Accurate estimation of crop evapotranspiration (ET_c) is crucial for effective irrigation and water management. To achieve this, support vector regression (SVR) was applied to estimate the daily ET_c of spring maize. Random forest (RF) as a data pre-processing technique was utilized to determine the optimal input variables for the SVR model. Particle swarm optimization (PSO) was employed to optimize the SVR model. This study used data obtained from field experiments conducted between 2017 and 2019, including crop coefficient and daily meteorological data. The performance of the innovative hybrid RF–SVR–PSO model was evaluated against a standalone SVR model, a back-propagation neural network (BPNN) model and a RF model, using different input meteorological variables. The ET_c values were calculated using the Penman–Monteith equation, which is recommended by the FAO, and used as a reference for the models’ estimated values. The results showed that the hybrid RF–SVR–PSO model performed better than all three standalone models for ET_c estimation of spring maize. The Nash–Sutcliffe efficiency coefficient (NSE), root mean square error (RMSE), mean absolute error (MAE) and coefficient of determination (R²) ranges were 0.956–0.958, 0.275–0.282 mm d⁻¹, 0.221–0.231 mm d⁻¹ and 0.957–0.961, respectively. It is proved that the hybrid RF–SVR–PSO model is appropriate for estimation of daily spring maize ET_c in semi-arid regions.

Traditional detection methods for soil nitrate nitrogen (NO₃^--N), a critical nutrient for crop growth, suffer from poor timeliness and susceptibility to matrix interference. To address these issues, this study presents a novel dual-band frequency-splitting coupled sensor with a concentric copper ring structure for in situ rapid detection of soil NO₃^--N via dielectric spectroscopy. The key technological innovation lies in using the low-frequency band (1-50 MHz) to isolate water and salinity interferences via stable impedance matching of the copper rings and capturing the characteristic NO₃^--N relaxation signals in the high-frequency band (100-500 MHz) via enhanced electromagnetic coupling. Field trials across five soil types (brown, black, red, saline-alkali, and loess) demonstrated excellent performance, with determination coefficient (R²) values of 0.943-0.987, mean absolute error values of ≤0.75 mg/kg, root-mean-square error values of ≤0.92 mg/kg, and millisecond-level response. Signal drift remained <0.25 mg/kg even under extreme conditions (-5 °C, 90% relative humidity (RH)), with Pearson correlation coefficient values of 0.995-0.999 in typical agricultural scenarios (pre/postfertilization and precipitation). The developed sensor eliminates the need for sampling and pretreatment, reducing the detection time from 3-5 days using traditional methods to milliseconds, and provides high-precision data for dynamic nitrogen management. Moreover, the IoT integration potential of the sensor advances smart agriculture and sustainable development.

Although the tasseling stage is a pivotal window for drought stress in maize, the photosynthetic physiological mechanism by which straw mulching compensates for drought-induced impairment relative to conventional rain-fed practice remains unclear. In this study, a two-factor randomized block experiment was conducted with two straw mulching levels (full mulching, FM, 9000 kg ha⁻¹; half mulching, HM, 4500 kg ha⁻¹) and three drought gradients (mild, 60–70% FC; moderate, 50–60% FC; severe, 40–50% FC) during the tasseling stage, with the conventional rain-fed control (CK) being non-mulched and well-watered (80% FC). The results show that only FM under mild drought (FM + LD) maintained the net photosynthetic rate, PSII photochemical efficiency (Fv/Fm), and grain yield at levels not significantly different from the CK (p > 0.05). Under mild drought, photosynthetic inhibition was dominated by stomatal limitation, with FM maintaining effective stomatal regulation. As drought intensified, the limiting factor shifted to non-stomatal limitation; however, FM showed significantly higher Fv/Fm and lower Vj than HM, indicating better PSII preservation. At 24 h after rewatering, donor-side function largely recovered across all treatments, whereas acceptor-side parameters remained elevated except in T1 (FM + LD). Overall, FM under mild drought at the tasseling stage achieved photosynthetic performance and yield comparable to conventional rain-fed management, providing a theoretical basis for water-saving cultivation in semi-arid regions.