Further investigating the performance of crop water stress index for maize from baseline fluctuation, effects of environmental factors, and variation of critical value

To further analyze the accuracy and applicability of empirical (CWSI_E) and theoretical (CWSI_T) crop water stress index calculation methods, study was conducted at the USDA-ARS Limited Irrigation Research Farm (LIRF), Colorado, USA during two maize growing seasons in 2013 and 2015. The growth stage and seasonal changes of non-water-stressed baseline (NWSB) and non-transpiration baseline (NTB), the effects of environmental factors, and the estimation performances of water stress, grain yield, and WUE were analyzed. Results show that significant correlations (p < 0.001) with vapor pressure deficit (VPD) were found for two baselines, however, the distributions with the changes of VPD and growth stage and seasonal changes of NWSB and NTB were different between two methods. Specifically, neither growth stage nor growing season significantly affects the baselines of CWSI_E, indicating that the baselines of the empirical method were more stable than those of the theoretical method. The lower baselines and smaller difference between NWSB and NTB were more likely to be observed in the theoretical method than the empirical method. The greater CWSI values were observed for the theoretical method because of the relatively smaller difference between NWSB and NTB. VPD with values greater than 1.5 kPa may a suitable environmental criterion to be used to indicate the applicability of the two CWSI methods for crop water stress estimation. Both methods could track maize water stress with R2 of 0.55 for CWSI_E and 0.49 for CWSI_T (n = 236) with sap flow measurements and could estimate grain yield with the highest R2 of 0.95 and WUE with the highest R2 of 0.87 (n = 24). However, greater values and a clear downtrend for three one-hour periods were observed for CWSI_T because of the relatively smaller difference between NWSB_T and NTB_T. This study contributes to the knowledge in the area of stable and accurate monitoring of crop water status based on CWSI.