Evaluation and Selection of Weather Variables for NRED Yield Simulation Model

Excerpts from the report: This review of large scale weather-crop yield equations applicable to NRED use suggests that: A) there is a common hydrological basis for all these weather-crop yield models and that these basic methods have been compared in literature. In addition, B) Procedures used to incorporate these basic hydrological variables tend to follow logical sequences which can be projected to almost any area which NRED seeks yield estimates. In regards to the Palmer index methods of Perrin and of Sakamoto, a comparison of these two methods has not been done but questions arise as to both their utilities. Because of the availability of the Palmer drought index data on magnetic tape for computer use for this project, a methodology based on either of these two indexes or a procedure suggested by these procedures should be employed. In conclusion, Lowrey in a review of crop-yield/weather models suggested that no really universal model is possible because of the great diversity of crops and cultural practices in the world as well as the diversity of crop-climate interactions. With this in mind, the use of dummy variables to represent "steps" in the time-series yield equation, and the selection of appropriate coefficients based on probable physiological responses to local climates appears to be a logical choice in weather-crop modeling methodology.