资源环境科技发展态势分析平台

A sensitivity-based initial condition (IC) perturbation method for convection-permitting ensemble forecasts (CPEFs) is presented and preliminarily tested. The distribution of the perturbations is based on the sensitivity patterns from the ensemble sensitivity analysis (ESA) and the fast-growing perturbations calculated by the breeding growth mode (BGM). Two convective precipitation cases are used to quantitatively and qualitatively evaluate the impacts of the perturbation schemes on the ensemble forecast skill based on the Weather Research and Forecasting (WRF) model. To generate the sensitivity-based IC perturbations, ESA is applied to the short-term convective precipitation cases to calculate the sensitivity patterns. The analysis reveals the influential factors related to the evolution of weather situations that impact the strength of the forecast precipitation. The sensitivity patterns are introduced to the IC perturbations by a sensitivity-based BGM (SeBGM) method proposed in this research, and the response of the forecast skill to different perturbation magnitudes is examined. Ensemble forecasts with sensitivity-based IC perturbations can adapt to changes in the weather regime and provide accurate simulations of the placement and strength of convective systems, leading to improvements in the precipitation forecast skill.