摘要
利用中国近岸的14个测站水位数据比较杜德森方法、鲁斯特方法、陈宗镛方法和25 h平均法计算不同时间尺度平均海平面的效果。研究结果表明,以杜德森方法作为基准,用鲁斯特方法、陈宗镛方法、25 h平均法与杜德森方法计算日平均海平面的平均均方根误差分别为29.6 mm、57.3 mm和40.3 mm。随着平均海平面时间尺度增大,鲁斯特方法和陈宗镛方法计算的结果与杜德森方法得到的结果越接近,31点平滑后与杜德森方法的均方根误差都在3 mm左右,而25 h平均法得到的长时间尺度的平均海平面则误差较大,91点平滑后与杜德森方法的均方根误差仍大于10 mm。由于杜德森方法和鲁斯特方法需要次日的潮位数据而会造成少量资料信息量的丢失,而陈宗镛方法不会造成资料信息丢失,且计算简单,因此利用月平均海平面数据或更长时间尺度海平面数据分析和研究海平面长期变化规律时,采用陈宗镛方法最为适宜。但资料较少且不能忽略海平面短周期变化时,用杜德逊法最佳。
This paper compares the effects of Dodson method,Rossiter method,Chen Chung-yung method and 25 hours average method to calculate the different time scales mean sea level,by using the water level data from 14 stations along the China coast.The results show that by using the Dodson method as a benchmark,the average RMS error between Rossiter method,Chen Chung-yung method,25 hours average method and Dodson method in calculating daily average sea level are 29.6 ram、57.3 mm and 40.3 mm,respectively.If the mean sea level time scale is larger,the results of Rossiter method and Chen Chung-yung method are closer to the result of Dodson method.Besides,and after 31 points smoothing,the RMS errors between the two methods and Dodson method have both been about 3 mm,while the long time scale mean sea level error from 25 hours average method is larger than 10 mm after 91 points smoothing.Dodson method and Rossiter method will lose of a small amount of data for needing the tide data of the next day,but Chert Chung-yung method will not,and the calculation is simple.Therefore,in analyzing and studying sea level long-term change rule by using monthly mean sea level data or a longer sea level data,the Chen Chung-yung method are effective.However,when the sea level changes in short time-scale cannot be ignored,Dodd Johnson method is best.
出处
《海洋预报》
北大核心
2013年第5期37-43,共7页
Marine Forecasts
基金
国家自然科学基金(41276018)
海洋公益性行业科研专项(201005019-3)
关键词
平均海平面
杜德森方法
陈宗镛方法
平滑滤波
mean sea level
dodson method
then chung-yung method
smoothing filter
