Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Belgique → Performance Evaluation of the AquaCrop Runoff Module for Semi-Arid Climate

KU Leuven (2014)

Performance Evaluation of the AquaCrop Runoff Module for Semi-Arid Climate

Jayag, Omar

Titre : Performance Evaluation of the AquaCrop Runoff Module for Semi-Arid Climate

Auteur : Jayag, Omar

Université de soutenance : KU Leuven

Grade : Master of Water Resources Engineering 2014


The amount of rainfall lost by surface runoff has a big impact on crop production in rainfed agriculture, particularly in semi-arid areas. Consequently, in the crop water productivity model AquaCrop, accurate runoff estimation is very crucial as it determines the amount of water that can infiltrate and will be stored in the soil profile, and will be available for crop growth and production. In addition, accurate runoff simulation is necessary for a realistic simulation of the effect of field management practices that affect runoff and may increase crop water productivity. In this study, the runoff modeling procedure of AquaCrop that is based on the curve number (CN) method was extensively evaluated. The evaluation was carried out by comparing the runoff algorithms adopted in AquaCrop with the recent developments of the CN method in literature. Furthermore, the AquaCrop runoff modeling procedure was tested based on experimental data from runoff field experiments in Tigray, Ethiopia. These runoff experiments with wheat (2011) and barley (2012) were carried out on fields with a clay and clay loam soil with slopes ranging between 5 to 16%. Recent literature shows that an initial abstraction (Ia) value of 0.05 (and corresponding adjustment of the CN value) would be more appropriate than the current value of 0.2. Moreover, the slope adaptation of the CN value has been used in different models like SWAT and SWRRB and could be advised. The comparison of simulation results with runoff data from Taye et al. (2013) showed that the current AquaCrop version has a poor model performance (RRMSE between 50 to 123% and NSE ranging between 0.26 to 0.46). An adaptation of the Ia value equal to 0.05 improved the model performance, while the slope adaptation of CN didn’t have a consistent improvement of the runoff simulation due to the increasing rock fragment cover which increased the infiltration at higher land slopes. Although some improvements were obtained, the overall model performance with the improvement scenario of Ia equal to 0.05S was still poor (RRMSE between 40 to 115% and NSE between -0.06 to 0.66. The results also showed that the simulated runoff remained underestimated no matter what improvement scenarios were carried out. For that reason, a corrected CN value in correlation to high rock fragment cover and soil texture might improve the simulation.


Page publiée le 2 novembre 2015, mise à jour le 10 décembre 2018