排序方式： 共有2条查询结果，搜索用时 0 毫秒
1
1.
不同大尺度强迫条件下考虑初始场与侧边界条件不确定性的对流尺度集合预报试验 总被引：5，自引：0，他引：5
研究了不同大尺度强迫条件下的暴雨个例中,考虑不同尺度特征的初始扰动与侧边界扰动相互作用构造对流尺度集合预报的可行性,为进一步构建“自适应”于不同强对流天气的对流尺度集合预报系统提供依据。结果表明,在大尺度强迫显著的个例1中,以大尺度扰动信息为主的动力降尺度的增长趋势较集合转换卡尔曼滤波(ETKF)更为显著,且总扰动能量在预报中后期超过集合转换卡尔曼滤波,而在大尺度强迫较弱的个例2中,集合转换卡尔曼滤波扰动能量始终高于动力降尺度。此外,当大尺度强迫显著时,初始扰动与侧边界扰动相匹配会产生相互促进的作用,而不匹配时初始扰动会在预报中后期抑制侧边界扰动的发展,当大尺度强迫较弱时,即使是互相间不匹配的初始扰动与侧边界扰动也能在大部分预报时段起到相互促进的作用,说明初始扰动与侧边界扰动的相互作用机制不仅与天气形势相关,也与二者是否匹配挂钩,另外,扰动的发展特征同样依赖于天气形势;从集合离散度的角度来看,当大尺度强迫明显时,侧边界扰动的作用会在更短的时间内取代初始扰动,从而对离散度起到主导地位;两种初始扰动方法相比,集合转换卡尔曼滤波在多数情况下对总离散度的贡献均大于动力降尺度;从降水量预报及概率预报情况来看,大尺度强迫明显的个例可预报性更高,且各集合成员间的差异较小,大尺度强迫较弱的个例则相反,且当两种初始扰动方案与侧边界扰动相结合时,较仅侧边界扰动均有一定提高。 相似文献
2.
The impact of applying three-dimensional variational data assimilation (3D-Var DA) on convective-scale forecasts is investigated by using two mesoscale models, the Weather Research and Forecasting model (WRF-ARW) and the Hirlam and Aladin Research Model On Non-hydrostatic-forecast Inside Europe (HARMONIE-AROME). One month (1 to 30 December 2013) of numerical experiments were conducted with these two models at 2.5 km horizontal resolution, in order to partly resolve convective phenomena, on the same domain over a mountainous area in Iran and neighboring areas. Furthermore, in order to estimate the domain specific background error statistics (BES) in convective scales, two months (1 November to 30 December 2017) of numerical experiments were carried out with both models by downscaling operational ECMWF forecasts. For setting the numerical experiments in an operational scenario, ECMWF operational forecast data were used as initial and lateral boundary conditions (ICs/LBCs). In order to examine the impact of data assimilation, the 3D-Var method in cycling mode was adopted and the forecasts were verified every 6 hours up to 36 hours for selected meteorological variables. In addition, 24 h accumulated precipitation forecasts were verified separately. Generally, the WRF and HARMONIE-AROME exhibit similar verification statistics for the selected forecast variables. The impact of DA on the numerical forecast shows some evidence of improvement in both models, and this effect decreases severely at longer lead times. Results from verifying the 24 h convective-scale precipitation forecasts from both models with and without DA suggest the superiority of the WRF model in forecasting more accurately the occurred precipitation over the simulation domain, even for the downscaling run. 相似文献
1