Early warning of heavy rainfall event using time-lagged ensemble prediction system (case study February 15th 2019)

Abstract

Abstract Convective clouds can be related to the development of intense storms that produce various extreme weather. The development of extreme weather could involve strong nonlinear interactions of many factors in the atmosphere, hence the ability to forecast extreme weather especially heavy rainfall and issued an early warning, becomes very important. BMKG has developed a time-lagged ensemble prediction system by utilizing the initial time difference, which is considered capable of providing data updates more closely to the forecasts final results. This study examines the percentile classification in the ensemble prediction system, to look for an extreme values distribution, then used it as extreme threshold. The extreme threshold was tested in a heavy rain case on February 15th 2019, on D-7, D-3, and D-1 of early warning dissemination. Based on this research, it was found that the use of the 90th and 95th percentile classification method was able to show a signal of extreme events on D-7 and D-3 events with a consistent probability pattern. In the D-1 prediction period, the probability value increases and the average precipitation value exceeds the extreme threshold.