OSA1.3

DWD's new Seamless INtegrated FOrecastiNg sYstem (SINFONY) is about to come to life in the next two years,
after 5 years of research and development.
For now, the system focuses on severe convective events in the very short time forecast range from minutes to 12~h.

There are different "optimal" forecast methods for different forecast lead times and different weather phenomena.
Focusing on precipitation and convective events up to some hours ahead, radar extrapolation techniques (Nowcasting)
show good skill up to about 2 h ahead (depending on the situation), while numerical weather prediction (NWP)
outperforms Nowcasting only at later hours. Ensembles of both Nowcasting and NWP help
to assess forecast uncertainties.
"Optimally" combining precipitation forecasts from Nowcasting and NWP
as function of lead time leads to the seamless forecasts of the SINFONY.

Different interdisciplinary teams work closely together in developing
a) radar Nowcasting ensembles for precipitation, reflectivity and convective cell objects,
b) a regional ICON-ensemble model with extensive data assimilation of high-resolution remote sensing data (3D radar volume scans of radial winds and reflectivity, cell objects, Meteosat VIS channels and lightning)
and hourly new rapid update cycle forecasts (SINFONY-RUC-EPS) on the km-scale,
c) optimal combinations of Nowcasting and NWP ensemble forecasts in observation space. Gridded combined precipitation and reflectivity ensembles are targeted towards hydrologic warnings. Combined Nowcasting- and NWP cell object ensembles help evolve DWD's warning process for convective events towards a flexible
"warn-on-objects".
d) systems for common Nowcasting and NWP verification of precipitation, reflectivity and objects.
   In particular the cell object based verification will provide new insights into the representation of deep convective cells in the model. 

For b), new innovative and efficient forward operators for radar volume scans and visible satellite data enable
direct operational assimilation of these data in an LETKF framework.
Advanced model physics (stochastic PBL scheme, 2-moment bulk cloud mircophysics) contribute to an improved forecast of convective clouds.

For c), the SINFONY-RUC-EPS outputs simulated reflectivity volume scan ensembles of the
entire German radar network every 5' online during its forecast runs.
Ensembles of composites and cell object tracks are generated
by the same compositing and cell detection- and tracking methods/software packages which are applied to the observations.

During the last year, all these methods have been further consolidated, and the RUC, along with some of the new nowcasting methods, has been run daily in a continuous test forecast mode.

This presentation will give a short overview on the activities of the last year and selected results of the test forecast mode.
Other presentations from SINFONY team members will give more details about the particular SINFONY components.

How to cite: Blahak, U. and the Team SINFONY: Current status of SINFONY - the combination of Nowcasting and Numerical Weather Prediction on the convective scale at DWD, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-440, https://doi.org/10.5194/ems2022-440, 2022.

The importance of accurate and near real-time radar-based quantitative precipitation estimation (QPE) and derived nowcasting products, which are key to feed hydrological models and enable reliable flash flood predictions, was illustrated again by the disastrous flooding in West Germany after a long-lasting intense stratiform precipitation event on 14 July 2021. We applied three state-of-the-art rainfall algorithms to four operational polarimetric C-band radars of the German Meteorological Service (DWD, Deutscher Wetterdienst): one is based on radar reflectivities Z only, while two hybrid algorithms use specific differential phase K_DP in heavy rain combined with Z or specific attenuation A in light rain, respectively. Since large vertical variability of the precipitation flux was observed below the melting layer during the warm-rain process, all QPE products showed significant underestimation. To mitigate this impact on the accuracy of QPE, two approaches have been proposed in this work: i) the inclusion of a local X-band radar, JUXPOL, with lower-altitude observations, and ii) a vertical profile correction for Z and K_DP using so-called range-defined quasi-vertical profiles (RD-QVP) to quantify vertical changes of these variables. When evaluated with DWD rain gauge measurements, JUXPOL and vertical profile correction have considerably improved the accuracy of the rainfall estimates by reducing at least 8% and 30% of normalized mean bias, respectively. QPEs with vertical profile correction even beat DWD’s operational rainfall product, which is based on Z only but hourly adjusted to rain gauge measurements. This greatly increases the value of radar-based QPE algorithms for warm-rain events and potential flood alerts.

How to cite: Chen, J.-Y., Reinoso-Rondinel, R., Trömel, S., Ryzhkov, A., and Simmer, C.: A radar-based quantitative precipitation estimation algorithm to overcome the impact of vertical gradients of warm-rain precipitation: the flood in western Germany on 14 July 2021, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-221, https://doi.org/10.5194/ems2022-221, 2022.

