Prediction of yellow lupin yield (Lupinus luteus L.) for northern Poland using weather-crop model
Abstract Prediction of yellow lupin yield (Lupinus luteus L.) for northern Poland using weather- -crop model. The paper presents an analysis of the impact of meteorological factors (solar radiation, maximum, minimum and mean temperature, precipitation) on the development and yield of yellow lupin Parys cultivar in the northern Poland in the years 1987-2008. Using multiple regression methods (linear and quadratic function) created regression equations that were estimated using the coefficients of determination (R2, R2 adj and R2 pred - using the Cross Validation procedure). Selected regression equation used to estimate the yield of yellow lupin, using generated - by means of model WGENK - daily values of global radiation, maximum and minimum temperature, precipitation, and climate change scenario GISS Model E for Central Europe. Examined dependencies weather-yield of lupine seeds (cultivar Parys) allowed the application of the chosen model to forecast yields from the time when the values are independent variables in the model by the end of the growing season. The comparison of distributions of actual and simulated yields shows that real yields are slightly (by 0.06 t·ha-1) higher than those generated for 2 × CO2 conditions.