Aiming at the problem of low data availability of FDR automatic soilmoisture station, based on the working principles of FDR automatic soil moisture sensors, combined with the data processing procedures and methods, about 450 secondarylevel calibration parameters of 60 station in Hunan and the related observation data are analyzed. It is noted that by the current calibration methods, it is difficult to conducted full scale calibration under natural conditions, and the unreasonable secondary calibration parameters are the root cause of the low data availability of FDR automatic soilmoisture data. The two calibration parameters are not reasonable, for the slope of the equation is too large, too small, or negative values, leading to such problems as too large increase amplitude, unchanging, or opposite to the reality of observation data. The suggestions of largesample undisturbed soil sampling and laboratory calibration are put forward aiming at the problems of current data calibration method. The method is verified preliminarily, and the results show that the method can effectively improve the quality of soil moisture observation data from the sources.