气象格点资料的准无损压缩方法  

Quasi-lossless Compression Encoding for Meteorological Grid Data

作　　者：罗坚[1] 赵苏璇[2] 姜勇强[1] 

机构地区：[1]解放军理工大学气象学院,南京211101 [2]南京军区空军气象中心,南京210018

出　　处：《数据采集与处理》2011年第3期320-325,共6页Journal of Data Acquisition and Processing

基　　金：国家自然科学基金(40905021)资助项目;国家重点基础研究发展规划("九七三"计划)(2004CB418301)资助项目

摘　　要：格点数据是气象资料存储交换的重要形式,资料量的爆炸性增长给数据的存储、传输带来了巨大压力。采用离散余弦变换(Discrete cosine transform,DCT)构造了气象格点资料的准无损压缩方案,对位势高度、经向风、纬向风及温度4个要素,分别按照3种精度对压缩效果进行了对比,并对该方案在要素场的直接分析和作为初始场进行模式积分两类应用进行了误差分析。结果表明,该方案可以大幅度提高气象格点数据的压缩效率,由此带来的压缩损失非常小,无论是用于数据的直接分析、诊断计算还是模式积分,其误差是可以接受的。该方案能提高气象海量数据的存储和传输效率,为提供快速气象数据响应和预报服务打下基础。Grid data is one of the important forms for the storage and exchange of the meteorological data.The explosive increae of the data brings huge pressure for storage and exchange.In this paper,a quasi-lossless compression encoding method for meteorological grid data is presented by using discrete cosine transform（DCT） method.The compression efficiency of geopotential height,meridional wind,zonal wind and temperature is comparatively analyzed according to three precision types.Furthermore,the error analysis is carried out in the direct decoding-data analysis and the model integration where the three types of data are taken as the initial fields.Results show that the method significantly improves the coding efficiency of meteorological grid data,and the coding loss is extremely small compared with original data.The errors of direct analysis,diagnosis calculation and model integration are all acceptable.The method can improve the storage and the exchange efficiency of the massive meteorological data,and provides a good foundation for rapid meteorological data respondence and weather forecasting service.

关 键 词：气象格点资料 准无损压缩 离散余弦变换 误差分析 

分 类 号：V321.2[航空宇航科学与技术—人机与环境工程] X16[环境科学与工程—环境科学]