around and above convective updraughts. Vertical profiles of ϵ suggest that turbulence is much stronger in deep convection; 95th percentile values increase with height from 0.03 to 0.1 m2/s3, compared to approximately constant values of 0.02–0.03 m2/s3 throughout the depth of shower cloud. In updraught regions on both days, the 95th percentile of ϵ has significant (p < 10−3) positive correlations with the updraught velocity, and the horizontal shear in the updraught velocity, with weaker positive correlations with updraught dimensions. The ϵ‐retrieval method presented considers a very broad range of conditions, providing a reliable framework for turbulence retrieval using high‐resolution Doppler weather radar. In applying this method across many observations, the derived turbulence statistics will form the basis for evaluating the parametrization of turbulence in NWP models. Turbulent mixing processes are important in determining the evolution of convective clouds, and the production of convective precipitation. To investigate turbulence in observed clouds, this contribution provides a comprehensive method to retrieve turbulence using Doppler weather radar. Statistics of convective cloud turbulence are presented through the analysis of retrievals (example in picture) over many observations in two contrasting case‐studies. [ABSTRACT FROM AUTHOR]

Estimation of precipitation from space‐based passive microwave (PMW) radiometric brightness temperature (TB) observations that adapts to the wide variety of Earth surface background and environmental conditions is a long‐standing issue. Since these conditions are generally unknown from the TB observations, PMW‐based precipitation estimation techniques commonly utilize independent ancillary data sources, such as interpolated prognostic variables from numerical weather prediction forecast models, and discrete surface emissivity classifications. In some situations, the selection of these variables may restrain the algorithm performance under particular surface and atmospheric conditions. The objective of this article is to examine the emissivity principal component (EPC) analysis as a common stratification method for indexing, searching and weighting candidate precipitation profiles from a priori databases, adaptable for Bayesian‐based precipitation estimation algorithms applied to the Global Precipitation Measurement (GPM) Microwave Imager (GMI) or other PMW sensors, to minimize dependence upon ancillary data sources. The EPC has been previously shown to track the joint variability between the 10–89 GHz surface emissivity, total column precipitable water vapour (TPW) and surface temperature (Ts) conditions directly from the TB observations, and identify global locations of similar conditions. A parallel GMI precipitation retrieval was carried out where the identical a priori database was indexed by TPW, Ts and a surface emissivity class index. An independent validation of each precipitation retrieval scheme was carried out using GMI pixel‐matched Multi‐Radar Multi‐Source (MRMS) ground radar data over the continental USA and surrounding ocean waters. While the EPC‐based estimates demonstrated similar performance to the TPW‐based estimates over ocean backgrounds, a markedly improved detection, and reduction in bias, was found for moderate and higher (>5 mm/hr) rainfall rates over other backgrounds, especially vegetated surfaces and coastlines. A new microwave surface emissivity‐based approach for indexing and searching extensive a priori passive microwave radiometric brightness temperature (TB) observations is proposed, appropriate for Bayesian‐based precipitation retrieval algorithms. This method identifies similar global surface conditions from within the extensive collection of a priori observations. The overall bias and probability of detection of precipitation rates greater than 10 mm/hr over land and coastal scenes are improved, relative to surface radar‐derived precipitation rates over the continental United States and surrounding waters. [ABSTRACT FROM AUTHOR]

We have evaluated the applicability of the Rayleigh–Gans approximation (RGA) and self‐similar Rayleigh–Gans approximation (SSRGA) for microwave scattering by rimed snowflakes. This study extends previous findings that showed that, for unrimed snowflakes, the RGA is in good agreement with the discrete dipole approximation (DDA), which we used as a reference method. When riming is introduced, the RGA‐derived scattering properties of individual snowflakes deviate significantly – up to 20–25 dB for the backscattering cross‐section in the W band – from the corresponding DDA results. In contrast, the average scattering properties given by RGA are in good agreement with DDA for all but the most heavily rimed snowflakes: the mean bias in the backscattering cross‐section rarely exceeds 1 dB for light and moderate riming. We also found that an adjustment that accounts for the non‐spherical shapes of ice crystals can help eliminate the small constant bias found in RGA in earlier studies. The SSRGA approximates the RGA results with good accuracy at all degrees of riming, indicating that it, too, can be used with up to moderately rimed snowflakes. We evaluate the usability of the Rayleigh–Gans approximation (RGA), a simple and computationally efficient scattering approximation, for microwave scattering by rimed snowflakes. The RGA compares well with reference calculations for all but the most heavily rimed snowflakes. The self‐similar Rayleigh–Gans approximation (SSRGA), which combines RGA with assumptions from fractal geometry, can also reproduce the RGA results well for rimed snowflakes. [ABSTRACT FROM AUTHOR]

