Evaluation of multi-source WRF data assimilation for a surface-reaching tropical hail event in Yogyakarta, Indonesia: a case study

Surface-reaching hail in the humid tropics is a rare but hazardous manifestation of deep convection, as hailstones must survive descent through a deep warm layer and high melting-level environment. This study investigates hazardous tropical convection producing surface-reaching hail over Yogyakarta, Indonesia, on 11 March 2025, using surface hail reports, Semarang radar observations, Himawari-9 infrared brightness temperature, sounding and ERA5 diagnostics, and WRF data assimilation experiments using WRFDA-3DVAR. A synthetic reflectivity-based hail-supporting proxy was developed from WRF hydrometeor fields, and storm representation was assessed using object-based and spatial metrics, including structure–amplitude–location, distance error, hit rate, intersection-over-union, and fractions skill score. Observations showed rapid convective intensification between 08:00 and 08:20 UTC, closely coinciding with the reported hail occurrence at approximately 08:15 UTC. Himawari-9 indicated cloud-top cooling from approximately − 53°C to − 72°C within 20 min, while environmental diagnostics showed moderate-to-strong instability, weak inhibition, a high freezing level, and organized low-level convergence. Among the assimilation experiments, radar reflectivity assimilation produced the most physically consistent storm-object representation, with the lowest total SAL value and distance error, whereas all-source assimilation increased hit rate but degraded storm-object compactness. These findings demonstrate that surface-reaching tropical hail should be interpreted as an indicator of hazardous convection and that WRF data assimilation experiments, combined with object-based verification, can improve the evaluation of compact hail-supporting storm cores.