Comparative Analysis Of Characteristics For Background Error Covariances In East-China And Plateau Region

It’s an important means of making use of data assimilation methods to extract useful information of observational data for improving weather forecasting system level.Background error covariances’ information plays a key role in the Data Assimilation.Background error covariances affect the performance of the assimilation system directly.Therefore, a reasonable estimation of the background error covariances are on a critical place in data assimilation.The characteristics of background error covariances have a great relationship with local meteorological features, Weather and climate characteristics of the region can be reflected in the background error covariances.In order to get a better understanding of the characteristics of background error covariances(BE) of East-China and Plateau region, the major parameters of BE,such as correlation coefficients, eigen-values, eigen-vectors and eigen-lengthscales were compared and analysised, based on the latest WRFDA Model of multivariate background error(MBE) statistics,using one month-long series of numerical forecast fields. The main conclusions are as follows:(1)Results indicate that the relationships between diferent variables in Plateau region are more notable than them in East-China.Unlike East-China,the diviation of vertical background in Plateau region is more obvious.In contrast to East-China,the atmosphere characteristics in Plateau region have smaller horizontal lengthscales and clearer local property,so the observations in Plateau region put forward higher requirements.(2) It’s the similar response on field U and fields V. Response range is significantly broader than the east China in plateau region, but the gradient of the east China is larger than the plateau region.The response to the temperature field in the east China is broader and higher-strength than the response in the plateau region.Relative humidity field in the plateau region is almost invisible,however,relative humidity field in east China’s not only large-scale, and also high-strength, in contrast with the plateau region brightly.(3) Whether in east China or the plateau region, the average error and the root mean square error of multivariate background error covariances are smaller than a single variable background error covariance,explaining that the multivariate background error covariances play a good role in improving assimilation effect.System error and root mean square error in the plateau region is much larger than them in the east China, describing that the forecast value and the actual value are quite different in the plateau region, the forecasting results are poorer.