Radar polarimetry and recently developed microphysical retrievals offer great potential for evaluating and improving the parameterizations of numerical models. In this study, we intend to inform the subgrid parameterizations of the ICON general circulation model (ICON-GCM), in particular to assess and improve the spatial heterogeneity of ice water content at ICON subgrid scales. Quasi-vertical profiles (QVPs) generated by azimuthal averaging of various polarimetric radar variables from Plan Position Indicators (PPIs) acquired during standard conical scans at antenna elevation angles of 18° successfully reduce statistical errors, especially for phase-based measurements such as specific differential phase (KDP). This methodology provides the ideal data basis for various robust polarimetric microphysical retrievals of IWC, Nt, and Dm. However, the use of QVPs reduces the information on sub-grid scale variability compared to higher-resolution PPIs. Thus, an important question is how much averaging is required for robust estimates of IWC, Nt, and Dm and how we can separate spatial variability from noise. Spatial variabilities or azimuthal standard deviations and statistical errors of different retrievals are analyzed using measurements of the polarimetric X-band radar in Bonn, Germany (BoXPol). Statistical errors are quantified by the standard error of the mean (σ_mean), calculated via Gaussian error propagation using Bienaymé's law. σ_mean is used to investigate the extent to which the azimuthal window size of 360° in the QVP methodology can be reduced while still ensuring acceptable statistical errors of IWC, Dm, and Nt. Range-defined QVPs (RD-QVPs), a variant making use of different antenna elevation angles, can further reduce the statistical errors due to the larger sample size. Shannon's information entropy is exploited within a new method to test the distributions of polarimetric variables and retrievals for homogeneity within PPIs. This is key for the use of σ_mean to ensure that samples are drawn from the same distribution. Statistics of a large BoXPol data set are presented and compared to simulations of ICON-GCM. Finally, an attempt is made to improve the specific threshold used in ICON-GCM for the onset of aggregation (particle diameters > 0.1 mm are defined as snow) by estimating the particle size distributions (PSDs) assuming an exponential function.

How to cite: Scharbach, T. and Trömel, S.: Radar-derived variability of ice water content (IWC), total number concentration (Nt), and mean volume diameter (Dm) for improved parameterizations, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-437, https://doi.org/10.5194/ems2022-437, 2022.

In recent years, a new nowcasting algorithm has been developed at DWD (Deutscher Wetterdienst), called KONRAD3D. It aims to automatically detect, track, and nowcast convective cells in order to support DWD’s warning decisions. The deterministic core of KONRAD3D consists of state of the art techniques, e.g. adaptive thresholding for the cell detection and Kalman filtering of cell centroids and velocities during the tracking step.

Originally, KONRAD3D made use of three-dimensional radar reflectivity data only. Currently, work is in progress to include lightning data and information on hydrometeor types that is based on polarimetric radar data. In this context, we will introduce a new polarimetric hail flag—a parameter that assesses a cell’s threat of hail—that rests upon the hydrometeor data and should roughly estimate the expectable near-ground hail size.

Studies about relationships of lightning and the hail amount of KONRAD3D cells showed strong correlations. A lightning jump detection within KONRAD3D turned out to be a promising approach for hail nowcasting. Statistics on 800 hailstorms over Germany between April and September 2019 revealed that lightning jumps occur 15 to 20 minutes before the maximum near-ground hail intensity on average.

One part of DWD’s project SINFONY (Seamless INtegrated FOrecastiNg sYstem) focuses on extending KONRAD3D towards an object-based ensemble nowcasting algorithm called KONRAD3D-EPS. It enables a suitable way of including cell life-cycle models in order to better predict intensification and weakening tendencies..

We give an overview of our deterministic cell detection and tracking algorithm KONRAD3D and present statistics of cell attributes. Moreover, we demonstrate the concept of our probabilistic nowcasting system. We also illustrate the basic functionalities of our algorithms for prominent example cases with focus on hail threat assessment.

How to cite: Josipovic, L., Werner, M., Feger, R., Wapler, K., Schultze, M., and Blahak, U.: Object-based Nowcasting at DWD using KONRAD3D, HYMEC, and Lightning Data, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-659, https://doi.org/10.5194/ems2022-659, 2022.