In this study, we present a unique 15‐year hail streak climatology for Switzerland based on volumetric radar reflectivity. Two radar‐based hail detection products and an automatic thunderstorm‐tracking algorithm were reprocessed for the extended convective season (April–September) between 2002 and 2016. More than 1.1 million convective cells were automatically tracked over the full radar domain, and over 191,000 storms and 31,000 hail streaks in the considered subdomain were selected for analysis following consistency and robustness tests. The year‐to‐year variability in the number of hailstorms reveals two types of convective seasons: (a) a few seasons with hail frequency far above the average, and (b) all other years with an average number of hailstorms. A high number of hailstorms in a particular year is not correlated with a higher number of convective storms in general, but is related to a greater fraction of severe storms. Convection initiation, hail initiation, and hail frequency maxima are located along the southern and northern foothills over the pre‐Alpine area and over the Jura mountains. Few hail streaks are present over the Alpine main ridge. Hail streak frequency and location is found to be strongly dependent on the synoptic‐scale weather regimes. This is important for monthly and seasonal outlooks, as well as for climate modelling. Analysis of storm life cycles shows that: (a) the majority of hail swaths contain only a single hail streak, (b) severe storms follow a more rapid evolution during their initial stages than do less severe storms, and (c) severe storms produce more spatially extended hail streaks. Finally, significant seasonal and diurnal cycles are present in most of the considered storm characteristics. Volumetric radar reflectivity is used to track and analyse hailstorms in Switzerland and adjacent areas. Two radar‐based hail detection products and an automatic thunderstorm tracking algorithm have been reprocessed for the convective season between 2002 and 2016. A unique 15‐year (2002–2016) severe storm climatology, which includes more than 191,000 storms and 31,000 hail streaks is presented. Image courtesy of Ioannis Sideris (MeteoSwiss) and Google Earth 2010. [ABSTRACT FROM AUTHOR]

This article describes the results of assimilation experiments of dual‐polarization (DPOL) radar observations in the AROME model with its 3D‐Var assimilation system. The assimilation of dual‐polarization observations is derived from the current operational assimilation method for radar reflectivities. It has two steps (1D+3D‐Var). Relative humidity pseudo‐observations are first retrieved from observed profiles of reflectivity (Zhh), with or without attenuation correction, from differential reflectivity (Zdr), from specific differential phase (Kdp), or from a combination of Zhh, Zdr and Kdp, through a unidimensional (1D) Bayesian inversion. The retrievals are then assimilated in the AROME 3D‐Var system. In the case of attenuation due to precipitation, the values of the relative humidity profiles retrieved with uncorrected reflectivity were found lower on average than with attenuation‐corrected reflectivity or with Kdp. In regions with large values of Kdp, both methods using Kdp produced the profiles with the largest values of relative humidity. Assimilation experiments were conducted for two convective cases. They demonstrated the benefit of the assimilation of Kdp and of attenuation‐corrected Zhh in the model analysis. The humidity increments were found higher in convective regions when assimilating either Kdp or attenuation‐corrected Zhh. Furthermore, the impact of DPOL observations on precipitation forecasts was found slightly positive for the periods of intense convection. This article describes an assimilation method for dual‐polarization radar observations. Relative humidity pseudo‐observations (squares) were retrieved from attenuation‐corrected reflectivities (Z experiment) or from reflectivities with no attenuation correction (ZNocorr) or from Kdp (ZK experiment), and were assimilated in the AROME 3D‐Var system. Thanks to attenuation‐corrected Zhh in Z, or to Kdp in ZK, relative humidity values were increased in the analysis in convective areas (indicated by 35 dBZ contours in black) compared to ZNocorr or to CTRL (which does not assimilate relative humidity pseudo‐observations). This is illustrated here in the relative humidity analysis fields (around 2700 m) at 06 UTC 24/09/2012. Circles denote 50 and 100 km radius from Nîmes radar. [ABSTRACT FROM AUTHOR]

