Assimilation of Dual-Pol Quality-Controlled Radial Velocity Data in the NOAA Operational Convective-Scale Forecast System

Weather radar network is the only data source which can routinely provide storm-scale observations of the atmosphere at a very high time frequency. Radar radial velocity observations contain storm-scale wind information and is considered very important for making quality convective-scale storm forecasts. In the past several years, many studies (Hu et al 2006, Hu and Xue 2007, Ge et al 2012, 2013, etc.) have shown that the assimilation of radial wind data improves forecasts of tornadic supercell storms. However, the impact of radial wind in the NOAA operational convective-scale forecast system (i.e. HRRR) is so far not explicitly evaluated.

The assimilation of radar wind data in an operational setting is not trivial as a lot of factors will affect the efficient use of this type of data. The amount of radial wind data per 5 minutes is huge and in-time delivery of these data in operation is challenging. Radar wind data contains noise, artifacts and this calls for careful quality-control procedures. Radar wind observation can only observe velocity up to a limited value which is determined by radar hardware parameters and called nyquist velocity. However, velocity inside storms often exceeds nyquist velocity and hence caused velocity aliasing. Raw radar wind observations often contain aliased data and it is critical to correctly de-aliasing these data before assimilating them into weather models. Radar data is much denser than the model resolution and a good thinning strategy is required to get the best use of these observations.

Our examination of current available operational radar velocity data revealed that due to legacy strict de-aliasing and quality-control algorithms, much of useful raw radar wind observation was tossed away. This greatly limit the impact of radar winds. In collaboration with the National Severe Storm Laboratory, we can obtain less-processed raw radar wind data in real-time since early 2019. These data only go through preliminary quality control (such as clutter removal) based on dual-pol information. These data also open a possibility to improve de-aliasing algorithm using HRRR short-term forecasts which provide a good quality model background. These quality-controlled and de-aliased data will then be assimilated into the HRRR model to examine the impact of radar winds on storm forecasts. The sensitivity to data thinning (or super-obbing) strategy, to data assimilation window and frequency (such as 15 min vs 60 min) , to other data assimilation settings (such as hybrid weighting, HRRRDAS BEC vs GDAS BEC) will also be reported.

Ge, G., J. Gao, and M. Xue, 2013: Impacts of Assimilating Measurements of Different State Variables with a Simulated Supercell Storm and Three-Dimensional Variational Method. Mon. Wea. Rev., 141, 2759–2777, https://doi.org/10.1175/MWR-D-12-00193.1

Ge, G., J. Gao, and M. Xue, 2012: Diagnostic Pressure Equation as a Weak Constraint in a Storm-Scale Three-Dimensional Variational Radar Data Assimilation System. J. Atmos. Oceanic Technol., 29, 1075–1092, https://doi.org/10.1175/JTECH-D-11-00201.1

Hu, M. and M. Xue, 2007: Impact of Configurations of Rapid Intermittent Assimilation of WSR-88D Radar Data for the 8 May 2003 Oklahoma City Tornadic Thunderstorm Case. Mon. Wea. Rev., 135, 507–525, https://doi.org/10.1175/MWR3313.1

Hu, M., M. Xue, J. Gao, and K. Brewster, 2006: 3DVAR and Cloud Analysis with WSR-88D Level-II Data for the Prediction of the Fort Worth, Texas, Tornadic Thunderstorms. Part II: Impact of Radial Velocity Analysis via 3DVAR. Mon. Wea. Rev., 134, 699–721, https://doi.org/10.1175/MWR3093.1