Abstract

The usefulness of a distance-dependent reduction of background error covariance estimates in an ensemble Kalman filter is demonstrated. Covariances are reduced by performing an elementwise multiplication of the background error covariance matrix with a correlation function with local support. This reduces noisiness and results in an improved background error covariance estimate, which generates a reduced-error ensemble of model initial conditions.

The benefits of applying the correlation function can be understood in part from examining the characteristics of simple 2 × 2 covariance matrices generated from random sample vectors with known variances and covariance. These show that noisiness in covariance estimates tends to overwhelm the signal when the ensemble size is small and/or the true covariance between the sample elements is small. Since the true covariance of forecast errors is generally related to the distance between grid points, covariance estimates generally have a higher ratio of noise to signal with increasing distance between grid points. This property is also demonstrated using a two-layer hemispheric primitive equation model and comparing covariance estimates generated by small and large ensembles. Covariances from the large ensemble are assumed to be accurate and are used a reference for measuring errors from covariances estimated from a small ensemble.

The benefits of including distance-dependent reduction of covariance estimates are demonstrated with an ensemble Kalman filter data assimilation scheme. The optimal correlation length scale of the filter function depends on ensemble size; larger correlation lengths are preferable for larger ensembles.

The effects of inflating background error covariance estimates are examined as a way of stabilizing the filter. It was found that more inflation was necessary for smaller ensembles than for larger ensembles.

Abstract

Ensemble Kalman filter (EnKF) techniques have been proposed for obtaining atmospheric state estimates on the scale of individual convective storms from radar and other observations, but tests of these methods with observations of real convective storms are still very limited. In the current study, radar observations of the 8 May 2003 Oklahoma City tornadic supercell thunderstorm were assimilated into the National Severe Storms Laboratory (NSSL) Collaborative Model for Multiscale Atmospheric Simulation (NCOMMAS) with an EnKF method. The cloud model employed 1-km horizontal grid spacing, a single-moment bulk precipitation-microphysics scheme, and a base state initialized with sounding data. A 50-member ensemble was produced by randomly perturbing base-state wind profiles and by regularly adding random local perturbations to the horizontal wind, temperature, and water vapor fields in and near observed precipitation.

In a reference experiment, only Doppler-velocity observations were assimilated into the NCOMMAS ensemble. Then, radar-reflectivity observations were assimilated together with Doppler-velocity observations in subsequent experiments. Influences that reflectivity observations have on storm-scale analyses were revealed through parameter-space experiments by varying observation availability, observation errors, ensemble spread, and choices for what model variables were updated when a reflectivity observation was assimilated. All experiments produced realistic storm-scale analyses that compared favorably with independent radar observations. Convective storms in the NCOMMAS ensemble developed more quickly when reflectivity observations and velocity observations were both assimilated rather than only velocity, presumably because the EnKF utilized covariances between reflectivity and unobserved model fields such as cloud water and vertical velocity in efficiently developing realistic storm features.

Recurring spatial patterns in the differences between predicted and observed reflectivity were noted particularly at low levels, downshear of the supercell’s updraft, in the anvil of moderate-to-light precipitation, where reflectivity in the model was typically lower than observed. Bias errors in the predicted rain mixing ratios and/or the size distributions that the bulk scheme associates with these mixing ratios are likely responsible for this reflectivity underprediction. When a reflectivity observation is assimilated, bias errors in the model fields associated with reflectivity (rain, snow, and hail–graupel) can be projected into other model variables through the ensemble covariances. In the current study, temperature analyses in the downshear anvil at low levels, where reflectivity was underpredicted, were very sensitive both to details of the assimilation algorithm and to ensemble spread in temperature. This strong sensitivity suggests low confidence in analyses of low-level cold pools obtained through reflectivity-data assimilation.

