Convection Forecasts From The Hourly Updated, 3-km High Resolution Rapid Refresh Model

Abstract

The next generation of the 1h Rapid Update Cycle is being readied for operational use at the 13 km scale. As described in a companion paper (Weygandt et al.), an effective technique for assimilating 3-d radar reflectivity data has been developed for the 13km RUC and upcoming 13km Rapid Refresh using a version of the WRF model. Here, we describe a new, hourly updated, high-resolution (3 km) nested model, the HRRR (High-Resolution Rapid Refresh) using the same radar reflectivity assimilation method. The HRRR has considerable promise for short-range severe weather prediction, and a first for hourly updated, storm-scale model forecasts including hourly radar reflectivity assimilation. The HRRR is currently nested over the northeastern US inside the 13km backup RUC run at NOAA/ESRL/GSD, run on a 1h cycle, and uses a version of the WRF model but does not include convective parameterization. It relies on the RUC13 data assimilation, which includes radar reflectivity assimilation based on a digital filter initialization (DFI) technique. Using the forward (diabatic) DFI inside the RUC (and in the future 13km Rapid Refresh) is shown to dramatically improve reflectivity forecasts from the HRRR. We will detail the configuration and environment of the HRRR runs and how the DFI-radar technique in the 13km RUC is applied to the 3km HRRR. The paper will include examples of the HRRR forecasts highlighting the effects of this technique, especially in the all-important first few hours of the HRRR forecasts. This information is critical in the location and intensity of the forecast precipitation and reflectivity. Case studies will also include 12-hour forecasts run hourly during summer 2008 for efficacy in improving guidance for air-traffic management over the busy northeast US air traffic corridor.