Multi-Diagnostic-based Ensemble (MDE) Probability Forecast for Deep Convective Area with Aviation Convective Index using Korean Meteorological Administration Model

Deep convection can cause serious weather hazards such as convectively-induced turbulence, lightning, icing threats, and downbursts, so that it can make adverse effects on safe and efficient aviation operations. Therefore, short- and mid-term forecast of thunderstorm are essential in planning the aviation operation. This study developed a Multi-Diagnostic-based Ensemble (MDE) probability forecast system using the Korean Integrated Model-based Global Data Assimilation and Prediction System (KIM-GDAPS) of the Korean Meteorological Administration with 13 km horizontal grid spacing. A radar, lightning detection, and GK-2A satellite data were used to validate the forecast performance of thunderstorm indices in the Incheon Flight Information Region (FIR) domain. Before conducting the validation, occurrence of deep convection was defined by comparing all observation data. In this study, we use a component diagnostic of the fuzzy-logic algorithm of Aviation Convective Index (ACI) as an input for the MDE forecast. The ACI developed in the previous study is calculated by combining the Surface-based CAPE (SBCAPE), outgoing long wave radiation, accumulative precipitation with the optimized membership functions in each season to account for the seasonal variability of convection in Korea. In addition, we use some other diagnostics of instability-related indices such as K-Index, Lifted Index, and Severe Weather Threat Index and variables of dynamic ingredients for developing deep moist convection such as low-level vertically integrated moisture flux convergence, most unstable CAPE (MUCAPE), and gradient of virtual potential temperature. All these new predictors were also converted to the potential values by using the membership functions based on each Probability and/or Cumulative Distribution Functions. Finally, the probability of occurrence of deep convection in the FIR region was calculated using the time-lagged members of multi-diagnostics derived from the KIM-GDAPS forecasts. Forecast skills of all individual diagnostics and MDE forecast was statistically verified against observed deep convective area in Korea. Detailed verification results with case studies will be presented in the conference.

Key Words: Multi-Diagnostic Ensemble-based Probability Forecast, Deep Convective Area, Aviation Convective Index

Acknowledgment: This work was funded by the Korea Meteorological Administration Research and Development Program under Grant KMI2022-00410.