Abstract
A rule-based, fuzzy logic crop water stress index (FL-CWSI) was developed using growth chamber data and tested on tall fescue canopies grown in the greenhouse. The FL-CWSI always gives an index value between zero and one. The selection of the FL-CWSI defuzzification methods-center of area (COA) or mean of maxima (MOM) is not important. COA and MOM defuzzification tests were quite similar for the growth chamber conditions (
r
2=0.92) and greenhouse conditions (
r
2=0.99), respectively. FL-CWSI values correlated well for growth chamber CWSI values (
r
2=0.70) and reasonably well for the greenhouse (
r
2=0.50). The results also demonstrated that FL-CWSI with 150 rules predicts the correct water stress development trend, similar to the empirical and more complex theoretical CWSI methods. FL-CWSI eliminated the necessity for measurements of canopy resistance, aerodynamic resistance, and soil heat flux, needed for predicting theoretical crop water stress indices. A FL-CWSI rule system is attractive for turf irrigation control, because it can be easily adapted and programmed on commercial micro-controllers.