Abstract:

With statistical regression methods, the use of meteorological and soil moisture data of different soil types in the three agricultural meteorological stations from 1988 to 2006, selecting four factors 10 days ago :temperature, precipitation, evaporation and soil moisture , the 0 ~ 40cm soil moisture dynamic forecasting models are established from May first ten days to the late ten days of Auguste of Hulunbeier city, 68.8% of the equation through a test of significance; The fitting relative error is less than ± 10% to 77.8%, less than ± 15% for 84.3% ; Using of the measured data to verify ,the model shows that an average error is about 20%, can be used to predict the 0 ~ 40cm soil moisture dynamic during plant growing season at all levels in Hulunbeier city.