Advances in data assimilation to improve the quality of weather and environmental forecasts

数据同化涉及利用从各种仪器获得的所有现有观测信息来估计大气的当前状态，以提高天气预报的质量。基于统计估计原理，通过数值天气预报模式（NWP）获得的短期预报提供大气的先验估计。其精度由背景误差协方差表示，这取决于气象条件。在与大气不稳定开始有关的情况下，初始条件需要包含能够触发天气系统发展的动力结构。这项建议涉及在数据同化方面取得进展，办法是在同化过程中引入依赖气流的结构，从而能够提高短期、中期和长期高影响天气事件的预报质量。这也将应用于适应性数据选择，以改进业务气象中心每天收到的大量卫星数据的使用。

Data assimilation is concerned with using information from all available observations obtained from a variety of instruments to estimate the current state of the atmosphere to improve the quality of weather forecasts. Based on statistical estimation principles, an a priori estimate of the atmosphere is provided by a short term forecast obtained from a numerical weather prediction (NWP) model. Its accuracy is represented by the background-error covariances, which depend on the meteorological situation. In cases associated with the onset of atmospheric instability, the initial conditions need to embed the dynamical structure that can trigger the development of a weather system. This proposal is concerned in making advances in data assimilation by introducing flow-dependent structures within the assimilation process that are in a position to improve the quality of forecasts of high impact weather events in the short, medium and long range. This will also be applied to adaptive data selection to improve the use of the vast amount of satellite data received daily at operational weather centres.