数值模式误差对降水四维变分资料同化及预报的影响

利用一个无量纲的水汽发展方程,针对同化时间窗口内出现和不出现降水两种情况,分析了不同模式误差和初始误差对降水四维变分资料同化预报效果的影响。结果表明,应用四维变分资料同化方法进行降水预报前,应该充分考虑数值模式中的误差,才能得到比较满意的同化及预报结果。假定同化窗口内获得的比湿观测场是准确的,当不存在模式误差时,四维变分资料同化方法可以有效地消除初始场误差,找到比湿真实初始场;而存在模式误差时,四维变分资料同化后的模式初始场会偏离真实的比湿初始场,并且模式误差越大,偏离程度越严重。在一些模式误差情况下,由于模式误差在同化窗口及延伸预报时段的作用不同,进行四维变分资料同化处理后,尽管累积降水量的预报结果在同化时间窗口内优于同化前的预报,而在最终预报时刻反而差于同化前。

The Effect of Model Errors on The 4D-VAR and Forecast of Rainfall

Using a scaled equation of specific humidity,an analysis of the effect of the model error and the initial data error on the 4D-VAR and the forecast result of rainfall is performed according to the two situations that with rainfall in the window of 4D-VAR and without rainfall in the window of 4D-VAR.The results indicates that the model error should be considerated sufficiently before the 4D-VAR of rainfall,then,a satisfied assimilation and forecast result will be obtained.In the case that the numerical model is correct and there is initial data error,if the observational data of specify humidity during the window of 4D-VAR is the true state,the error in model initial data will be effectively removed and the true initial data will be retrial after the 4D-VAR.If there are model errors,the initial data after the 4D-VAR will deviate the true initial data of specify humidity,and the deviation as large as the model errors.When the model has errors and the errors work differently at the window of assimilation and the window of forecast time,although the forecast result of the accumulated rainfall can approach the observational data during the assimilation window,it will become less advantageous than the forecast without 4D-VAR in the window of forecast time,even becoming worse than it at the end of the forecast time.