Three strategies on data quality early warning of automatic soil water observations are setup: (1) Abnormal value detection of the automatic soil observation data is conducted based on the regularity of the soil moisture evolution. (2) From the relationship between soil observations, the soil hydrological constants are calculated, which are used to determine the theoretically soil moisture boundary for the automatic soil observation data limit tests. (3) Referring to the relationships between soil water content and soil water flow trend, the water potential flow trend coefficient is set up by using the real water contents as well as its field capacity and wilting moisture, which is used to analyze the unreasonable distributions of soil water among the different soil layers. Based on the data of early warning strategies, the operatiing scheme is setup for data quality check in real time. The outputs from the early warning system shows that the relative higher detection rate implies to the suspicious soil hydrological constants, which is used in calculating the soil moisture from the instrument signals, and the inappropriate observation environment of the instrument due to the soil hole, cracks, and so on around the instrument after installation. The warning results are consistent with the conclusions from the automatic soil water observation station field investigations. The warning effects are objective, reliable, and suitable for different levels of operation sectors to use automatic soil water data in routine services.