Abstract:

To realize the soybean yield forecast at the provincial and municipal levels by every ten days, this paper analyzed the spatial and temporal distribution characteristics of soybean yield data in Heilongjiang Province. Established a temperature, precipitation, sunshine and comprehensive climate suitability model based on the meteorological data of each period, and analyzed their correlation with the relative meteorological yield. A dynamic yield forecast model based on climate suitability index was constructed to forecast soybean yield in Heilongjiang Province. The results showed that: (1) the annual per unit area yield of soybean decreased from south to north in space and showed interdecadal variation in time. There was a significant difference in annual total yield among cities and counties, and Nenjiang City had the highest soybean yield; (2) the climate suitability index of Heilongjiang Province and Nenjiang City from 1995 to 2015 was significantly correlated with the corresponding relative meteorological yield of soybean, the climate suitability model could objectively reflect the meteorological conditions in each growth period of soybean; (3) from 1995 to 2015, the average accuracy rate of model back substitution test was more than 80%, the trend accuracy year of each period was more than 12 years, and the accuracy of model extrapolation forecast from 2017 to 2019 exceeded 85%. The established yield forecast model can provide a reference basis for soybean yield prediction in Heilongjiang Province.

Key words: Heilongjiang Province, soybean, climate suitability model, yield forecast model, dynamic yield forecast