Probabilistic Seasonal Outlook of the Rainy Season over Southeast Asia by Monitoring its Onset

The SouthEast Asian Summer Rainy Season (SEASuRS) exhibits significant interannual variations, which is a challenge for reliable seasonal prediction. In this study we provide a complementary approach to existing methodologies for the seasonal outlook of SEASuRS based on observational monitoring of the onset of the wet season. The onset and demise date of SEASuRS is objectively diagnosed from the first and the last days of the year when the daily rain rates exceed the annual mean climatological rainfall at the granularity of the rainfall analysis. An ensemble of diagnosed onset/demise dates are generated by perturbing the original timeseries of observed daily rain rates to account for uncertainty at meso to synoptic scales so that rain events unconnected to the annual cycle do not significantly influence the diagnosis of these dates. Our study shows that the interannual variations in the start dates of SEASuRS are significantly associated with variations in the season's length and seasonal rainfall. These local relationships are leveraged to generate probabilistic seasonal forecasts, which reveals that longer and wetter seasons associated with early onset or shorter and drier seasons associated with late onset seasons yield very skillful seasonal outlooks at the granularity of the available rainfall analysis.