Implications of a 30-second Update Cycle for a Convective-Scale Ensemble Radar Data Assimilation System

As we enter the era of post peta-scale computing, convective-scale NWP will be performed at increasingly higher model resolutions, using more sophisticated data assimilation (DA) schemes and advanced observational datasets. Here we explore the implications for a regional-scale numerical weather prediction system that uses a unique 30-second update for a 500-m grid, using observations from an advanced multi-parameter phased array weather radar (MP-PAWR), on forecasts of convective weather systems. Experiments showed a rapid buildup in the level of atmospheric dynamical activity in the analyses from the start of cycling that promoted the initialization of spurious and often overly-strong convection in forecasts. This was found to be the consequence of substantial differences between the initial conditions and observations and the rapid updating process, which together introduced large perturbations to the analyses during early cycling, leading to the generation of an atmospheric state that was characterized by strong low-level winds and regions of high convective instability. These conditions would remain at a near constant level well after the period of initial cycling, continuing to be a strongly determining factor on the level of development of convection in the forecasts. It was subsequently demonstrated that we could reduce the level of convective activity in forecasts and so improve forecast skill by reducing the localization scale parameter to near model grid resolution, which acted to force initial conditions closer to the initial set of observations following the first update and reduce the large pertubations that caused these conditions to develop.