基于空间梯度信息的质量控制方法在红外高光谱AIRS资料同化中的应用研究

Application of Quality Control Method Based on Spatial Gradient Information in Assimilation of Infrared High Spectrum Data AIRS

本文利用中尺度非静力WRF(Weather Research Forecast,Version 3.4)模式,针对1013号"鲇鱼"台风个例,通过对红外高光谱AIRS(Atmospheric Infrared Sounder)资料经过基于空间梯度信息的质量控制之后同化进入模式,来评估新的质量控制方法对同化效果的影响以及对台风数值模拟的改善情况。研究结果发现,如果仅仅基于WRFDA(WRF Data Assimilation system,Version 3.4)模式自带的质量控制系统,将会有部分梯度距平值明显较大超过阈值的资料被同化进入模式,而这些可能受到"污染"且误差较大的资料同化进入模式必将会导致同化结果有较大误差,影响分析结果的质量。而对AIRS资料经过基于空间梯度信息质量控制之后再同化进入模式,确实可将梯度距平值大于阈值的"坏点"剔除掉,从而使初始场的描述更加准确,台风路径的模拟精度在一定程度上得到提高。综上可知,基于空间梯度信息的质量控制方法整体上对改善同化效果有较好的正效应,对台风的数值模拟也起到一定的促进作用。

In this study, using nonhydrostatic mesoscale Weather Research Forecast （WRF） Version 3.4 model, a new quality control （QC） method is introduced that uses infrared high spectrum Atmospheric Infrared Sounder （AIRS） data based on spatial gradient information. To access the effect of the new QC method on numerical typhoon simulation, assimilation and simulation experiments were performed using Typhoon Megi （2010）. The results show that if using only the previous QC method provided by the WRF data assimilation system, some data points with gradient anomalies exceeding the threshold value are assimilated into the model. These data points may be contaminated by clouds or misidentified and can be regarded as ＂bad＂ points. When such bad data are assimilated into the numerical model, large errors can be brought into the model, which can significantly affect the quality of the initial analysis field. In the new QC method, gradient information checking should be done first to find and eliminate the bad data points. The new method results in a more accurate initial field and improvement in the simulated typhoon tracks. Therefore, the new QC method based on spatial gradient information has a positive effect on improving assimilation results and plays an important role in the numerical simulation.