Study on Anomaly Data Detection Method for Automatic Soil Moisture Observation

Abstract

Soil moisture plays a crucial role in the study of agricultural drought monitoring, yield prediction, soil erosion, and so on, which is of great significance for agriculture, drought, and climate. With the advantages of high precision, high spatial and temporal resolution, and non-destructive, the automatic soil moisture observing instrument has become an important component of automatic soil moisture observation station for the meteorological department in china. In the process of observing soil moisture, the accuracy of data is seriously affected by the calibration methods, equipment stability, soil texture, etc. Therefore, it is especially important to establish a quality control method for the automatic soil moisture observation data from the origin of affecting the observation quality. To solve the outstanding quality problems in the automatic soil moisture observation data, firstly, based on the historical data, the inherent variation characteristics of soil moisture were studied. Secondly, combined with the instrument observation principle, data characteristics and error sources of abnormal data, classified and statistical analyzed the form of abnormal data caused by various reasons and given the threshold, this paper preliminarily puts forward a practical set of quality control methods for the hourly soil moisture observation data. Finally, the application effect of the quality control method was verified by using the data in china in 2019. The result shows that: (1) Four mainly categories of abnormal data are gross errors, mutation, abnormal high values, and stiffness, which are mostly caused by instrument failure, abnormal soil hydrological constant, and unreasonable calibration of sensors. (2) The method of frequency characteristic detection can identify the error caused by instrument fault. (3) This quality control method can effectively detect abnormal data in china. At present, the method has been applied to the Integrated Meteorology Observation Data Quality Control System.