نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشجوی کارشناسی ارشد آبیاری و زهکشی، گروه علوم و مهندسی آب، واحد اهواز، دانشگاه آزاد اسلامی، اهواز، ایران.
2 استادیار، گروه علوم و مهندسی آب، واحد اهواز، دانشگاه آزاد اسلامی، اهواز، ایران
3 دکتری آبیاری و زهکشی، سازمان آب و برق استان خوزستان، اهواز، ایران.
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Sensitivity analysis is the most important step before calibrating crop models. It helps researchers to have enough information about the effectiveness of each parameter, and changes them during calibration stage. This issue is more important due to the increasing use of AquaCrop model for crop simulation. Therefore, in the study, the sensitivity of AquaCrop to change some crop growth parameters; normalized water productivity (WP *), maximum transpiration coefficient (KCTrx), initial canopy cover (CC0), canopy growth coefficient (CGC), canopy decline coefficient (CDC) and harvest index (HI) were assessed using Beven (1979) method. For this purpose, the data collected in a research farm in Ahvaz during 2014 were used. The studied treatments include irrigation method (D: sprinkler irrigation using saline water, F: sprinkler irrigation using both saline and fresh water and S: furrow irrigation using saline water) with five irrigation water qualities (S1: 2.5, S2: 2.3, S3 : 3.9, S4: 4.6 and S5: 1.5 dS m-1). The results showed that the highest sensitivity was to changes in normalized water productivity (0.95≤Spi≤1.04) and maximum transpiration coefficient (0.95≤Spi≤0.67). After that, the sensitivity of harvest index (0.51≤Spi≤0.56) was in the middle category. The range of yield changes in different values of normalized water productivity, maximum transpiration coefficient, harvest index and canopy decline coefficient were 1.3-3.3, 0.8-1.6, 0.6-1.16 and 0.32-0.64 ton ha-1, respectively. Sensitivity coefficients were positive for all parameters except CDC. Therefore, by increasing the CDC, AquaCrop suffers from underestimation error and by increasing the value of other parameters; the model suffers from overestimation error. Therefore, in situations where the difference between observed and simulated yield is large, it is better to consider the parameters of normalized water productivity and maximum transpiration coefficient. Otherwise, the parameters of harvest index and canopy decline coefficient should be considered.
کلیدواژهها [English]