Radar Radial Velocity Observations Research Articles

Improving the Analysis and Forecast of Hurricane Dorian (2019) with Simultaneous Assimilation of GOES-16 All-Sky Infrared Brightness Temperatures and Tail Doppler Radar Radial Velocities

AbstractRecent studies have shown that the assimilation of all-sky infrared (IR) observations can be beneficial for tropical cyclone analyses and predictions. The assimilation of Tail Doppler Radar (TDR) radial velocity observations has also been shown to improve tropical cyclone analyses and predictions; however, there is a paucity of literature on the impacts of simultaneously assimilating them with all-sky infrared IR brightness temperatures (BTs). This study examines the impacts of assimilating combinations of GOES-16 all-sky IR brightness temperatures, NOAA P-3 TDR radial velocities, and conventional observations from the Global Telecommunications System (GTS) on the analyses and forecasts of Hurricane Dorian (2019). It is shown that including IR and/or TDR observations on top of conventional GTS observations significantly reduces both track and intensity forecast errors. Track errors are reduced the most (25% at lead times greater than 48 h) when TDR and GTS observations are assimilated. In terms of intensity, errors are always lower at lead times greater than 48 h when IR BTs are assimilated. Simultaneously assimilating TDR and IR observations has the potential to further improve the intensity forecast by as much as 37% at a lead time of 48 h to 72 h. The improved intensity forecasts produced by the experiments assimilating all three observation sources are shown to be a result of the competing effects of IR assimilation producing an overly broad area of strong cyclonic circulation and TDR assimilation constraining that circulation to a more realistic size and intensity. Interestingly, the order in which observations are assimilated has non-negligible impacts on the analyses and forecasts of Dorian.

A Study of Extrapolation Nowcasting Based on IVAP-Retrieved Wind

AbstractIn this study, we propose a new way to obtain motion vectors using the integrating velocity–azimuth process (IVAP) method for extrapolation nowcasting. Traditional tracking methods rely on tracking radar echoes of a few time slices. In contrast, the IVAP method does not depend on the past variation of radar echoes; it only needs the radar echo and radial velocity observations at the latest time. To demonstrate it is practical to use IVAP-retrieved winds to extrapolate radar echoes, we carried out nowcasting experiments using the IVAP method, and compared these results with the results using a traditional method, namely, the tracking radar echoes by correlation (TREC) method. Comparison based on a series of large-scale mature rainfall cases showed that the IVAP method has similar accuracy to that of the TREC method. In addition, the IVAP method provides the vertical wind profile that can be used to anticipate storm type and motion deviations.

A Variational Method for Analyzing Vortex Flows in Radar-Scanned Tornadic Mesocyclones. Part II: Tests with Analytically Formulated Vortices

AbstractThe variational method formulated in Part I for analyzing vortex flow (VF), called VF-Var, is tested with simulated radar radial-velocity observations from idealized and pseudo-operational Doppler scans of analytically formulated benchmark vortices with spiral-band structures to resemble VFs in observed tornadic mesocyclones. The idealized Doppler scans are unidirectional in parallel along horizontal grid lines of a coarse-resolution grid, so they measure only the horizontal components of three-dimensional velocities in the analysis domain. The pseudo-operational Doppler scans mimic a scan mode used by operational WSR-88Ds for severe storms. Paired numerical experiments are designed and performed to test the two-step analysis versus single-step analysis formulated in VF-Var. Both analyses perform very well with dual-Doppler scans and reasonably well with single-Doppler scans. Errors in the analyzed velocities from single-Doppler scans are mainly in the unobserved velocity components and only in fractions of the benchmark velocities. When the vortex is upright or slanted in the direction perpendicular to idealized single-Doppler scans, the two-step analysis slightly outperforms the single-step analysis for idealized Doppler scans and pseudo-operational dual-Doppler scans. When the vortex becomes slanted in the direction largely along or against Doppler scans, both analyses become less (more) accurate in analyzing the horizontal (slantwise vertical) velocity, and the single-step analysis outperforms the two-step analysis especially for single-Doppler scans. By considering the projections of analyzed velocity on radar beams in the original Cartesian coordinates, useful insights are gained for understanding why and how the analysis accuracies are affected by vortex slanting.