Tropospheric products from Global Navigation Satellite Systems (GNSS) have become a vital data source for Numerical Weather Prediction (NWP) and meteorology in general over the last decades. Commonly derived parameters like path delays and integrated water vapor are utilized for data assimilation and have proven to be beneficial for precipitation forecasts. In recent years, also other meteorological phenomena and parameters, such as soil moisture variations, snow properties or foehn winds, have been studied by means of GNSS remote sensing. However, typical troposphere products only consider the gaseous constituents of water to influence GNSS observations. Therefore, GNSS troposphere products neglect the effect of hydrometeors (liquid and solid particles of water) present in the air. Typically, their contribution to tropospheric delays is small, but becomes significant for severe weather events.

This study investigates the signature of hydrometeors, in particular hail, in time series of GNSS data, aiming to detect precursors of hail formation in the vicinity of GNSS stations. Hail represents one of the most treacherous types of severe weather and, at the same time, can only be detected from radar images. As radar observing systems are cost-intensive and therefore very sparsely distributed, the possibility of hail detection from GNSS observations would be very beneficial, especially for nowcasting applications.

The study presents an analysis of different hail events, e.g. the severe thunderstorm happening over the city of Zurich in the night of 12/13.07.2021, which caused tremendous damages on infrastructure. Therefore, a combination of different GNSS observations (tropospheric delays, signal-to-noise ratio, carrier phase residuals) will be used to investigate potential signs of the hail formation and compared to operational radar observations and event reports. Finally, an outlook on the usability of data-driven methods such as machine-learning-based detection methods for this use case will be given.  

How to cite: Aichinger-Rosenberger, M.: Can we detect hail from GNSS observations? Case studies from severe weather events in Switzerland 2021, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-397, https://doi.org/10.5194/ems2022-397, 2022.

In the framework of the project SINFONY at Deutscher Wetterdienst, we work towards seamless prediction at the very-short range blending over from observation-based nowcasting to numerical weather prediction. The key goals that we pursue in this context are:
1.    To deliver forecasts earlier to be displayed in the meteorological workstation NinJo of our forecasters, which is realized by hourly forecast initialization in our newly-developed Rapid Update Cycle (RUC) and shorter latency for observation arrival ahead of data assimilation. 
2.    To provide seamlessly combined products integrating nowcasted and forecasted radar reflectivities as well as precipitation from both forecasting systems.
3.    To achieve a better representation of precipitation processes and convective cells in our NWP model to allow for the seamless blending with nowcasts. For this purpose, we use a two-moment microphysics scheme that predicts not only mixing ratios of hydrometeor species, but also their particle size distribution. This is also of great importance for the data assimilation of geostationary all-sky satellite data assimilation, for data assimilation of lightning data and essentially radar reflectivities.

In this presentation, we explain how data assimilation of cloudy visible satellite data can help to improve the accuracy of clouds and precipitation processes in NWP forecasts to assist a seamless blending of nowcasting and NWP in terms of radar reflectivities mentioned in 2) and 3). 

Visible satellite data are directly sensitive to liquid water path, ice water path and specific humidity  which are integral quantities related to precipitation processes. Moreover, cloud positioning can be improved by deleting false alarm clouds and convective cells and introducing missing ones to the forecast. A key advantage is that visible data are particularly sensitive to water clouds, which allows to constrain convective cells already at their state of initiation in the initial conditions of our RUC forecasts.

We elaborate on the basic principles of satellite data assimilation in our ICON-D2-KENDA system making use of an ensemble Kalman filter. Case studies will be shown to demonstrate how data assimilation of all-sky satellite data reduces analysis and forecast error of clouds and precipitation. Finally, we show the impact in our rapid update pre-operational system over longer periods of time. 

How to cite: Bach, L., Deppisch, T., Scheck, L., Schomburg, A., Ulbrich, S., Blahak, U., Schraff, C., Faulwetter, R., Stephan, K., Köpken-Watts, C., and Potthast, R.: How can data assimilation of visible satellite data assist seamless prediction?, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-628, https://doi.org/10.5194/ems2022-628, 2022.

The study of radar data assimilation (reflectivity, radial wind and cell object) in NWP models and its effect particularly on short-term forecast has been intensified recently at DWD. In particular, the seamless Integrated ForecastiNg sYstem (SINFONY) project, which develops a short-term forecasting system with a focus on convective events from minutes up to 12 hours ahead, shows clearly the benefit of radar data assimilation in improving the short-term forecast. This system integrates Nowcasting techniques for radar data with numerical weather prediction (NWP based on the new ICON-model) in a seamless way with initial focus on severe summertime convective events and associated hazards such as heavy precipitation, hail and wind gusts.