The occurrence of noise and spurious convective cells in analyses of convective-scale data assimilation (DA) is a problem that is not quantified sufficiently to develop reliable set-ups for the DA of radar observations in an ensemble Kalman filter. This article presents two approaches to quantifying these phenomena using a local ensemble transform Kalman filter testbed, where simulated radar observations are assimilated with varying spatial and temporal DA parameters. Firstly, the dynamical coupling of cold pools and emerging cells of organized deep convection is quantified in terms of spatio-temporal correlations, so that spurious perturbations can be identified. Secondly, the abundance of gravity-wave noise is characterized by the variance of the vertical velocity field during DA cycling and the ensemble forecasts. This evaluation is performed separately for subsets of model points inside observed storms, in their vicinity and in the surrounding clear-air regions. Varying horizontal localization lengths, observation averaging scales and the length of the assimilation cycling interval are compared, with respect to the two diagnostic approaches. It is concluded that the cold-pool coupling method and the partitioned variance of vertical velocity fields are viable measures to quantify the influence of DA parameters and algorithms on the balance of analysis states that are used to initiate forecasts of convection. [ABSTRACT FROM AUTHOR]

This article describes a forward observation operator for polarimetric radar variables, developed within the research mesoscale non-hydrostatic model Meso-NH. The forward operator enables direct comparisons of atmospheric simulations with polarimetric radar observations and constitutes a first step towards the assimilation of polarimetric radar data. The parameters of the polarimetric forward operator were defined in order to be consistent (whenever possible) with the model microphysical parametrization. Subjective and quantitative comparisons between observations and simulations are analyzed for S-, C- and X-band radars for two convective cases that were observed in autumn 2012 in southeastern France during the first observing period (SOP1) of the Hydrological cycle in the Mediterranean Experiment (HyMeX) campaign. The good consistency between the medians of the observed and simulated distributions of Zdr and Kdp as a function of Zhh, in rain and in dry snow, enables the validation of the parametrization of the polarimetric forward operator. However, the higher spread in observed Zdr and Kdp (due to noise and microphysical variability) reveals a need for careful data quality control and pre-processing for use in assimilation. Comparisons between observed and simulated vertical profiles of Zhh, Kdp and Zdr in convective areas also show generally good agreement. Some differences between observations and simulations are attributed to model limitations, in particular its one-moment microphysics scheme and coarser resolution. [ABSTRACT FROM AUTHOR]

Radial velocity measurements collected by the airborne radar RASTA during Hydrological Cycle in the Mediterranean Experiment ( HyMeX) Special Observation Period ( SOP) 1 are used to assess the quality of real-time multiple-Doppler winds produced in southern France from six ground-based radars of the French operational radar network ARAMIS. An original verification procedure based upon the exploitation of RASTA radial velocities is proposed and applied to 10 h of data collected during HyMeX Intensive Observing Periods ( IOPs) 6, 8 and 15. The statistical evaluation of multiple-Doppler winds produced in a variety of weather situations shows that errors in ARAMIS-derived horizontal wind components are generally comprised between 0.5 and 2 m s−1 through most of the troposphere. Secondly, multiple-Doppler winds are used to evaluate the performance of three-dimensional (3 D) wind retrieval inferred from the analysis of RASTA radial velocities collected during HyMeX IOP 8. The mean difference between the two datasets is close to zero for all three wind components suggesting that errors on multiple-Doppler and RASTA-derived 3D cloud winds should be very similar. [ABSTRACT FROM AUTHOR]

