Impacts of Multi–Observation Interests on Rainstorm Forecasts in East China: An OSSE–Based Study

Existing Observation System Simulation Experiment (OSSE) studies have limitations in evaluating regional rainstorms in East China due to the lack of known atmospheric truth and varied interests. This study introduces a comprehensive framework that combines multiple forecast metrics (synoptic, objective, energy, economic) with various observation interests (element, region, density, layer) using identical twin designs. The framework assesses their impacts on a central East China rainstorm event (19 July 2016) through flexible assimilation-forecast systems. Key findings include: (1) significant disparities in mutual sensitivity between metrics and observation interests, highlighting the limitations of single-interest approaches; (2) coordinated optimization strategies that identify specific observation interests (e.g., surface temperature, mid-east region, southern density, 500 hPa layers) that enhance forecast accuracy, while others (e.g., dew element, central-west region, northwestern density, 925 hPa layers) require cautious reduction; (3) universal improvements in forecasts through assimilation across different reference truths, but with notable benefit-cost paradoxes that underscore the need for enhanced validation of synoptic metrics and reference authenticity. These results provide actionable insights for improving the harmonization of observation and forecast systems in regional numerical models, ultimately enhancing extreme weather prediction and disaster mitigation.