Опыт прогнозирования урожайности сельскохозяйственных культур с использованием имитационных моделей

The analysis of the components and parameters of the EPIC bioproductivity simulation model for agricultural conditions in the central group of agricultural districts of the Krasnoyarsk Region was made. For the component of the EPIC model “Crop Growth Model”, the data sources were selected to evaluate its weather conditions parameters. The map of the fields with their coordinates, information about cultivated crops, as well as the access to remote distance Earth zonding sensing data was carried out through the agro-monitoring service ISIT SFU. Simulation model parameters usually not measured at meteorological stations - daily total solar radiation, leaf area index - was made. Software implementing the core of the yield forecast model was developed. The time series of the total daily solar radiation and leaf area index were calculated as a function of the normalized difference vegetation index NDVI for the vegetation period of 2018. The data from meteorological station ‘Sukhobuzimskoye' were processed into the necessary for the model format: the average, maximum and minimum daily temperatures were calculated, and the cloud index was estimated. The simulation of daily increase in the aboveground biomass of crops depending on the air temperature in the current growing season was carried out. The simulation results of daily biomass growth and the sum of these growths during the growing season - the aboveground biomass - were consistent with yield data characteristic of the Sukhobuzimo district of Krasnoyarsk Region - 23-24 c/hectare. For adverse weather conditions modeling, EPIC used stress coefficients having the values from 1 (the best conditions for this factor) to 0 (the culture does not grow at these parameter values). The demonstration of model behavior in conditions of lack of moisture by direct construction of a series of water stress coefficient was shown.