Multimodel and Multidiagnostic Ensemble-Based Deep Convective Area Forecast for Aviation Operations Using the Global Unified Model and Korean Integrated Model

To provide safe and efficient guidance for aircraft operation to avoid deep convective areas (DCAs) in Korea and broader East Asia, we developed a multimodel and multidiagnostic ensemble (MMDE)-based forecast system. This system utilizes two global numerical weather prediction (NWP) models based on the Unified Model (UM) and the Korean Integrated Model (KIM) operated by the Korea Meteorological Administration. To predict hazardous weather conditions, we defined DCAs using ground-based radar mosaic data where the 15-dBZ echo-top height exceeds flight level (FL) 250 (about z = 6.5 km) and FL350 (about z = 9 km). For predictors, we employed a total of 22 diagnostics, which were derived either directly from physical parameterization schemes within the NWP models or indirectly from various mesoscale forcings responsible for deep convection. Performance skills of the individual DCA diagnostics from both the UM and KIM models were evaluated against radar-based DCA observations from June to September 2022. Finally, the normalized individual DCA diagnostics from the two NWP models were used as ensemble members for deterministic and probabilistic forecast systems. As a result, the newly developed MMDE-based deterministic and probabilistic DCA forecasts outperformed both individual diagnostics and single-model-based forecasts. Eventually, this newly developed DCA forecasting system is expected to be highly valuable for strategic planning of aviation operations in Korea and East Asia. SIGNIFICANCE STATEMENT: Deep convective areas pose dangerous weather conditions to flight operations. This paper describes a deterministic and probabilistic forecast system using multimodel and multidiagnostic ensemble–based methods, which takes into account possible uncertainties in numerical weather prediction models. The newly developed forecast product is superior to the individual diagnostics and single-model-based methods. We anticipate that the new product will enhance the safety and efficiency of air traffic management in terminal areas and broader airspace across the world.