Abstract: In order to explore the potential of models introduced with pollen variable for forecasting harvest yields,progress in the forecast model research was summarized. It was indicated that the accuracy of the forecasting results could be enhanced by introducing pollen variable in traditional statistical models which were composed only of meteorological variables. Due to the limitation in the monitoring technologies,the forecast model was more applicable to the single crop in the same growing period. In the area where various crops were separated in small size fields in the same growing period,the forecasting accuracy of this kind of model was largely depended on the quality and quantity of the pollen data obtained by using the relevant monitoring technologies. In order to improve the application potential of pollen variable models in China.It is necessary to build the complete monitoring network,improve and develop the pollen recognition technology,and also accumulate enough pollen variable data for a stable and comprehensive yield forecast model.

Key words: Pollen density, Pollen density, Pollen flux, Pollen index, Crop yield, Forecast model