The Simulation Assessment And Prediction Method Research On The Major Modes Of The Asian-Australian Monsoon Interannual Variability

Based on the NCEP/NCAR reanalysis data set and NOAA extended reconstructed sea surface temperature and precipitation reconstruction over land,26 climate system models which participate in the Coupled Model Intercomparison Project Phase 5 are assessed on the simulation capability of the main modes for Asian-Australian monsoon interannual variability. Then the predicition capacity and predictability of the second generation climate system model of Beijing Climate Center,BCC-CSM1.1(m) on the main modes of Asian- Australian monsoon interannual variability is studied. The results are as follows:(1) The simulation result of the first mode is generally better than that of the second mode. There are 22 models which can simulate the high correlation between the first mode and ENSO and 12 models which can simulate the high correlation between the second mode and 1-year lag ENSO. On the synthesize of pattern, time series, and the relationship between ENSO and the main modes, the best four models are:BCC-CSM1.1(m)、GFDL-CM3、GISS-E2-R and MIROC5. However,the deviation exists in the period simulation of these models. The simulation results of the first mode is affected by the simulation consequences of the ENSO pattern, but that of the second mode is influenced by the simulation of the precipitation climatology of the north part of the tropical western pacific warm pool. Multi-models ensemble can simulate the three features of the wind field for the first mode and the ENSO signal of the second mode successfully, but deviation still exists in point of details.(2) The data set of BCC-CSM1.1(m) hindcasts in June can predict the relationship between the main modes and ENSO well, but it can not predict the ENSO phase transformation accurately. The prediction wind amplitude of the first mode and the later three seasons of the second mode is less than observation.BCC-CSM1.1(m) can predict the first feature of the wind anomaly for the first mode, but it has some deviation in predicting the second feature and the third feature of that as well as the time of maritime continent being wet from dry. BCC-CSM1.1(m) can barely predict the pattern of the second mode and the ENSO signal. However, the performance of predicting the time series of the first and the second modes is satisfying. The correlation coefficient of the first and the second mode’s time series between observation and the prediction after projection on the observation pattern are respectively 0.86 and 0.48.(3) The prediction capacity of BCC-CSM1.1(m) of the different lead months on the first mode is better than that of the second mode. Through summarizing the prediction performance of the pattern, the time series and the relationship between it and ENSO, the predictability of BCC-CSM1.1(m) on the first mode can continue to 4-months lead, but that on the second mode can only continue to 0-month lead.