Towards operational prediction of El Niño by a coupled ocean-atmosphere model
Climate Research Department, MRI/JMA, 1-1 Nagamine, Tsukaba, Japan.
: +81-298-53-8590, Fax: +81-293-55-2552,
The 1997-98 El Niño has turned out to be one of the largest events in history in terms of the anomaly of sea surface temperature in the Pacific equatorial area. The atmospheric general circulation in and around the Pacific tropical area has been greatly affected by the above El Niño, resulting in a number of abnormal or extreme weather events. In the later part of 1997, dry weather prevailed in Indonesia and Australia, higher temperature in India and Southeast Asian region, high temperature with dry weather in the northern part of South America, heavy rainfall in the southern part of the USA and in the northern part of Mexico, etc. Even such weather as heavy rainfall in eastern equatorial Africa in late 1997 might have been a result. They have caused various devastating disasters such as floods, or long-lasting forest fires, especially those in Indonesia.
These disasters remind us how important it is to get accurate information on El Niño, particularly early warnings by way of prediction. Recently, besides statistical methods, dynamical ways have been introduced in several institutions which use coupled atmosphere-ocean models on a research and/or operational basis. The Meteorological Research Institute (MRI), the El Niño Monitoring and Prediction Centre and other relevant divisions of the Japan Meteorological Agency (JMA) have developed such a model by coupling a global atmosphere model with an ocean model. In order to run the model, data assimilation is vitally important. The JMA has already developed an Ocean Data Acquisition System (ODAS) with higher resolution in tropical areas. Data are obtained as SHIP or BATHY from research vessels and voluntary observing ships. In addition, essentially important data are derived from TOGA/TAO buoys operated by NOAA/PMEL of U.S.A. in the equatorial region of the Pacific Ocean. Triangle Trans-Ocean Buoy Network (TRITON) buoys of the Japan Marine Science & Technology Centre (JAMSTEC) have recently been installed in the western part of it.
We have made a number of test experiments for the cases of past El Niño events. Based upon these results, we are going to launch operational prediction later this year. We know our model is not perfect, with various problems still left to be solved. They will be coped through our efforts to go forward step by step through real-time model output. Basic description and performance of the model will be briefly introduced together with existing issues such as climate drift, further improvement of data assimilation, etc.