Ensemble-Based Targeted Observation Method Applied to Radar Radial Velocity Observations on Idealized Supercell Low-Level Rotation Forecasts: A Proof of Concept

Abstract The potential future installation of a multifunction phased-array radar (MPAR) network will provide capabilities of case-specific adaptive scanning. Knowing the impacts adaptive scanning may have on short-term forecasts will influence scanning strategy decision-making in hopes to produce the most optimal ensemble forecast while also benefiting human severe weather warning decision-making. An ensemble-based targeted observation algorithm is applied to an observing system simulation experiment (OSSE) where the impacts of synthetic idealized supercell radial velocity observations are estimated before the observations are “collected” and assimilated. The forecast metric of interest is the low-level rotation forecast metric (0–1-km updraft helicity), a surrogate for tornado prediction. It is found that the ensemble-based targeted observation approach can reasonably estimate the true error variance reduction when an effective method that treats sampling error is applied, the period of model forecast is associated with less degrees of nonlinearity, and the observation information content relative to the background forecast is larger. In some scenarios, a subset of a full-volume scan assimilation produces better forecasts than all observations within the full volume. Assimilating the full-volume scan increases the number of potential spurious correlations arising between the forecast metric and radial velocity observation induced state perturbations, which may degrade the forecast metric accuracy.

Practical Uncertainties in the Limited Predictability of the Record-Breaking Intensification of Hurricane Patricia (2015)

AbstractHurricane Patricia (2015) was a record-breaking tropical cyclone that was difficult to forecast in real time by both operational numerical weather prediction models and operational forecasters. The current study examines the potential for improving intensity prediction for extreme cases like Hurricane Patricia. We find that Patricia’s intensity predictability is potentially limited by both initial conditions, related to the data assimilation, and model errors. First, convection-permitting assimilation of airborne Doppler radar radial velocity observations with an ensemble Kalman filter (EnKF) demonstrates notable intensity forecast improvements over assimilation of conventional observations alone. Second, decreasing the model horizontal grid spacing to 1 km and reducing the surface drag coefficient at high wind speed in the parameterization of the sea surface–atmosphere exchanges is also shown to notably improve intensity forecasts. The practical predictability of Patricia, its peak intensity, rapid intensification, and the underlying dynamics are further investigated through a high-resolution 60-member ensemble initialized with realistic initial condition uncertainties represented by the EnKF posterior analysis perturbations. Most of the ensemble members are able to predict the peak intensity of Patricia, but with greater uncertainty in the timing and rate of intensification; some members fail to reach the ultimate peak intensity before making landfall. Ensemble sensitivity analysis shows that initial differences in the region beyond the radius of maximum wind contributes the most to the differences between ensemble members in Patricia’s intensification. Ensemble members with stronger initial primary and secondary circulations beyond the radius of maximum wind intensify earlier, are able to maintain the intensification process for longer, and thus reach a greater and earlier peak intensity.

Effect of background error tuning on assimilating radar radial velocity observations for the forecast of hurricane tracks and intensities

AbstractThe background error covariance is an important component in data assimilation systems, which largely dominates the error correlations between different analysis variables. This study aimed to assess the impact of the determination of the background error length and variance scale factors on the track and intensity forecast of hurricane systems for radar radial velocity (Vr) data assimilation. To test the sensitivity of the background error length scale and variance scale, multiple data assimilation analyses were performed using a 3D variational data assimilation method with a parameter‐sweeping strategy for the case of Hurricane Ike (2008) by varying the decorrelation length scale (LEN_SCALING) and the variance (VAR_SCALING) of the background error step by step. It is shown that, in radar Vr data assimilation, generally smaller variance scale and length scale factors tend to improve the prediction of the hurricane track. Setting both VAR_SCALING and LEN_SCALING equivalent to 0.4 provides a better track than other combinations. In particular, the impact of the variance scale factor on the track forecasts is larger than that of the length scale factor. There is no direct impact of tuning variance scales and length scales on the forecast intensity for the first few forecast hours. Tuning length scale factors less than 0.4 provides smaller forecast errors of minimum sea level pressure and maximum surface wind with increasing forecast lead time.

