The Capability of ENSEMBLES Models in Predicting the Principal Modes of Pan-Asian Monsoon Precipitation

Abstract

he predictability of the dominant modes of summer (June?September) precipitation in the pan-Asian monsoon region is evaluated based on 1-month-lead retrospective forecasts in five state-of-the-art coupled models from the ENSEMBLES project for the period 1979?2005. The results show that the models and their multimodel ensemble mean (MME) perform well in reproducing the interannual variability of the climatology and the spatiotemporal distribution of the first mode of summer precipitation in the pan-Asian monsoon region. The associated oceanic and atmospheric circulation indicators are also well captured, such as the spatiotemporal structures of the simultaneous El Niño?Southern Oscillation (ENSO) and Antarctic Oscillation in the Pacific Ocean (AAOSP). Moreover, the interannual variation of the preceding AAOSP can also be captured by some of the coupled models. For individual models, the ECMWF, Météo-France, and Met Office models exhibit better skill with respect to the first mode of summer precipitation in the pan-Asian monsoon region, which displays a tripole pattern from north to south over 80°?140°E. In addition, these models can successfully predict the intensity and location of the associated ENSO, as well as the simultaneous summer AAOSP distributions. By contrast, the prediction capabilities of the Leibniz Institute of Marine Sciences (IFM-GEOMAR) and Euro-Mediterranean Center for Climate Change (CMCC-INGV) models are relatively weaker. Furthermore, the predictions of the second mode of the summer precipitation in the pan-Asian monsoon region are investigated. Some of the ENSEMBLES models show good capability in predicting the spatiotemporal distribution of the second mode, owing to the successful prediction of the atmospheric convection activities over the tropical Indian Ocean.