An ensemble data assimilation system for 3D radar reflectivity data is introduced for the convection-permitting numerical weather prediction model of the COnsortium for Small-scale MOdelling (COSMO) based on the Kilometre-scale ENsemble Data Assimilation system (KENDA), developed by Deutscher Wetterdienst and its partners. KENDA provides a state-of-the-art ensemble data assimilation system on the convective scale for operational data assimilation and forecasting based on the Local Ensemble Transform Kalman Filter (LETKF). In this study, the Efficient Modular VOlume RADar Operator is applied for the assimilation of radar reflectivity data to improve short-term predictions of precipitation. Both deterministic and ensemble forecasts have been carried out. A case-study shows that the assimilation of 3D radar reflectivity data clearly improves precipitation location in the analysis and significantly improves forecasts for lead times up to 4 h, as quantified by the Brier Score and the Continuous Ranked Probability Score. The influence of different update rates on the noise in terms of surface pressure tendencies and on the forecast quality in general is investigated. The results suggest that, while high update rates produce better analyses, forecasts with lead times of above 1 h benefit from less frequent updates. For a period of seven consecutive days, assimilation of radar reflectivity based on the LETKF is compared to that of DWD's current operational radar assimilation scheme based on latent heat nudging (LHN). It is found that the LETKF competes with LHN, although it is still in an experimental phase. [ABSTRACT FROM AUTHOR]

The Met Office has developed and demonstrated an hourly cycling 1.5 km resolution NWP-based nowcast system (0-6 h forecasts) using four-dimensional variational data assimilation (4D-Var). This was known as the Nowcasting Demonstration Project ( NDP), and was principally for prediction of convective storms for flood forecasting. The NDP was run in real time from March 2012 to April 2013 to cover the period of the London Olympics 2012. The system was run on a domain covering southern England and Wales nested in the UK variable resolution model ( UKV). The UKV used a UK-wide 1.5 km domain with 3 hourly cycling three-dimensional variational data assimilation (3D-Var) and produced 36 h forecasts every 6 h. The NDP 4D-Var included standard observations, Doppler radar radial winds, humidity derived from a 3D cloud cover analysis and geostationary satellite upper-tropospheric water vapour radiances not contaminated by cloud. This was used in combination with latent heat nudging of radar-derived precipitation rates. Example case-studies compare the NDP precipitation forecasts to both the operational extrapolation/merged nowcast system and the UKV forecasts. Objective comparison of fraction skill score for the period June to August 2012 shows that the NDP skill was greater than the latest UKV forecasts, available to forecasters at the same time as the NDP, for the whole 6 h forecast period. The skill of the NDP was greater than the operational extrapolation/merged nowcast beyond T + 2 h. [ABSTRACT FROM AUTHOR]

Networks of rain-gauges provide an accurate but highly localized measure of rainfall, with limited coverage and resolution, whereas radars provide rain-rate and accumulation estimates over wide areas at high spatial and temporal resolution but low accuracy. When quantifying rainfall accumulations for applications such as flood forecasting, combining the two sets of data can be beneficial for producing a high-resolution merged product with a lower error than the gauge-only or the radar-only product. In this study, three kriging methods (kriging with external drift (KED), kriging with radar-based error correction (KRE) and ordinary kriging (OK)) as well as a multiquadric (MQ) scheme have been used to merge radar and gauge data. The results were cross-validated with true rainfall readings at the surface for a number of different meteorological events covering England and Wales. Overall, all the merging schemes trialled produced a merged product that was superior to the individual radar or gauge data. The KED was the best performing method across all rainfall thresholds and meteorological conditions, with the use of a parametric variogram best suited to short (15 min) accumulation periods and a non-parametric variogram preferable for hourly accumulations. The study also shows that the merged product deteriorates when the gauge-network density is reduced; exceptions to this are made in situations where spatially isolated rainfall is observed by the radar and where the gauges that are found least likely to represent the rainfall are carefully removed prior to merging. The latter process was found to improve the quality of the merged product, regardless of the meteorological conditions, providing a viable method for producing a KED-based merging scheme that is applicable to all meteorological conditions. [ABSTRACT FROM AUTHOR]

