The development process and characteristic analysis of precipitation cloud system is an important issue in the field of cloud precipitation physics.Here, the 700 hPa Cloud Water Content (CWC) and the 1h value of airflow velocity (omega, OMG) in the vertical direction of the atmosphere are used to measure the chaos degree of CWC distribution via the information entropy and judge the cloud development via OMG time series, hence a combined prediction model is proposed based on hybrid multi-scale decomposition, Holt model, Autoregressive Integrated Moving Average model (ARIMA) and Lagrange Multiplier.The results show that, the CWC entropy has nonlinear and non-stationary characteristics;the clouds over the north have smaller means of the CWC entropy sequence and larger variance compared with those over the south regardless of the cloud development stage;a good temporal corresponding relationship is found between the regional average OMG and the extreme point reconstructed by the wavelet low-frequency of the CWC entropy, and close extreme value points account for 50% in clouds over the south and 83.3% in clouds over the north, showing that CWC entropy can somehow reflect the cloud development;the multiple timescale features of CWC entropy sequences make the multi-scale decomposed Holt-ARIMA-Lagrange Multiplier model more accurate than the single prediction method and single-layer decomposed prediction model, with accuracy improvement of more than 3%.