Comparability analyses are performed to investigate similarities/differences of the standard precipitation index (SPI) and the reconnaissance drought index (RDI), respectively, utilizing precipitation and ratio of precipitation over potential evapotranspiration (ET 0 ). Data are from stations with different climatic conditions in Iran. Drought characteristics of the 3-month, 6-month and annual SPI and RDI time series are developed and Markov chain order dependencies are investigated by the Log-likelihood, AIC and BIC tests. Steady state probabilities and Markov chain characteristics, i.e., expected residence time in different drought classes and time to reach "Near Normal" class are investigated. According to results, both indices exhibit an overall similar behaviour; particularly, they follow the first order Markov chain dependency. However, climatic variability may produce some differences. In several cases, the "Extremely Dry" class has received a more critical value by RDI. Furthermore, the expected residence time of "Near Normal" class and expected time to reach "Near Normal" class are quite different in a number of cases. The results show that the RDI by utilizing the ET 0 can be very sensitive to climatic variability. This is rather important, since if the drought analyses are for agricultural applications, utilization of the RDI would seem to serve a better purpose.

The AquaCrop model was used to simulate maize growth and soil water content under full and deficit irrigation managements as 1.2, 1, 0.8, and 0.6 of the potential crop water requirement. Generally, the RMSEs in simulating soil water content in calibration and validation were 0.01-0.039 and 0.012-0.037 m 3 m −3 , respectively, that overall corresponds to 3-14 % error. For the in-season biomass development, the RMSEs in calibration varied between 2.16 and 2.73 Mg ha −1 , while they varied between 1.97 and 5.19 Mg ha −1 in validation for the four irrigation managements. The model showed poor performance for simulating biomass late in the season under deficit irrigation managements. The RMSEs of final grain yield simulation were 0.71 and 1.77 Mg ha −1 that corresponded to 7 and 18 % error in calibration and validation, respectively. Likewise, the RMSEs for simulating the final biomass in calibration and validation were 1.29 and 2.21 Mg ha −1 that equals to 6 and 10 % error, respectively. Results demonstrated that AquaCrop is a useful decision-making tool for investigating deficit irrigations and maize growth in the region. However, in agreement with the findings in earlier studies on AquaCrop, the model showed insufficient accuracy in simulating final grain yield and biomass under moderate to severe water stresses. It is suggested that AquaCrop would benefit of including some calibrating parameters about the root distribution pattern in the soil because it is a water-driven model and highly depends on the accurately simulated water uptake from the soil profile.

No abstract

No abstract

No abstract