Abstract

A hybrid ensemble transform Kalman filter–three-dimensional variational data assimilation (ETKF–3DVAR) system for the Weather Research and Forecasting (WRF) Model is introduced. The system is based on the existing WRF 3DVAR. Unlike WRF 3DVAR, which utilizes a simple, static covariance model to estimate the forecast-error statistics, the hybrid system combines ensemble covariances with the static covariances to estimate the complex, flow-dependent forecast-error statistics. Ensemble covariances are incorporated by using the extended control variable method during the variational minimization. The ensemble perturbations are maintained by the computationally efficient ETKF. As an initial attempt to test and understand the newly developed system, both an observing system simulation experiment under the perfect model assumption (Part I) and the real observation experiment (Part II) were conducted. In these pilot studies, the WRF was run over the North America domain at a coarse grid spacing (200 km) to emphasize synoptic scales, owing to limited computational resources and the large number of experiments conducted. In Part I, simulated radiosonde wind and temperature observations were assimilated. The results demonstrated that the hybrid data assimilation method provided more accurate analyses than the 3DVAR. The horizontal distributions of the errors demonstrated the hybrid analyses had larger improvements over data-sparse regions than over data-dense regions. It was also found that the ETKF ensemble spread in general agreed with the root-mean-square background forecast error for both the first- and second-order measures. Given the coarse resolution, relatively sparse observation network, and perfect model assumption adopted in this part of the study, caution is warranted when extrapolating the results to operational applications.

Abstract

The hybrid ensemble transform Kalman filter–three-dimensional variational data assimilation (ETKF–3DVAR) system developed for the Weather Research and Forecasting (WRF) Model was further tested with real observations, as a follow-up for the observation system simulation experiment (OSSE) conducted in Part I. A domain encompassing North America was considered. Because of limited computational resources and the large number of experiments conducted, the forecasts and analyses employed relatively coarse grid spacing (200 km) to emphasize synoptic scales. As a first effort to explore the new system with real observations, relatively sparse observation datasets consisting of radiosonde wind and temperature during 4 weeks of January 2003 were assimilated. The 12-h forecasts produced by the hybrid analysis produced less root-mean-square error than the 3DVAR. The hybrid improved the forecast more in the western part of the domain than the eastern part. It also produced larger improvements in the upper troposphere. The overall magnitude of the ETKF ensemble spread agreed with the overall magnitude of the background forecast error. For individual variables and layers, the consistency between the spread and the error was less than the OSSE in Part I. Given the coarse resolution and relatively sparse observation network adopted in this study, caution is warranted when extrapolating these results to operational applications. A case study was also performed to further understand a large forecast improvement of the hybrid during the 4-week period. The flow-dependent adjustments produced by the hybrid extended a large distance into the eastern Pacific data-void region. The much improved analysis and forecast by the hybrid in the data void subsequently improved forecasts downstream in the region of verification. Although no moisture observations were assimilated, the hybrid updated the moisture fields flow dependently through cross-variable covariances defined by the ensemble, which improved the forecasts of cyclone development.

Abstract

A nonlocal quasi-phase radio occultation (RO) observation operator is evaluated in the assimilation of Challenging Minisatellite Payload (CHAMP) radio occultation refractivity using a Weather Research and Forecasting (WRF) ensemble data assimilation system at 50-km resolution. The nonlocal operator calculates the quasi phase through integration of the model refractivity along the observed ray paths. As a comparison, a local refractivity operator that calculates the model refractivity at the observed ray perigee points is also evaluated. The assimilation is done over North America during January 2003 in two different situations: in conjunction with dense, high-quality radiosonde observations and with only satellite cloud drift wind observations. Analyses of temperature and water vapor with the RO refractivity assimilated using the local and nonlocal operator are verified against nearby withheld radiosonde observations. The bias and RMS errors of the analyses of water vapor and temperature using the nonlocal operator are significantly reduced compared with those using the local operator in the troposphere when the only additional observations are satellite cloud drift winds. The reduction of the bias and RMS errors is reduced when radiosonde observations are assimilated.

