基于乡镇拟合数据的农业气象灾害风险研究

发布时间：2018-05-23 07:28

  本文选题：乡镇 + 气象数据　； 参考：《齐鲁工业大学》2015年硕士论文

【摘要】：本文通过气象数据特征分析,建立数据拟合方法,在此基础上开发系统软件,并对实际应用效果校验和评述。根据1981-2010年逐日气压、气温、相对湿度、风速、降水量五要素气候标准值的分布特征分析,气压、气温、相对湿度要素变化基本呈正态分布,风速分布曲线为左偏太,降水量分布曲线大致为指数。再利用2007-2014年气象数据计算五要素相关性。结合要素分布特征和相关系数分析,气压、气温、相对湿度可直接利用线性方程,拟合乡镇区域站气象资料,风速、降水量需做正态分布转换。最后对拟合得到的2014年气象数据与乡镇区域站观测数据进行校验,只有个别月份气象数据超出仪器标准误差,说明拟合数据可用。其中气温数据符合最好,误差基本在0.2℃之内,风速和降水量拟合误差对小观测量影响较大。系统结合现有气象业务开发,以实用为主,保障拟合的准确性。系统包括系统管理、报文收集转换、质量控制、数据拟合、小麦风险度计算、特色农业风险度计算、数据查询、应用工具等模块。利用系统多次插补观测期数据、重复计算回归方程系数,保障拟合气象数据最接近观测值。利用农业气象灾害风险评估系统计算东阿县农业灾害风险度,得到如下结论:小麦气象灾害风险按旱涝、高温、低温、倒伏四项指标计算,小麦综合农业气象灾害风险度在0.47~0.51之间,各乡镇综合气象灾害风险度处于中、高风险区。特色农业油用牡丹的综合农业气象灾害风险度在0.25~0.65之间,处于综合气象灾害低到高风险区。课题建立了以乡镇为基点的山东省气象拟合资料集,使气象资料分布达到0.1°x0.1°,为气象资料的共享提供了基础;并在观测和实验得到气象灾害数据基础上建立部分特色农业的气象灾害风险模型,为特色农业生产提供了技术指导。
[Abstract]:Based on the feature analysis of meteorological data, the method of data fitting is established, and the system software is developed on the basis of which, and the practical application effect is verified and reviewed. According to the analysis of the distribution characteristics of daily air pressure, air temperature, relative humidity, wind speed and precipitation from 1981 to 2010, the variation of air pressure, temperature and relative humidity is basically normal distribution, and the distribution curve of wind speed is too far left. The precipitation distribution curve is approximately exponential. The correlation of the five elements is calculated by using the meteorological data of 2007-2014. Based on the analysis of the distribution characteristics of elements and the correlation coefficient, the linear equation can be directly used to fit the meteorological data of township regional stations. The transformation of normal distribution of wind speed and precipitation is needed. Finally, the fitting meteorological data of 2014 and the observation data of township regional stations are verified. Only the meteorological data in a few months exceed the standard error of the instrument, which shows that the fitting data can be used. Among them, the temperature data is the best, the error is within 0.2 鈩，

本文编号：1923819