Reconstruction of Typhoon Structure Using 3-Dimensional Doppler Radar Radial Velocity Data with the Multigrid Analysis: A Case Study in an Idealized Simulation Context

Extracting multiple-scale observational information is critical for accurately reconstructing the structure of mesoscale circulation systems such as typhoon. The Space and Time Mesoscale Analysis System (STMAS) with multigrid data assimilation developed in Earth System Research Laboratory (ESRL) in National Oceanic and Atmospheric Administration (NOAA) has addressed this issue. Previous studies have shown the capability of STMAS to retrieve multiscale information in 2-dimensional Doppler radar radial velocity observations. This study explores the application of 3-dimensional (3D) Doppler radar radial velocities with STMAS for reconstructing a 3D typhoon structure. As for the first step, here, we use an idealized simulation framework. A two-scale simulated “typhoon” field is constructed and referred to as “truth,” from which randomly distributed conventional wind data and 3D Doppler radar radial wind data are generated. These data are used to reconstruct the synthetic 3D “typhoon” structure by the STMAS and the traditional 3D variational (3D-Var) analysis. The degree by which the “truth” 3D typhoon structure is recovered is an assessment of the impact of the data type or analysis scheme being evaluated. We also examine the effects of weak constraint and strong constraint on STMAS analyses. Results show that while the STMAS is superior to the traditional 3D-Var for reconstructing the 3D typhoon structure, the strong constraint STMAS can produce better analyses on both horizontal and vertical velocities.

Assimilation of Satellite Infrared Radiances and Doppler Radar Observations during a Cool Season Observing System Simulation Experiment

Abstract An observing system simulation experiment is used to examine the impact of assimilating water vapor–sensitive satellite infrared brightness temperatures and Doppler radar reflectivity and radial velocity observations on the analysis accuracy of a cool season extratropical cyclone. Assimilation experiments are performed for four different combinations of satellite, radar, and conventional observations using an ensemble Kalman filter assimilation system. Comparison with the high-resolution “truth” simulation indicates that the joint assimilation of satellite and radar observations reduces errors in cloud properties compared to the case in which only conventional observations are assimilated. The satellite observations provide the most impact in the mid- to upper troposphere, whereas the radar data also improve the cloud analysis near the surface and aloft as a result of their greater vertical resolution and larger overall sample size. Errors in the wind field are also significantly reduced when radar radial velocity observations were assimilated. Overall, assimilating both satellite and radar data creates the most accurate model analysis, which indicates that both observation types provide independent and complimentary information and illustrates the potential for these datasets for improving mesoscale model analyses and ensuing forecasts.

Assimilating Doppler radar radial velocity and reflectivity observations in the weather research and forecasting model by a proper orthogonal‐decomposition‐based ensemble, three‐dimensional variational assimilation method

Doppler radar observations with high spatial and temporal resolution can effectively improve the description of small‐scale structures in the initial condition and enhance the mesoscale and microscale model skills of numerical weather prediction (NWP). In this paper, Doppler radar radial velocity and reflectivity are simultaneously assimilated into a weather research and forecasting (WRF) model by a proper orthogonal‐decomposition‐based ensemble, three‐dimensional variational assimilation method (referred to as PODEn3DVar), which therefore forms the PODEn3DVar‐based radar assimilation system (referred to as WRF‐PODEn3DVar). The main advantages of WRF‐PODEn3DVar over the standard WRF‐3DVar are that (1) the PODEn3DVar provides flow‐dependent covariances through the evolving ensemble of short‐range forecasts, and (2) the PODEn3DVar analysis can be obtained directly without an iterative process, which significantly simplifies the assimilation. Results from real data assimilation experiments with the WRF model show that WRF‐PODEn3DVar simulation yields better rainfall forecasting than radar retrieval, and radar retrieval is better than the standard WRF‐3DVar assimilation, probably because of the flow‐dependence character embedded in the WRF‐PODEn3DVar.