Abstract

One aspect of implementing a limited-area ensemble Kalman filter (EnKF) involves the specification of a suitable ensemble of lateral boundary conditions. Two classes of methods to populate a boundary condition ensemble are proposed. In the first class, the ensemble of boundary conditions is provided by an EnKF on a larger domain and is approximately a random draw from the probability distribution function for the forecast (or analysis) on the limited-area domain boundary. The second class perturbs around a deterministic estimate of the state using assumed spatial and temporal covariance relationships. Methods in the second class are relatively flexible and easy to implement. Experiments that test the utility of these methods are performed for both an idealized low-dimensional model and limited-area simulations using the Weather Research and Forecasting (WRF) model; all experiments employ simulated observations under the perfect model assumption. The performance of the ensemble boundary condition methods is assessed by comparing the results of each experiment against a control “global” EnKF that extends beyond the limited-area domain. For all methods tested, results show that errors for the limited-area EnKF are larger near the lateral boundaries than those from a control EnKF, but decay inside the limited-area domain so that errors there are comparable to the control case. The relatively larger errors near the boundaries in the limited-area EnKF originate from not assimilating observations outside the limited-area domain and, in the second class of methods, from deficiencies in boundary spatial and temporal covariances. Overall, these experiments suggest that for observation densities typical in numerical weather prediction models, ensemble boundary conditions can be specified in the absence of a global ensemble without significant penalty in the domain interior by perturbing around an ensemble mean.

Abstract

A global atmospheric analysis and forecast system is constructed based on the atmospheric component of the Model for Prediction Across Scales (MPAS-A) and the Data Assimilation Research Testbed (DART) ensemble Kalman filter. The system is constructed using the unstructured MPAS-A Voronoi (nominally hexagonal) mesh and thus facilitates multiscale analysis and forecasting without the need for developing new covariance models at different scales. Cycling experiments with the assimilation of real observations show that the global ensemble system is robust and reliable throughout a one-month period for both quasi-uniform and variable-resolution meshes. The variable-mesh assimilation system consistently provides higher-quality analyses than those from the coarse uniform mesh, in addition to the benefits of the higher-resolution forecasts, which leads to substantial improvements in 5-day forecasts. Using the fractions skill score, the spatial scale for skillful precipitation forecasts is evaluated over the high-resolution area of the variable-resolution mesh. Skill decreases more rapidly at smaller scales, but the variable mesh consistently outperforms the coarse uniform mesh in precipitation forecasts at all times and thresholds. Use of incremental analysis updates (IAU) greatly decreases high-frequency noise overall and improves the quality of EnKF analyses, particularly in the tropics. Important aspects of the system design related to the unstructured Voronoi mesh are also investigated, including algorithms for handling the C-grid staggered horizontal velocities.

Abstract

The impact of assimilating radiance observations from the Advanced Microwave Sounding Unit-A (AMSU-A) on forecasts of several tropical cyclones (TCs) was studied using the Weather Research and Forecasting Model (WRF) and a limited-area ensemble Kalman filter (EnKF). Analysis/forecast cycling experiments with and without AMSU-A radiance assimilation were performed over the Atlantic Ocean for the period 11 August–13 September 2008, when five named storms formed. For convenience, the radiance forward operators and bias-correction coefficients, along with the majority of quality-control decisions, were computed by a separate, preexisting variational assimilation system. The bias-correction coefficients were obtained from 3-month offline statistics and fixed during the EnKF analysis cycles. The vertical location of each radiance observation, which is required for covariance localization in the EnKF, was taken to be the level at which the AMSU-A channels’ weighting functions peaked.

Deterministic 72-h WRF forecasts initialized from the ensemble-mean analyses were evaluated with a focus on TC prediction. Assimilating AMSU-A radiances produced better depictions of the environmental fields when compared to reanalyses and dropwindsonde observations. Radiance assimilation also resulted in substantial improvement of TC track and intensity forecasts with track-error reduction up to 16% for forecasts beyond 36 h. Additionally, assimilating both radiances and satellite winds gave markedly more benefit for TC track forecasts than solely assimilating radiances.

