摘要
基于农田特有气象观测设备和环境属性,建立农业气象数据清洗标准和质控方法,以提升农业气象观测数据质量。针对数据属性异常和重复记录情形,选取Bohn数据清洗模型的空缺值清洗方法和噪声数据清洗方法。通过农业微气象观测站点空间内观测要素历史数据统计,获取清洁数据指标,应用于数据质量动态阈值生成方法,建立农业微气象数据质量控制模型。清洗质控后的数据评估指标表明,经过数据清洗和质控模型后数据准确率和重复性均有明显改善。数据清洗质控方法有助于准确获取农业气象灾害监测信息,为农业的防灾减灾提供有效决策支撑。
In order to improve the quality of agrometeorological observation data,the cleaning standard and quality control method of agrometeorological data are established based on the unique meteorological observation equipment and environmental attributes of farmland.For the case of abnormal data attributes and repeated records,the method of cleaning the blank value of Bohn data cleaning model and the method of cleaning the noise data are selected.Through the historical data statistics of observation elements in the space of agricultural micro meteorological observation station,the clean data index is obtained and applied to the dynamic threshold generation method of data quality,and the quality control model of agricultural micro meteorological data is established.The data evaluation indexes after cleaning and quality control showed that the accuracy and repeatability of the data are significantly improved after data cleaning and quality control model.The data cleaning quality control method is helpful to obtain the monitoring information of agrometeorological disaster accurately and provide effective decision support for agricultural disaster prevention and reduction.
作者
周强
ZHOU Qiang(Shandong Meteorological Service Center,Jinan 250031,China)
出处
《湖北农业科学》
2020年第14期37-40,51,共5页
Hubei Agricultural Sciences
