Crop status evaluations and yield predictions

A model was developed for predicting the day 50 percent of the wheat crop is planted in North Dakota. This model incorporates location as an independent variable. The Julian date when 50 percent of the crop was planted for the nine divisions of North Dakota for seven years was regressed on the 49 variables through the step-down multiple regression procedure. This procedure begins with all of the independent variables and sequentially removes variables that are below a predetermined level of significance after each step. The prediction equation was tested on daily data. The accuracy of the model is considered satisfactory for finding the historic dates on which to initiate yield prediction model. Growth prediction models were also developed for spring wheat.