Prediction Skills of Rainfall Diurnal Cycle over the Maritime Continent and the Influences from Madden-Julian Oscillation in ECMWF and NCEP S2S Reforecasts

The Maritime Continent is one of the rainiest regions in the globe. The diurnal cycle of rainfall contributes significantly to rainfall variability over this region and can also be modulated by the Madden-Julian Oscillation (MJO). However, our understanding of the prediction skills for the diurnal cycle and its relevance to the MJO predictions in subseasonal forecast model outputs is still limited. This study aims to evaluate the subseasonal prediction skills of rainfall diurnal cycle characteristics over the Maritime Continent using subseasonal-to-seasonal (S2S) prediction models and explore the influences of MJO activity on the prediction skill of rainfall diurnal cycle. The rainfall diurnal cycle characteristics include the daily mean, diurnal amplitude, and timing of the maxima and minima of the diurnal cycle. The prediction skills of rainfall diurnal cycle characteristics are evaluated against satellite rainfall and radar observational data collected during the “Years of Maritime Continent (YMC)” field campaign in winter 2017/18. The capability of S2S models to forecast the rainfall diurnal cycle during winter 2017/18 will also be compared to other years in hindcast datasets. Finally, we will stratify the forecasted rainfall diurnal cycle by MJO phase to find how the MJO influences the prediction skills of rainfall diurnal cycle variability in S2S models. The hypothesis of this study is that the prediction skills for the diurnal cycle of rainfall would be better when significant MJO events are present at the forecast initialization time compared to when weak MJO events occur. The present study contributes to our understanding of the subseasonal prediction skill of the rainfall diurnal cycle and the role of MJO in rainfall predictions over the Maritime Continent.