Assimilation of Doppler Weather Radar Radial Velocity and Reflectivity Observations in WRF-3DVAR System for Short-Range Forecasting of Convective Storms

In this paper the impact of Doppler weather radar (DWR) reflectivity and radial velocity observations for the short range forecasting of a tropical storm and associated rainfall event have been examined. Doppler radar observations of a tropical storm case that occurred during 29–30 October 2006 from SHARDWR (13.6° N, 80.2° E) are assimilated in the WRF 3DVAR system. The observation operator for radar reflectivity and radial velocity is included within latest version of WRF 3DVAR system. Keeping all model physics the same, three experiments were conducted at a horizontal resolution of 30 km. In the control experiment (CTRL), NCEP Final Analysis (FNL) interpolated to the model grid was used as the initial condition for 48-h free forecast. In the second experiment (NODWR), 6-h assimilation cycles have been carried out using all conventional (radiosonde and surface data) and non-conventional (satellite) observations from the Global Telecommunication System (GTS). The third experiment (DWR) is the same as the second, except Doppler radar radial velocity and reflectivity observations are also used in the assimilation cycle. Continuous 6-h assimilation cycle employed in the WRF-3DVAR system shows positive impact on the rainfall forecast. Assimilation of DWR data creates several small scale features near the storm centre. Additional sensitivity experiments were conducted to study the individual impact of reflectivity and radial velocity in the assimilation cycle. Radar data assimilation with reflectivity alone produced large analysis response on both thermodynamical and dynamical fields. However, radial velocity assimilation impacted only on dynamical fields. Analysis increments with radar reflectivity and radial velocity produce adjustments in both dynamical and thermodynamical fields. Verification of QPF skill shows that radar data assimilation has a considerable impact on the short range precipitation forecast. Improvement of the QPF skill with radar data assimilation is more clearly seen in the heavy rainfall (for thresholds >7 mm) event than light rainfall (for thresholds of 1 and 3 mm). The spatial pattern of rainfall is well simulated by the DWR experiment and is comparable to TRMM observations.

Applications of a Velocity Dealiasing Scheme to Data from the China New Generation Weather Radar System (CINRAD)

Abstract An improved velocity dealiasing algorithm is developed as an extension of the Next Generation Weather Radar (NEXRAD) dealiasing algorithm. The algorithm described in this paper is evaluated on selected China Next Generation Weather Radar (CINRAD) S-band radar radial velocity observations. This algorithm includes four modules for removing weak signals and determining the starting radial as a prelude to identifying and correcting aliased velocities. The proposed dealiasing algorithm was tested on 14 different weather systems, composed of typhoons, squall lines, and heavy rains. The results show that the algorithm is robust and stable for dealiasing S-band CINRAD radial velocity measurements. The performance levels for the typhoon and heavy rain cases are slightly better than for squall-line cases.

A 3.5‐Dimensional Variational Method for Doppler Radar Data Assimilation and Its Application to Phased‐Array Radar Observations

A 3.5‐dimensional variational method is developed for Doppler radar data assimilation. In this method, incremental analyses are performed in three steps to update the model state upon the background state provided by the model prediction. First, radar radial‐velocity observations from three consecutive volume scans are analyzed on the model grid. The analyzed radial‐velocity fields are then used in step 2 to produce incremental analyses for the vector velocity fields at two time levels between the three volume scans. The analyzed vector velocity fields are used in step 3 to produce incremental analyses for the thermodynamic fields at the central time level accompanied by the adjustments in water vapor and hydrometeor mixing ratios based on radar reflectivity observations. The finite element B‐spline representations and recursive filter are used to reduce the dimension of the analysis space and enhance the computational efficiency. The method is applied to a squall line case observed by the phased‐array radar with rapid volume scans at the National Weather Radar Testbed and is shown to be effective in assimilating the phased‐array radar observations and improve the prediction of the subsequent evolution of the squall line.

