Sensitivity of Weather and Plant Parameters Used in Fuzzy-Neuro Model for Irrigation Scheduling of Greenhouse Rose

Fuzzy neuro model for drip irrigation scheduling of greenhouse rose was developed using the observed values of temperature, solar radiation, humidity and leaf area. Experiments were conducted with First Red rose variety grown in climate controlled greenhouse. Three treatments namely irrigation scheduling based on 100%, 80% and 60% of ET_c were considered to evaluate the effect of irrigation on yield and other plant parameters. Model was developed using the observations obtained from the best treatment. Calibrated and validated model was subjected to sensitivity analysis for identifying the parameters to which model was relatively sensitive in terms of prediction of irrigation quantity and interval. Sensitivity analysis was carried out for temperature, solar radiation, humidity and leaf area with respect to reference values of each input parameters, which were taken as average values recorded during the study period. Irrigation quantity and irrigation interval were predicted by changing the input parameters above and below the reference values. Results revealed that model was very sensitive to changes in leaf area followed by solar radiation, humidity and temperature.