Besides, radar data assimilation is being operationally used in the short-range ensemble numerical weather prediction (SRNWP) system (ICON-D2-KENDA LETKF system) at DWD since 2020 (radial wind starting in March 2020 and reflectivity starting in June 2020). This is in addition to the traditional Latent Heat Nudging (LHN) of 2D radar-derived precipitation rates. For both systems, SRNWP and SINFONY, the usage of 3D radar data is not only advantageous but crucial to improve the forecast skills related to convection and precipitation.

We will present the latest results of our research in radar assimilation at DWD including the application of radar data assimilation together with a more sophisticated cloud microphysiscs parameterization (a 2-moment bulk scheme) and in combination with the LHN in the SINFONY forecasting system. We also study to assimilate radar information in the alternative form of convective cell objects. Of particular interest are for example the specification of the radar observation error, but also other topics related to the improvement of short-term forecasts.

How to cite: Khosravian, K., Stephan, K., Blahak, U., Neef, L., Vobig, K., De Lozar, A., Potthast, R., Schraff, C., and Welzbacher, C. A.: Assimilation of 3D radar information in convective scales at Deutscher Wetterdienst (DWD), EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-636, https://doi.org/10.5194/ems2022-636, 2022.

A primary goal of the upcoming Rapid Update Cycle (RUC) at the German weather service is to close the gap between nowcasts (NWC) and numerical weatherp prediction (NWP) by adding cloud- and precipitation-related observations to the operational data assimilation. However, the NWP and NWC worlds differ not just by timescale but also more fundamentally in their approach:  while NWCs deal with individual convective cells, i.e. coherent objects whose positions and physical features are tracked, NWP systems and their associated data assimilation deal with gridded information, i.e. pixels of data.

To bridge these two worlds, we have developed a unique aproach of assimilating nowcast objects into an NWP model. The crux of the idea is to identify objects first, and then map the individual physical features of each object onto a regular model grid. In this talk we explore two ways of implementing this idea: The first defines objects simply by whether or not the observed radar reflectivity exceeds a given threshold, and then assimilates the gridded fraction of gridpoints that meet this criterion within a given spatial scale. The second approach defines objects using a more complex cell identification and tracking algorithm, and then grids the associated cell attributes (e.g. cell area) based on the distance of each gridpoint from the object centroid. We then go on to show how both of these approaches allow us to assimilate object-based information into an ensemble filter, focusing in particular on the difficulties of such an unconventional observation operator, as well as the possible complimentarity to conventional radar reflectivity assimilation.

How to cite: Neef, L., Blahak, U., Welzbacher, C., Pante, G., Potthast, R., and Zacharuk, M.: Assimilation of Nowcast Objects in the Regional Forecast Model ICON-LAM, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-697, https://doi.org/10.5194/ems2022-697, 2022.

Dual-polarization radar (DPR) observations provide additional information about bulk properties of clouds and precipitation such as hydrometeor size, shape, orientation and composition compared to single-polarization radar data. Thus, the use of DPR data for model evaluation and data assimilation has the potential to improve the representation of cloud-precipitation microphysical processes in numerical weather prediction (NWP) models, weather analyses and short-term quantitative precipitation forecasts (QPFs). Two frequently used approaches for radar data assimilation are 1) the use of radar forward operators and 2) the use of radar-estimated microphysical model state variables. Approach 1 is challenging as particle size, shape and orientation distributions and the composition of mixed-phase particles, which all impact polarimetric radar observables, are still rather rudimentarily represented in NWP models. Approach 2 circumvents these difficulties but may suffer from uncertainties in the retrievals. Here, we present first results of the latter approach.

Estimates of liquid water content (LWC) and ice water content (IWC) derived from observations of the operational dual-polarimetric C-band radar network of the German national meteorological service (DWD, Deutscher WetterDienst) are assimilated into DWD’s operational convective-scale NWP model ICON-D2 using the KENDA (Kilometre-sale ENsemble Data Assimilation) system. We compare the results of assimilating A) only conventional observations, B) conventional and radar reflectivity observations (approach 1 for single-polarization radar observations), and C) conventional and radar reflectivity observations as in B) and additionally LWC-/IWC-estimates below/above the melting layer. We focus on predicted hourly precipitation accumulations resulting from the three assimilation configurations for an intense three-day stratiform and a two-day convective precipitation period in summer 2017. Configurations B and C, which include radar observations, clearly improve both the deterministic and ensemble first guess QPFs for both precipitation periods compared to configuration A. Configuration C shows better results than configuration B only in some situations. Results also suggest that the assimilation of LWC is superior to the assimilation of IWC, possibly due to larger observation errors above the melting layer and/or the fact that the LWC estimator has been adjusted to the central European climatology. Investigation of the impact of the LWC/IWC-assimilation on QPFs with longer lead times and more events is in progress.

