Determining a suitable “dry sowing and wet emergence” water control program for cotton fields in the arid regions of Northwest China is of great significance in saving water resources, improving economic efficiency, and promoting sustainable development of agriculture. The objective of this study was to analyze the effects of different “dry sowing and wet emergence” water control treatments on dry matter accumulation, chlorophyll fluorescence, yield quality, and water productivity of cotton, and to determine the optimal “dry sowing and wet emergence” water control program for cotton growth in arid areas. A two-year experiment was carried out in 2021 and 2022 in mulched drip-irrigated cotton fields, with a total of 13 treatments of different seedling water quantities (2021: 67.5 mm, 90 mm, 112.5 mm; 2022: 6 mm, 10.5 mm, 15 mm) and different drip frequencies (frequencies means number of drops at seedling stage) (2021: one, two, three times; 2022: two times, four times) in the “dry sowing and wet emergence”. Results indicated a positive correlation between increased seedling water quantity and growth indexes. High seedling water quantity treatment demonstrated a 14.33% higher cotton yield than the low seedling water quantity treatment. In comparison with low-frequency treatment, the high-frequency treatment exhibited significantly larger cotton plant height, dry matter accumulation, and yield. Over two years, the average values increased by 8.69%, 16.4%, and 15.91%, respectively, with a 14.55% increase in the coefficient of photochemical quenching of the leaf blade (qP). The high frequency and larger amount of seedling water quantity treatments showed significantly higher irrigation water productivity, with increases of 39.2% and 70.2% compared to the winter irrigation control treatment. In summary, the appropriate “dry sowing wet emergence” water regulation mode (the first drip: 15 mm, the second drip: 4.5 mm, the third drip: 22.5 mm, the fourth drip: 15 mm) can ensure crop yield quality under the premise of significantly reducing the agricultural irrigation water, which can provide certain theoretical support for the green, efficient, and sustainable development of the local cotton industry.

Soil amendments such as humic acid (HA), a biochar-based microbial agent (M), and vermicompost (V) can improve soil quality and promote crop growth. However, it remains unclear whether the co-application of the three soil amendments (HMV) has a synergistic effect on alleviating soil quality deterioration obstacles caused by dry sowing and wet emergence technology in Xinjiang cotton fields. A three-year field experiment was conducted in saline–alkali soils using plastic-film-mulched drip irrigation in Xinjiang, China. Through the orthogonal experiment method, the application amounts of HA, M and V were 75 kg ha⁻¹, 75 kg ha⁻¹ and 225 kg ha⁻¹ respectively in 2021. In 2022, three application amount gradients were used for HA, M and V: 60 kg ha⁻¹, 90 kg ha⁻¹ and 120 kg ha⁻¹ respectively. In 2023, the application amounts of HA, M, and V were 60 kg ha⁻¹, 120 kg ha⁻¹, and 120 kg ha⁻¹. It should be pointed out that V contains HA in the range of 20–35%. This study aimed to explore the improvement effect of a single or combined application of HA, M, and V on soil quality and cotton emergence rate using dry sowing and wet emergence technology in Xinjiang cotton fields. The results showed that the single and combined applications of HA, M, and V improved the soil quality and water–heat–salt environment of the cultivated layer. In the combined application, the cotton seedling emergence rate and yield increased by 1.9–22.8% and 7.0–54.1%. Therefore, it is recommended to jointly apply HA, M, and V to promote cotton seedling emergence and increase yield using dry sowing and wet emergence technology in Xinjiang cotton fields.

In order to investigate the changes in winter wheat yield and the factors influencing it, five meteorological factors—such as rainfall and soil moisture—collected from the experimental area between 2010 and 2022 were used as characteristic features. A combined model of GNN (Graph Neural Network), based on the Fourier transform and the Random Forest algorithm was developed to predict winter wheat yield. Matrix multiplication in Fourier space was performed to predict yield, while the Random Forest algorithm was employed to quantify the contribution of various yield factors to winter wheat yield. The combined model effectively captured the dynamic dependencies between yield factors and time series, improving predictive accuracy by 5.00%, 10.00%, and 27.00%, and reducing the root mean square error by 26.26%, 29.31%, and 88.20%, respectively, compared to the StemGNN, Informer, and Random Forest models. The predicted outputs ranged from 520 to 720 g/m², with an average error of 2.69% compared to the actual measure outputs. Under the insufficient real-time irrigation mode, winter wheat yield was highest at 90% irrigation upper limit and 70% irrigation lower limit, with a medium fertilization level (850 mg/kg). The yield showed an overall decreasing trend as both irrigation limits and fertilizer application decreased. Rainfall and soil moisture were the most significant factors influencing winter wheat yield, followed by air temperature and evapotranspiration. Solar radiation and sunshine duration had the least impact. The results of this study provide a valuable reference for accurately predicting winter wheat yield.