JpGU-AGU Joint Meeting 2017

Presentation information

[EE] Oral

A (Atmospheric and Hydrospheric Sciences) » A-CG Complex & General

[A-CG45] [EE] Multi-scale ocean-atmosphere interaction in the tropical Indo-Pacific region

Sat. May 20, 2017 9:00 AM - 10:30 AM 302 (International Conference Hall 3F)

convener:Motoki Nagura(Japan Agency for Marine-Earth Science and Technology), H Annamalai(University of Hawaii at Manoa), Ayako Seiki(Japan Agency for Marine-Earth Science and Technology), Yukiko Imada(Meteorological Research Institute, Japan Meteorological Agency), Chairperson:Motoki Nagura(Japan Agency for Marine-Earth Science and Technology), Chairperson:Ayako Seiki(Japan Agency for Marine-Earth Science and Technology)

9:15 AM - 9:30 AM

[ACG45-02] The impact of full 3D ocean coupling to MJO simulations using the global cloud/cloud-system resolving model NICAM.

★Invited papers

*Tomoki Miyakawa1 (1.Atmosphere and Ocean Research Institute University of Tokyo)

Keywords:MJO, ENSO, global cloud resolving model

The global cloud/cloud-system resolving model NICAM and its new fully-coupled version NICOCO is run on one of the worlds top-tier supercomputers, the K computer. NICOCO couples the full-3D ocean component COCO of the general circulation model MIROC using a general-purpose coupler Jcup. We carried out multiple MJO simulations using NICAM and the new ocean-coupled version NICOCO to examine their extended-range MJO prediction skills and the impact of ocean coupling. NICAM performs well in terms of MJO prediction, maintaining a valid skill up to 27 days after the model is initialized (Miyakawa et al 2014). Here we focus on the initial 100 days to estimate the early drift of the model, and subsequently evaluate MJO prediction skills of NICOCO. Results show that in the initial 100 days, NICOCO forms a La-Nina like SST bias compared to observation, with a warmer Maritime Continent warm pool and a cooler equatorial central Pacific. The enhanced convection over the Maritime Continent associated with this bias project on to the real-time multi-variate MJO indices (RMM, Wheeler and Hendon 2004), and contaminates the MJO skill score. However, the bias does not appear to demolish the MJO signal severely. The model maintains a valid MJO prediction skill up to nearly 4 weeks when evaluated after linearly removing the early drift component estimated from the 54 simulations. Furthermore, NICOCO outperforms NICAM by far if we focus on events associated with large oceanic signals, such as the 1998 MJO event that is suggested to have ended the intense 1997/1998 El Niño.