re discrimination indexes with practical significance. By means of continuous observation data from automatic weather stations, combined with the artificially observed weather phenomena records, based on the research findings on the formation causes of dew, frost, and freezing, the basic indicators qualified as a comprehensive criterion (air temperature, relative humidity, ground temperature, air grass surface temperature, and near surface air temperature) are selected as the main factors, considering the influence of rainfall and weather conditions on the formation of these weather phenomena. The minute to minute observations and artificially observed weather phenomena records of two stations are used for statistical analysis. After factor selection and indicator determination, by means of the multi element combination method, the discrimination indexes of dew, frost, and freezing occurrence under the natural environment are extracted, which are satisfactory with a fitting rate of above 80%. These indexes can be used to provide references for further understanding of the occurrence and variation regularities of these weather phenomena to provide the reference and the basis for automatic weather phenomena observation.