Rain intensity spectra as seen by space sensors feed numerous applications at global scales ranging from water budget studies to forecasting natural hazards related to extreme rainfall events. Rainfall variability at scales finer than what is resolved by current space sensors affects their detection capabilities, the characterization of rainfall types, as well as the quantification of rainfall rates. A high-resolution surface rainfall product is used to evaluate the impact of rainfall variability within the field of view (FOV) of the Tropical Rainfall Measurement Mission (TRMM) Precipitation Radar (PR) quantitative precipitation estimation (QPE) at ground. The primary contribution of this study is to assess the impact of rainfall variability in terms of occurrence, types and rate at PR's pixel resolution on PR precipitation detection, classification and quantification. Several aspects of PR errors are revealed and quantified including sensitivity to non-uniform beam filling. While the error structure of the PR is complicated because of the interaction of these factors, simple error models are developed to describe the PR performances. The methodology and framework developed herein applies more generally to rainfall rate estimates from other sensors on board low Earth-orbiting satellites such as microwave imagers and dual-frequency radars such as with the Global Precipitation Measurement (GPM) mission. [ABSTRACT FROM AUTHOR]

This article describes the development and trials at the Met Office of the assimilation of radar Doppler radial winds in a pre-operational framework. It uses a high-resolution 1.5 km convective-scale version of the Unified Model for nowcasting, on a small domain covering southern England and Wales. The assimilation was performed using a version of the Met Office 3D-variational data assimilation system at 1.5 km resolution. The Doppler radial wind observations were provided by the original four Dopplerized C-band single-polarization radars in the UK radar network, all situated in the south of England. This article provides a description of the Doppler radial wind observation operator, pre-processing, observation errors and data assimilation configuration. The system is tested on and verified for four case studies using continuous hourly cycling with 7 h forecasts. A significant positive impact for rain and wind is found in the first half of the forecasts (nowcasting time-scale). [ABSTRACT FROM AUTHOR]

Dual-polarisation radar measurements provide valuable information about the shapes and orientations of atmospheric ice particles. For quantitative interpretation of these data in the Rayleigh regime, common practice is to approximate the true ice crystal shape with that of a spheroid. Calculations using the discrete dipole approximation for a wide range of crystal aspect ratios demonstrate that approximating hexagonal plates as spheroids leads to significant errors in the predicted differential reflectivity, by as much as 1.5 dB. An empirical modification of the shape factors in Gans's spheroid theory was made using the numerical data. The resulting simple expressions, like Gans's theory, can be applied to crystals in any desired orientation, illuminated by an arbitrarily polarised wave, but are much more accurate for hexagonal particles. Calculations of the scattering from more complex branched and dendritic crystals indicate that these may be accurately modelled using the new expression, but with a reduced permittivity dependent on the volume of ice relative to an enclosing hexagonal prism. [ABSTRACT FROM AUTHOR]

Observations have been obtained within an intense (precipitation rates > 50 mm h−1) narrow cold-frontal rainband ( NCFR) embedded within a broader region of stratiform precipitation. In situ data were obtained from an aircraft which flew near a steerable dual-polarisation Doppler radar. The observations were obtained to characterise the microphysical properties of cold frontal clouds, with an emphasis on ice and precipitation formation and development. Primary ice nucleation near cloud top (−55°C) appeared to be enhanced by convective features. However, ice multiplication led to the largest ice particle number concentrations being observed at relatively high temperatures (> −10°C). The multiplication process (most likely rime splintering) occurs when stratiform precipitation interacts with supercooled water generated in the NCFR. Graupel was notably absent in the data obtained. Ice multiplication processes are known to have a strong impact in glaciating isolated convective clouds, but have rarely been studied within larger organised convective systems such as NCFRs. Secondary ice particles will impact on precipitation formation and cloud dynamics due to their relatively small size and high number density. Further modelling studies are required to quantify the effects of rime splintering on precipitation and dynamics in frontal rainbands. Available parametrizations used to diagnose the particle size distributions do not account for the influence of ice multiplication. This deficiency in parametrizations is likely to be important in some cases for modelling the evolution of cloud systems and the precipitation formation. Ice multiplication has significant impact on artefact removal from in situ particle imaging probes. [ABSTRACT FROM AUTHOR]