Assessment of crop water stress under prevailing irrigation practice in Faisalabad, Pakistan using hyperspectral remote sensing

Inefficient irrigation under prevailing farmer practices induces crop water stress, reducing yield in semi-arid agricultural systems. This study, conducted from August 2019 to January 2020 at Faisalabad, Pakistan, compared water stress in maize and spinach grown under deficit-excess irrigation and prevailing irrigation practice (PIP) using hyperspectral remote sensing (325–1075 nm). A randomized complete block design was used with five irrigation levels (0, 33, 66, 100, and 133% of field capacity) across four growth stages, and soil moisture ranged from 7% to 24% across these levels. Thirty spectral vegetation indices (SVIs) of water, structural, pigment, and red-edge values were calculated and related to soil moisture and crop yield using ANOVA, Tukey post hoc tests, and linear regression. Water stress altered spectral reflectance, particularly in the Near-Infrared region. SVIs showed strong stage-related correlations with soil moisture (R2 up to 0.67 in maize and 0.89 in spinach), with spinach exhibiting greater spectral sensitivity. The normalized photochemical reflectance index (PRInorm) alone showed strong correlations with soil moisture across most growth stages in both crops, confirming its reliability for stress detection. Maize reached its highest yield (1320 g) at 100% FC and spinach at 66% FC (1581 g); PIP reduced yields by 31.6% and 26.6%, respectively (P < 0.001). Spectral analysis revealed alternating mild-to-moderate under-irrigation and mild-to-severe over-irrigation stress under PIP throughout the crop cycle. A stage-specific irrigation program was proposed based on optimum field capacity thresholds. Hyperspectral spectroscopy offers a feasible platform for designing accurate irrigation strategies to improve water productivity under water scarcity.