自动土壤水分观测资料应用误差分析

根据从国家气象信息中心实时资料数据库读取的自动土壤水分监测资料,对比同日人工测定的土壤相对湿度数据,发现白天各时次自动监测数据与人工监测数据之间均存在15%左右的差异。具体针对监测仪器自身的结构特点以及中国气象局的业务要求,从两种监测方法自身的监测地段代表性、数据测量和换算、监测土层深度、监测时间、土壤水分常数测算、土壤结构变化等多种角度分析数据间存在差异的原因;认为多种可以估算的差异"叠加"在一起时,对比差异最大可能达到20%左右,加上其他误差因素的影响,实际应用中出现15%左右的差异是可以解释的;选择较长序列的站点实测资料进行了数据差异的实况分析。结合业务中常用的墒情等级判断指标,分析数据差异对客观判断构成的数据干扰和数据损失。建议直接利用体积含水量进行业务应用分析,探究全新的自动土壤水分监测数据应用方法,在具体应用中建立相应的指标体系或指标模型,充分发挥自动土壤水分监测的优势。

Error Analysis of Automatic Soil Moisture Observation Data

Comparison is conducted between automatic soil moisture monitoring data and manual measured relative soil moisture data from the Meteorological Data Storage System (MDSS) of the National Meteorological Information Center (NMIC), and it is found that there is about 15% of difference between different observing times in daytime. According to the structural characteristics of the two sets of measurements and the operational requirements of the Chinese Meteorological Administration (CMA), the reasons for the differences are analyzed from several aspects, such as measurement and calculation, repeatable errors, differences in monitoring soil depth and monitoring time, calculation of soil moisture constant, and soil structure change, etc. It is concluded that when the various differences are overlaid together, the maximum difference of about 20% can be found; other factors are also known to impact the difference, so a difference of about 15% is reasonable in practice. The differences of the measured data between observing sites are analyzed by selecting a long sequence observation data. According to commonly used operational indicators of moisture level, the data disturbance and data loss influencing objective judgment induced by the data differences between the two methods are analyzed. It is suggested that the volumetric water content should be used directly in operational application analysis; new application methods of automatic soil moisture monitoring data should be developed, and the corresponding indicator systems or models should be established; the advantages of automatic soil moisture monitoring should be fully taken.