PARAMETER ADJUSTMENT TO A CROP MODEL USING A SENSOR-BASED DECISION SUPPORT SYSTEM

A knowledge-based system was developed to adjust input parameters to t he soil-water and rooting components of PNUTGRO, a process-oriented pe anut growth model. The system was developed to provide a better repres entation of temporal water status in the root zone of a growing crop. Soil water sensors provided input to adjust appropriate parameters bas ed on interpretation of their readings. These interpretations were pro grammed using human expertise combined with data from peanuts grown in lysimeters. A separate expert system screened sensor readings to insu re their validity before using their readings to adjust parameters. Te sts of the system over one season showed that model-based representati ons of soil-water status converged on sensor-based representations in the soil water regulation zone as the adjusted input parameters conver ged on new static values early in the season.