Comparison of Time-Lagging and Neighborhood Methods on the Performance of the Rapid Refresh Forecast System

NOAA is transitioning to the Unified Forecast System (UFS) for both global and regional applications. Part of this process will involve replacing operational regional forecast systems with the Finite Volume on a Cubed Sphere (FV3) based convection-allowing Rapid Refresh Forecast System (RRFS), which is under development. The current convection-allowing ensemble model, the High Resolution Ensemble Forecast (HREFv3), is a multi-dynamical core (dycore) and multi-physics ensemble that employs time lagging. However, the RRFS will be a single-dycore and single-physics ensemble. Often, these single-dycore, single-physics ensembles are under dispersive, or lack sufficient spread. This project aims to optimize ensemble design of the RRFS by examining the use of time-lagging, neighborhood methods, initial condition perturbations, and stochastic physics. The goal is to evaluate their potential to provide sufficient forecast spread without degrading skill.

In this presentation, we focus on evaluating the impact of including time-lagged members as well as neighborhood probabilistic methods on model performance. The analyzed data comes from one of the RRFS prototypes run during the Hazardous Weather Testbed (HWT) 2021 Spring Forecasting Experiment (SFE), which was conducted between 4 May and 4 June 2021. Objective verification is performed using probabilistic and ensemble-based metrics on multiple variables and thresholds, including discontinuous fields. This presentation compares the performance of these different ensemble designs using the UFS Short Range Weather App, which runs the enhanced Model Evaluation Tools (METplus) verification package.