Cloud-Resolving Hurricane Initialization and Prediction through Assimilation of Doppler Radar Observations with an Ensemble Kalman Filter

This study explores the assimilation of Doppler radar radial velocity observations for cloud-resolving hurricane analysis, initialization, and prediction with an ensemble Kalman filter (EnKF). The case studied is Hurricane Humberto (2007), the first landfalling hurricane in the United States since the end of the 2005 hurricane season and the most rapidly intensifying near-landfall storm in U.S. history. The storm caused extensive damage along the southeast Texas coast but was poorly predicted by operational models and forecasters. It is found that the EnKF analysis, after assimilating radial velocity observations from three Weather Surveillance Radars-1988 Doppler (WSR-88Ds) along the Gulf coast, closely represents the best-track position and intensity of Humberto. Deterministic forecasts initialized from the EnKF analysis, despite displaying considerable variability with different lead times, are also capable of predicting the rapid formation and intensification of the hurricane. These forecasts are also superior to simulations without radar data assimilation or with a three-dimensional variational scheme assimilating the same radar observations. Moreover, nearly all members from the ensemble forecasts initialized with EnKF analysis perturbations predict rapid formation and intensification of the storm. However, the large ensemble spread of peak intensity, which ranges from a tropical storm to a category 2 hurricane, echoes limited predictability in deterministic forecasts of the storm and the potential of using ensembles for probabilistic forecasts of hurricanes.

Trade-Offs between Measurement Accuracy and Resolutions in Configuring Phased-Array Radar Velocity Scans for Ensemble-Based Storm-Scale Data Assimilation

Abstract Assimilation experiments are carried out with simulated radar radial-velocity observations to examine the impacts of observation accuracy and resolutions on storm-scale wind assimilation with an ensemble square root filter (EnSRF) on a storm-resolving grid (Δx = 2 km). In this EnSRF, the background covariance is estimated from an ensemble of 40 imperfect-model predictions. The observation error includes both measurement error and representativeness error, and the error variance is estimated from the simulated observations against the simulated “truth.” The results show that the analysis is not significantly improved when the measurement error is overly reduced (from 4 to 1 m s−1) and becomes smaller than the representativeness error. The analysis can be improved by properly coarsening the observation resolution (to 2 km in the radial direction) with an increase in measurement accuracy and further improved by properly enhancing the temporal resolution of radar volume scans (from every 5 to 2 or 1 min) with a decrease in measurement accuracy. There can be an optimal balance or trade-off between measurement accuracy and resolutions (in space and time) for configuring radar scans, especially phased-array radar scans, to improve storm-scale radar wind analysis and assimilation.

Assessing 4D-VAR for dynamical mapping of coastal high-frequency radar in San Diego

The problem of dynamically mapping high-frequency (HF) radar radial velocity observations is investigated using a three-dimensional hydrodynamic model of the San Diego coastal region and an adjoint-based assimilation method. The HF radar provides near-real-time radial velocities from three sites covering the region offshore of San Diego Bay. The hydrodynamical model is the Massachusetts Institute of Technology general circulation model (MITgcm) with 1 km horizontal resolution and 40 vertical layers. The domain is centered on Point Loma, extending 117 km offshore and 120 km alongshore. The reference run (before adjustment) is initialized from a single profile of T and S and is forced with wind data from a single shore station and with zero heat and fresh water fluxes. The adjoint of the model is used to adjust initial temperature, salinity, and velocity, hourly temperature, salinity and horizontal velocities at the open boundaries, and hourly surface fluxes of momentum, heat and freshwater so that the model reproduces hourly HF radar radial velocity observations. Results from a small number of experiments suggest that the adjoint method can be successfully used over 10-day windows at coastal model resolution. It produces a dynamically consistent model run that fits HF radar data with errors near the specified uncertainties. In a test of the forecasting capability of the San Diego model after adjustment, the forecast skill was shown to exceed persistence for up to 20 h.

Analysis and Prediction of a Squall Line Observed during IHOP Using Multiple WSR-88D Observations

