Snow water equivalent modeling and predicting method based on zero-expansion space-time regression model

The invention provides a snow water equivalent modeling and predicting method based on a zero-expansion space-time regression model, and mainly solves the problem that in the snow water equivalent and environmental factor modeling process, the time-space regression model is not influenced. The problems of a large number of zero observation values and snow water equivalent time lag effects occurring along with time-space dramatic change of accumulated snow, and model missetting, variance expansion, coefficient deviation and the like caused by spatial distribution autocorrelation are solved. According to the method, two parts of binary regression and continuous regression are constructed by introducing a time lag term and utilizing a generalized linear space regression model to represent two processes of judging whether a zero value exists or not and estimating a numerical value of a non-zero value, so that the space-time effect and the zero expansion effect are considered at the same time, the modeling and pre