Abstract

Many of the current seasonal prediction systems show useful skill in Madden–Julian Oscillation (MJO) prediction but limited skill in forecasting tropical cyclones (TC). Quantifying the influences of MJO on TC in an operational prediction system is therefore useful for prediction purposes. The results in this study show that forecast skill of boreal summer TC genesis potential index (GPI) anomaly by operational seasonal prediction systems in Met Office Hadley Centre (MOHC) and Beijing Climate Center (BCC) are significantly higher than persistence out to four weeks. In the prediction system of MOHC, stronger MJO initial conditions are conducive to better GPI anomaly forecast, and MJO phases 4–7 make for higher GPI anomaly skill at lead 11–20 days. However, the MJO influence on TC forecasting skill is not so significant in BCC's prediction system. Both the prediction systems could reproduce the relationship that there are larger positive GPI anomalies in MJO phases 4–5 and negative ones in MJO phases 8–1 over the South China Sea (SCS) and northwest Pacific (WNP). However, the intensity of MJO effect on GPI is generally underestimated in BCC_CSM1.2, which may be due to the response of inadequate ascending motion and negative outgoing long‐wave radiation accompanied by weak MJO over SCS and WNP. In GloSea5, the absolute vorticity, relative humidity, vertical velocity and wind shear terms are relatively closer to ERA‐Interim during lead 1–10 days. Considering the intrinsic predictability of the MJO of more than 2 weeks, the forecasted MJO phase information has significant implications for TC genesis prediction on intraseasonal time‐scales.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call