Intraseasonal Forecasting of Asian Summer Monsoon, MJO, and Tropical Cyclones

Intra-Seasonal Oscillation (ISO) is one of the major climate variability in the tropics with a period of 30-60 days, which offers an opportunity to predict weather statistics from two weeks to two months. During boreal summer, the ISO initiates in western equatorial Indian Ocean and propagates northeastward to modulate tropical cyclones (TC) in Indo-Pacific Ocean as well as active and break spells of Asian summer monsoon. The eastward-propagating MJO component will modulate North American monsoon and Hurricances in the eastern North Pacific, Gulf Mexico, and Atlantic Basin. In a series of numerical experiments, we have demonstrated that interactive air-sea coupling plays an important role to the realistic simulation of the northward-propagating ISO: its intensity and phase relationship between convection and sea surface temperature; air-sea coupling also significantly extends the predictability of the ISO by generating a coherent positive SST anomaly in front of the convection, which helps organizing the convection and leading the propagation of the ISO, thus advancing the intraseasonal predictability of Asian summer monsoon too. In addition to surface conditions (or air-sea coupling), initial conditions also play a significant role on intraseasonal forecasting of Asian summer monsoon, which has also been demonstrated in our recent studies.

During boreal winter, the ISO has a dominant eastward propagation (also known as Madden-Julian Oscillation or MJO), which strongly modulates the tropical cyclones in South Indo-Pacific Ocean and active-break spells of Australian summer monsoon. Through tropical-extratropical tele-connection, the MJO regulates the heavy rainfall and snow storm events over western and eastern coasts of the United States. An inter-comparison study during DYNAMO period has revealed that current operational and research climate models show very good skill in forecasting successive MJO (with a preceding event) rather than the primary MJO (with no preceding event). Major remained issues include: i), the problem to reproduce MJO initiation and propagation and, ii) the difficulty to propagate model MJO over the Maritime Continent.

Improved representation of the ISO makes extended-range TC forecasting possible. Current numerical models have shown capability in forecasting some TC occurrences (e.g., Debby in late June over the Gulf of Mexico and Daniel in early July over the Eastern North Pacific) with a lead time of one-to-three weeks. The potential societal benefits of extended-range TC forecasting are tremendous. At the same time, great efforts are still needed to make routine extended-range forecasting of tropical cyclones more useful for applications. On intraseasonal time scale, current model forecasting of tropical cyclones still has various uncertainties: such as too many false alarms, some missing events and “jumpy” forecasts at different lead times.