同化技术在渤海溢油应急预报系统中的应用

溢油应急预报对溢油事故现场处理具有重要指导意义。国内外已开展大量溢油数值预报技术研究,但由于各类误差的引入(尤其风和流数值预报误差的引入)以及模型本身的不完善等各种原因导致溢油数值预报无法满足日益提高的溢油预报精度需求。随着现场观测技术和监测水平的提高,如何充分利用实时观测数据提高业务化溢油应急预报精度,并满足应急预报迅速快捷的要求,成为目前业务化溢油应急预报的首要问题。国家海洋环境预报中心于2008年实现了渤海溢油业务化预报系统的建立和业务化应用,本文针对当前渤海溢油业务化应急预报中存在的现实问题,利用已有渤海海上5个石油平台从2010年1月至2011年2月的风场观测数据,初步开展最优插值方法(optimal interpolation assimilation method,OI)同化技术在国家海洋环境预报中心渤海溢油应急预报系统风场订正的应用研究。本文采用交错订正方法,确定了OI同化技术中相关尺度因子的选取,从而实现在这5个观测站地理分布情况下,OI同化技术应用中参数的最优化,之后在理想实验和实际案例的应用中,该同化方法明显提高渤海溢油预报精度。本文为如何进一步利用同化方法迅速快捷地实现溢油应急预报精度的提高提供了一定研究基础。

The assimilation technology application in the oil spill emergency forecasting system of the Bohai Sea

The trajectory numerical forecasting of oil spill is very important to the accident handling of oil spill. Many researchers did a lot of work on oil spill forecasting technology. However, the success of each model depends on the model structure and schemes, the accuracy of input data (such as wind, current, oil source information) and the interpretation of results. Input data, especially wind data and current data, are subjected to many error sources. Now, the accuration of oil spill numerical forecasting still can't reach the level we need. With the developing of observation technology and capability, it becomes a major question that how to use these real-time observation data to improve the accuration of oil spill numerical forecasting in order to meet the requirements of emergency forecast. The oil spill operational numerical forecasting system of the Bohai Sea was established on 2008 by National Marine Environmental Forecasting Center(NMEFC). In the paper, in order to resolve these real troubles in the oil spill operational numerical forecasting system of the Bohai Sea, the OI assimilation technology is applied in the oil spill emergency forecasting system of the Bohai Sea by using the observation data of wind field of five oil platforms from January 2010 to February 2011 in the Bohai Sea. The relevant scale factor, which is a very important factor in OI assimilation method, is determined by cross-correct method. Then the optimal value of parameter in OI assimilation technology is determined in the case of the five observation stations such this distribution. After that, the technology is applied in an ideal case and a real case, the results both approve that the applying of the OI assimilation technology improve the accuration of oil spill numerical forecasting results. This work provide a simple method to improve the accuration of oil spill numerical forecasting results, and it also meet the requirements of the oil spill emergency prediction in term of convenience and quickness.