| 统计方法在中国农业气象中的应用进展 |
| |
| 引用本文： | 杨霏云,朱玉祥,李文科,等.统计方法在中国农业气象中的应用进展[J].气象与环境科学,2016,39(3):121-129. |
| |
| 作者姓名： | 杨霏云 朱玉祥 李文科 等 |
| |
| 摘 要： | 统计学是目前农业气象工作中最有效并广泛应用的研究工具,农业气象统计比一般生物统计和气象统计有着更广泛的内容和应用更复杂的特点。统计方法被广泛应用于农业气象试验数据处理与监测评估和农业气象预报中,其中有统计假设检验、分布函数、相关和回归分析、聚类分析、小波分析、经验正交函数分解、灰色关联分析、时间序列分析、人工神经网络等,每种统计方法都有自身的优点和不足。统计方法在农业气象中的运用,取得了许多成果,但也存在不足。今后,应结合农作物生长发育的机理,深入理解各种统计方法的数学原理,加强统计方法应用的规范性,以减少应用统计方法解决农业气象研究问题中常出现的简单照搬、应用盲目的现象;应强化假设检验在农业气象试验数据分析方面的应用、概率密度和分布函数的适用性等方面的研究,加强相关和回归分析方法应用的数据独立性检验,注意灰色关联分析法、人工神经网络法和回归分析法等不同预测方法应用结果的比较,为统计方法在农业气象中的规范化应用提供参考。
|
| 关 键 词： | 统计 农业气象 应用 进展 |
Application Progress of Statistical Methods in Chinese Agrometeorology |
| |
| Yang Feiyun,Zhu Yuxiang,Li Wenke,et al.Application Progress of Statistical Methods in Chinese Agrometeorology[J].Meteorological and Environmental Sciences,2016,39(3):121-129. |
| |
| Authors: | Yang Feiyun Zhu Yuxiang Li Wenke |
| |
| Abstract: | Statistics has been an effective and widely used research tool for agrometeorology, and agrometeorological statistics has more extensive contents and features of more complex application than general biological and meteorological statistics. Statistical methods are widely used in the processing and monitoring assessment of agrometeorological experimental data and agrometeorological forecasting. The statistical methods include statistical hypothesis testing, distribution function, correlation and regression analysis, cluster analysis, wavelet analysis, empirical orthogonal function decomposition, grey relational analysis, time series analysis, artificial neural network, etc, and each method has its own merits and demerits. The application of statistical methods in agrometeorology has got a lot of achievements, and also has some shortages. It is pointed out that in the future we should combine with crop growth and development mechanism, deeply understand the mathematical principle of various statistical methods, strengthen the standard application of statistical methods to reduce the phenomenon of simple copy and blind application in agrometeorological statistics application. In addition, some concrete suggestions are proposed, including strengthen the application of hypothesis testing in agrometeorological experimental data analysis, the applicability of probability density and distribution functions, the data independence test in application of correlation and regression analysis, and pay more attention to comparison of different forecasting methods such as gray relational analysis, artificial neural network and regression analysis. So we can provide reference for the standard application of statistical methods in agrometeorology. |
| |
| Keywords: | statistics agrometeorology application progress |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《气象与环境科学》浏览原始摘要信息 |
| 点击此处可从《气象与环境科学》下载免费的PDF全文 |
|
