Abstract
The over‐exploitation of groundwater in agriculture necessitates the adoption of water‐efficient practices. A field study (2020–2022) under cotton–wheat cropping in semi‐arid Punjab, India, evaluated subsurface drip irrigation (SSDI) with lateral depths (20, 25 and 30 cm), emitter spacings (20, 30 and 40 cm) and irrigation levels (60%, 80% and 100% ETc). Bayesian optimization was used for multicriteria decision‐making. The results revealed wheat sensitivity to lateral depth, with a significant yield reduction at 30 cm compared to 20 cm, whereas the deep root system of cotton mitigated such effects. Overall, SSDI at 80% ETc maintained system productivity (SP; 4343 kg ha −1 ) and net returns (NR; 2001.1 $ ha −1 ) comparable to full irrigation, in addition to 20% higher irrigation water productivity (WP). Additionally, it also resulted in 5%, 11%, 5% and 23% greater SP, NR, irrigation and net economic WP, respectively, than the recommended surface drip practice. Furthermore, multiresponse optimization identified 25.5‐cm lateral depth, 31.2‐cm emitter spacing and irrigation at 81% ETc as the optimum combination for SSDI. This configuration optimized productivity, water saving and profitability within permissible drainage losses (RMSE 0.4223). These findings not only underscore the resource conservation potential of SSDI but also demonstrate its capacity to maximize economic and environmental benefits.