基于遗传优化的西太平洋副高异常活动年份的动力预报模型反演

Retrieving dynamic forecast model of the western pacific subtropical high in abnormal years based on GA

基于2000—2010年NECR/NECP的500hPa位势场资料,用EOF时空分解方法和动力模型重构思想,通过遗传算法的全局优化搜索和并行计算途径,开展了500hPa位势场动力预报模型反演,建立了刻画副高活动的非线性动力预报模型,实现了副高活动的中长期预报.模型预报试验表明,该模型对副高的中长期活动,尤其是异常活动能够较好地描述和预报,进而为副高等复杂天气系统的预报探索了新的方法思路.

On the basis of the potential field data on 500 hPa from 2000-2010 from NCEP/NCAR, introducing the ideas of EOF（empirical orthogonal function） time-space separation and the dynamic system reconstruction of time series, with the advantages of the global optimization and parallel calculation by genetic assistance, dynamic model inversion is carried out, thus a nonlinear forecast model of the subtropical high activity and aberrance is established. And a mid/long-range forecast of subtropical high activity was carried out. The results of dynamic model forecast experiment showed that the mid/long-range forecast of subtropical high pressure by our model can be very actual. Especially, the aberrance of the subtropical high pressure can be drawn and forecast. A new method of idea is presented for diagnosing and forecasting such complicated weathers as subtropical high activity.