Abstract This paper presents a case study on the assimilation of observations from multiple Doppler radars of the Next Generation Weather Radar (NEXRAD) network. A squall-line case documented during the International H2O Project (IHOP_2002) is used for the study. Radar radial velocity and reflectivity observations from four NEXRADs are assimilated into a convection-permitting model using a four-dimensional variational data assimilation (4DVAR) scheme. A mesoscale analysis using a supplementary sounding, velocity–azimuth display (VAD) profiles, and surface observations from Meteorological Aerodrome Reports (METAR) are produced and used to provide a background and boundary conditions for the 4DVAR radar data assimilation. Impact of the radar data assimilation is assessed by verifying the skill of the subsequent very short-term (5 h) forecasts. Assimilation and forecasting experiments are conducted to examine the impact of radar data assimilation on the subsequent precipitation forecasts. It is found that the 4DVAR radar data assimilation significantly reduces the model spinup required in the experiments without radar data assimilation, resulting in significantly improved 5-h forecasts. Additional experiments are conducted to study the sensitivity of the precipitation forecasts with respect to 4DVAR cycling configurations. Results from these experiments suggest that the forecasts with three 4DVAR cycles are improved over those with cold start, but the cycling impact seems to diminish with more cycles. The impact of observations from each of the individual radars is also examined by conducting a set of experiments in which data from each radar are alternately excluded. It is found that the accurate analysis of the environmental wind surrounding the convective cells is important in successfully predicting the squall line.

An unconventional approach for assimilating aliased radar radial velocities

An aliasing operator is introduced to mimic the effect of aliasing that causes discontinuities in radial-velocity observations, and to modify the observation term in the costfunction for direct assimilations of aliased radar radial-velocity observations into numerical models. It is found that if the aliasing operator is treated as a part of the observation operator and applied to the analysed radial velocity in a conventional way, then the analysis is not ensured to be aliased (or not aliased) in consistency with the aliased (or not aliased) observation at every observation point. Thus, the analysis-minus-observation term contains a large alias error whenever an inconsistency occurs at an observation point. This causes fine-structure discontinuities in the costfunction. An unconventional approach is thus introduced to apply the aliasing operator to the entire analysis-minus-observation term at each observation point in the observation term of the costfunction. With this approach, the costfunction becomes smooth and concave upwards in the vicinity of the global minimum. The usefulness of this approach for directly assimilating aliased radar radial-velocity observations under certain conditions is demonstrated by illustrative examples.

Using Wavelet Analysis to Detect Tornadoes from Doppler Radar Radial-Velocity Observations

Abstract A wavelet-based algorithm is developed to detect tornadoes from Doppler weather radar radial-velocity observations. Within this algorithm, a relative region-to-region velocity difference (RRVD) is defined based on the scale- and location-dependent wavelet coefficients and this difference represents the relative magnitude of the radial velocity shear between two adjacent regions of different scales. The RRVD fields of an idealized tornado and a realistic tornado from a high-resolution numerical simulation are analyzed first. It is found that the value of RRVD in the tornado region is significantly larger than those at other locations and large values of RRVD exist at more than one scale. This characteristic forms the basis of the new algorithm presented in this work for identifying tornadoes. Different from traditional tornadic vortex signature detection algorithms that typically rely on the velocity difference between adjacent velocity gate pairs at a single spatial scale, the new algorithm examines region-to-region radial wind shears at a number of different spatial scales. Multiscale regional wind shear examination not only can be used to discard a nontornadic vortex signature to reduce the false alert rate of tornado detection but also has the ability of capturing tornadic signatures at various scales for improving the detection and warning. The potential advantage of the current algorithm is demonstrated by applying it to the radar data collected by Oklahoma City, Oklahoma (KTLX), Weather Surveillance Radar-1988 Doppler (WSR-88D) on 8 May 2003 for a central Oklahoma tornado case.

Physical initialization applied in WRF‐Var for assimilation of Doppler radar data

WRF‐Var is further developed with Physical Initialization (PI) to assimilate Doppler radar radial velocity and reflectivity observations. In this updated 3D‐Var, specific humidity and vertical velocity are first derived from reflectivity with PI, and then the model fields of specific humidity are replaced by the modified ones, finally, the estimated vertical velocity is added to the cost‐function of the existing WRF‐Var (version 2.0) as a new observation type, and radial velocity observations are assimilated directly by the method afforded by WRF‐Var. It is tested with a rainfall event that occurred in Hubei province near the Yangtze River on 19 June 2002. Results show that the updated 3D‐Var shows better capability to forecast the precipitation than the WRF‐Var does, and the forecast reflectivity field correlates well with the observations for about 4‐h prediction period.