How to cite: Reimann, L., Simmer, C., Potthast, R., and Trömel, S.: On the assimilation of dual-polarization radar observations via estimators of hydrometeor mixing ratios in Germany, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-197, https://doi.org/10.5194/ems2022-197, 2022.

The SINFONY project at Deutscher Wetterdienst (DWD) aims to produce seamless precipitation and radar reflectivity ensemble forecast products for a time-range from minutes up to 12 hours. It combines numerical weather predictions (NWP) and forecasts from Nowcasting (NWC) with an initial focus on severe summertime convective events with associated hazards such as heavy precipitation, hail and wind gusts. Nowcasts at DWD are currently initialized with an update frequency of 5 minutes, while standard short-range numerical weather prediction (SRNWP) systems are initialized every three hours. Hence, in the worst case a three-hour old SRNWP-forecast had to be combined with an up-to-date Nowcast.

To overcome this issue, a rapid update cycle (RUC) is implemented, which initializes forecasts every hour with a potentially much shorter observation cutoff for the subsequent data assimilation step due to time criticality. To avoid growing differences between the atmospheric states in RUC and SRNWP, the RUC data assimilation cycle has a limited lifetime and branches off the SRNWP-cycle every 24 hours.

Besides the initialization frequency of the forecasts, the RUC also differs from the standard SRNWP in the data assimilation and the atmospheric model . The prediction of extreme convective events benefits for example from additionally available observation systems with huge data amount (e.g. satellites). Also, a more sophisticated microphysics scheme is applied differing from the SRNWP leading to a spin-up phase after branching off. However, this microphysics scheme helps to improve the results in reflectivity space compared to the SRNWP making a combination with NWC more feasible.

The large amount of involved additional observational data and more frequent forecasts is a challenge regarding stable data production, timeliness and handling at DWD.

We present ideas in terms of atmospheric parameters, related data flow, involved infrastructure, and results for the convective season of 2022. The latter topic will also touch upon outcomes of a first evaluation of the SINFONY-RUC with the DWD forecasters.

How to cite: Ulbrich, S., Welzbacher, C., Hanisch, T., Potthast, R., and Blahak, U.: Rapid Update Cycle in DWD's Seamless INtegrated FOrecastiNg sYstem (SINFONY), EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-308, https://doi.org/10.5194/ems2022-308, 2022.

IMPROVER (Integrated Model Post-Processing and Verification) has been developed by the Met Office as an open-source probability-based post-processing system to fully exploit our convection permitting, hourly cycling ensemble forecasts. Post-processed MOGREPS-UK model forecasts are blended with deterministic UKV model forecasts and data from the coarser resolution global ensemble, MOGREPS-G, to produce seamless probabilistic forecasts from now out to 7 days ahead. For precipitation, an extrapolation nowcast is also blended in at the start. Forecasts are converted to probabilities at the start, and all initial stages of post-processing are performed on gridded data, with site-specific forecasts extracted as a final step, helping to ensure consistency. Data are processed on a 10km global grid and on a 2km UK-centred grid. An overview of IMPROVER will be given in a separate talk.

The IMPROVER post-processing system includes a precipitation phase discriminator to determine the phase of precipitation after accounting for the change in orography from the modelled grid to the standardised, unsmoothed grid so that forecasts from different resolution models can be blended together seamlessly. The method calculates the melting experienced by precipitation as a product of positive wet-bulb temperature and atmospheric depth. This discriminator includes two empirical thresholds for where snow starts to melt and where melting is complete which are used to separate falling snow, mixed-phase precipitation, and falling rain. Where liquid precipitation falls onto a frozen surface, freezing rain is diagnosed.

This talk describes how the phase discrimination is performed, and how the empirical thresholds have been tuned by comparing spot forecasts with current weather SYNOP observations.

How to cite: Moseley, S.: Tuning the precipitation phase discriminator in IMPROVER, EMS Annual Meeting 2022, Bonn, Germany, 5–9 Sep 2022, EMS2022-257, https://doi.org/10.5194/ems2022-257, 2022.