Supercooled water is a crucial parameter for assessing the potential for weather modification in stratiform clouds. Utilising data from 32 aircrafts detected by the Airborne Meteorological Detection Device from 2017 to 2022, we conduct a statistical analysis of the altitude characteristics where supercooled water is present, along with temperature, humidity, vertical wind speed, and other atmospheric parameters. We also examine the variation of supercooled water content in relation to these environmental factors. Based on these findings, we develop a diagnostic forecasting method for supercooled water content in clouds over the Shanxi region. The study reveals the following insights: (1) The supercooled water content obtained by the Airborne Meteorological Detection Device shows good agreement with the liquid water content detected by the Cloud Particle Probe (when the temperature is below 0 ℃, the liquid water content detected by the CDP is the supercooled water content), which indicates that both instruments are better at detecting the supercooled water content in the cloud. (2) The supercooled water content predominantly ranges from 0.06 to 0.22 g·m^-3, with an average value of 0.18 g·m^-3. Its probability density function (PDF) exhibits a distinct single-peaked normal skewed distribution, and the cumulative distribution function (CDF) growth rate decreases as the supercooled water content increases. (3) The supercooled water region is predominantly found between 3589 and 4667 metres, approximately 1011 to 2316 metres above the 0 ℃ isotherm. The temperature range is from -8.52 ℃ to -3.52 ℃, humidity levels vary between 86.68% and 100%, and the supercooled water area is predominantly influenced by updrafts. (4) The supercooled water content tends to increase with decreasing altitude from the 0 ℃ isotherm, rising temperature, increasing relative humidity, and stronger updrafts within the clouds. (5) The relationships between the content of supercooled water and various factors such as temperature, relative humidity, altitude above the 0 ℃ isotherm, and vertical wind speed are determined through polynomial fitting. The analysis reveals that the inversion magnitude is substantial when the supercooled water content is low, and conversely, it is minimal when the content is high. However, when considering the overall distribution, there is a distinct positive correlation between the two variables. The retrieval results for six cases outside the statistical sample suggest that some supercooled water is also forecasted during periods when it is actually detected, with the trends of the two being largely consistent and accuracies exceeding 65%. However, the forecasted values are slightly higher than the observed values. This diagnostic method for forecasting supercooled water can, to a certain extent, qualitatively assess the magnitude of supercooled water content under specific conditions, locate areas rich in supercooled water, and offer guidance for the scientific advancement of artificial weather modification operations.