Assessing Aquacrop model for pearlmillet (Pennisetum glaucum) under in-situ water conservation in a rainfed semi-arid environment

Abstract

AquaCrop (v4. 0) model has been commonly used to simulate the crop yield under irrigated environments, but this model has not yet been calibrated and validated for simulating the yield of pearlmillet (Pennisetum glaucum (L.) R. Br.) under in-situ water conservation. Thus, a three year (2011−13) field experiment was conducted at ICAR-Indian Agricultural Research Institute, New Delhi with six in-situ water conservation treatments,viz. trench-cum-bund (TCB: 20 cm depth of trench, 20 cm height of bund), bund (30 cm height), ridge-and-furrow (R&F: 15 cm height), skip-row (SR: 3:1) planting, basin-tillage (BT: 45 cm × 45 cm) and control (no-water conservation). Calibration of AquaCrop (v 4. 0) model for pearlmillet grain yield and total soil moisture content was done using experimental data of 2011 and validated separately for 2012 (deficit rainfall year) and 2013 (excess rainfall year) data. For the year (2012), absolute prediction errors of grain yield were 1.7, 8.7, 14.1, 14.9, 3.3 and 7.3% for BT, R&F, TCB, bunds, SR and control, respectively, whereas Nash Sutcliffe Efficiency, root mean square error and coefficient of determination were 0.95, 0.07 and 0.96, respectively during calibration period and 0.73, 0.15 and 0.93 during validation period of deficit year. Thus, model predictions were satisfactoryfor less rainfall (<600 mm) year. Coefficient of determination R2 (0.4 to 1.0) was better for soil moisture simulation during dry-spell periods and for BT practice. The validated AquaCrop model can be used for prediction of pearlmillet yield and soil moisture under different water conservation practices in a semi-arid environment.