Abstract

Approximations to flow-dependent analysis-error covariance singular vectors (AEC SVs) were calculated in a dry, T31 L15 primitive-equation global model. Sets of 400-member ensembles of analyses were generated by an ensemble-based data assimilation system. A sparse network of simulated rawinsonde observations were assimilated, and a perfect model was assumed. Ensembles of 48-h forecasts were also generated from these analyses. The structure of evolved singular vectors was determined by finding the linear combination of the forecast ensemble members that resulted in the largest forecast-error variance, here measured in a total-energy norm north of 20°N latitude. The same linear combination of analyses specifies the initial-time structure that should evolve to the forecast singular vector under assumptions of linearity of error growth.

The structures of these AEC SVs are important because they represent the analysis-error structures associated with the largest forecast errors. If singular vectors using other initial norms have very different structures, this indicates that these structures may be statistically unlikely to occur. The European Centre for Medium-Range Weather Forecasts currently uses singular vectors using an initial total-energy norm [“total-energy singular vectors” or (TE SVs)] to generate perturbations to initialize their ensemble forecasts. Approximate TE SVs were also calculated by drawing an initial random ensemble with perturbations that were white in total energy and applying the same approach as for AEC SVs. Comparing AEC SVs and approximate TE SVs, the AEC SVs had maximum amplitude in midlatitudes near the tropopause, both at the initial and evolved times. The AEC SVs were synoptic in scale, deep, and did not appear to be geographically localized nor tilted dramatically upshear. This contrasts with TE SVs, which started off relatively smaller in scale, were tilted upshear, and had amplitudes typically largest in the lower to midtroposphere.

The difference between AEC SVs and TE SVs suggests that operational ensemble forecasts based on TE SVs could be improved by changing the type of singular vector used to generate initial perturbations. This is particularly true for short-range ensemble forecasts, where the structure of the forecast ensemble is more closely tied to the analysis ensemble.

Abstract

The statistical properties of analysis and forecast errors from commonly used ensemble perturbation methodologies are explored. A quasigeostrophic channel model is used, coupled with a 3D-variational data assimilation scheme. A perfect model is assumed.

Three perturbation methodologies are considered. The breeding and singular-vector (SV) methods approximate the strategies currently used at operational centers in the United States and Europe, respectively. The perturbed observation (PO) methodology approximates a random sample from the analysis probability density function (pdf) and is similar to the method performed at the Canadian Meteorological Centre. Initial conditions for the PO ensemble are analyses from independent, parallel data assimilation cycles. Each assimilation cycle utilizes observations perturbed by random noise whose statistics are consistent with observational error covariances. Each member’s assimilation/forecast cycle is also started from a distinct initial condition.

Relative to breeding and SV, the PO method here produced analyses and forecasts with desirable statistical characteristics. These include consistent rank histogram uniformity for all variables at all lead times, high spread/skill correlations, and calibrated, reduced-error probabilistic forecasts. It achieved these improvements primarily because 1) the ensemble mean of the PO initial conditions was more accurate than the mean of the bred or singular-vector ensembles, which were centered on a less-skilful control initial condition—much of the improvement was lost when PO initial conditions were recentered on the control analysis; and 2) by construction, the perturbed observation ensemble initial conditions permitted realistic variations in spread from day to day, while bred and singular-vector perturbations did not. These results suggest that in the absence of model error, an ensemble of initial conditions performs better when the initialization method is designed to produce random samples from the analysis pdf. The perturbed observation method did this much more satisfactorily than either the breeding or singular-vector methods.

The ability of the perturbed observation ensemble to sample randomly from the analysis pdf also suggests that such an ensemble can provide useful information on forecast covariances and hence improve future data assimilation techniques.