11:00 〜 11:15
[AAS03-07] What determines the timing of MJO propagation into the western Pacific?: Analysis of 14-km mesh large-member ensemble simulation by NICAM
キーワード：マッデン・ジュリアン振動、赤道波、スケール間相互作用、大気海洋相互作用
The Madden-Julian oscillation (MJO) convection over the western Pacific (WP) provides a source of weather disturbances (e.g., tropical cyclone) and climate variabilities (e.g., teleconnection, ENSO). Thus, it is of great interest to understand and predict how and when the MJO propagates into the WP. While past statistical analyses have emphasized the role of moisture advection via MJO-scale circulations in propagating MJO events, several case studies imply that synoptic-scale disturbances and/or large-scale SST conditions can be responsible for MJO propagation. Due to this lack of our understanding about key spatio-temporal variations in the MJO, it remains elusive what determines the timing of MJO development in the WP.
In this study, we aim to reveal the factors that control the timing of MJO propagation into the WP, using about 4000-member ensemble simulation by nonhydrostatic icosahedral atmospheric model (NICAM) at 14-km horizontal mesh. Two MJO events realized in November and December 2018 (denoted as Nov-MJO and Dec-MJO) were targeted for this ensemble simulation. For each of the Nov- and Dec-MJO event, we conducted 45-day 1000-member experiments (100 members for each initial date) initialized at 00UTC from 25 October to 3 November and from 23 November to 2 December, respectively, when the respective organization of MJO convection began in the Indian Ocean.
In both of the MJO ensemble simulations, MJO propagation was well reproduced in ample number of ensemble members. A comparison of the propagation characteristics between Nov-MJO and Dec-MJO simulation indicates that while the timing of Nov-MJO propagation was almost uniquely determined, the timing of Dec-MJO propagation was bifurcated among the ensemble members. Sensitivity experiments showed that whether the timing of MJO propagation is unique or bifurcated depended on the SST distributions; the bifurcation tends to be allowed under the SSTs in December 2018 irrespective of initial atmospheric conditions.
The bifurcation seen in Dec-MJO propagation corresponded to the difference in the moistening time scale over the WP, which was strongly related to whether MRG/TD-like disturbances with plentiful moisture propagating from the east of the date line can develop and efficiently moisten the WP (as in the observation) or not. The MRG/TD-like disturbances were developed by wave accumulation and wave-convection coupling, and these processes were more appropriately simulated when the initial conditions had the equatorial intraseasonal westerlies and plentiful moisture to the east of the date line.
Our results suggest that MJO propagation into the WP is largely controlled by the interaction between intraseasonal fields and synoptic-scale disturbances, rather than MJO-scale dynamics alone, and that SST distributions may determine degrees of freedom in processes supporting MJO propagation.
[This work was supported by MEXT as “Program for Promoting Researches on the Supercomputer Fugaku” (JPMXP1020200305) and used computational resources provided by the RIKEN Center for Computational Science (Project ID: hp200128, hp210166).]
In this study, we aim to reveal the factors that control the timing of MJO propagation into the WP, using about 4000-member ensemble simulation by nonhydrostatic icosahedral atmospheric model (NICAM) at 14-km horizontal mesh. Two MJO events realized in November and December 2018 (denoted as Nov-MJO and Dec-MJO) were targeted for this ensemble simulation. For each of the Nov- and Dec-MJO event, we conducted 45-day 1000-member experiments (100 members for each initial date) initialized at 00UTC from 25 October to 3 November and from 23 November to 2 December, respectively, when the respective organization of MJO convection began in the Indian Ocean.
In both of the MJO ensemble simulations, MJO propagation was well reproduced in ample number of ensemble members. A comparison of the propagation characteristics between Nov-MJO and Dec-MJO simulation indicates that while the timing of Nov-MJO propagation was almost uniquely determined, the timing of Dec-MJO propagation was bifurcated among the ensemble members. Sensitivity experiments showed that whether the timing of MJO propagation is unique or bifurcated depended on the SST distributions; the bifurcation tends to be allowed under the SSTs in December 2018 irrespective of initial atmospheric conditions.
The bifurcation seen in Dec-MJO propagation corresponded to the difference in the moistening time scale over the WP, which was strongly related to whether MRG/TD-like disturbances with plentiful moisture propagating from the east of the date line can develop and efficiently moisten the WP (as in the observation) or not. The MRG/TD-like disturbances were developed by wave accumulation and wave-convection coupling, and these processes were more appropriately simulated when the initial conditions had the equatorial intraseasonal westerlies and plentiful moisture to the east of the date line.
Our results suggest that MJO propagation into the WP is largely controlled by the interaction between intraseasonal fields and synoptic-scale disturbances, rather than MJO-scale dynamics alone, and that SST distributions may determine degrees of freedom in processes supporting MJO propagation.
[This work was supported by MEXT as “Program for Promoting Researches on the Supercomputer Fugaku” (JPMXP1020200305) and used computational resources provided by the RIKEN Center for Computational Science (Project ID: hp200128, hp210166).]