Agro-pastoral ecotone is an important livestock production area in the north of China, and alfalfa is the main pasture crop in this area. Aiming to address the issues of groundwater overexploitation in the area with water demand, we assessed the consumption pattern, irrigation scheduling, and water usage efficiency of alfalfa under subsurface drip irrigation. Alfalfa was used as the research object in this study. A DSSAT model was used to simulate the soil moisture, yield, and other alfalfa grow characteristics during a two-year in situ observation study and provide information on the best irrigation techniques and the water-use efficiency of alfalfa in the agro-pastoral ecotone of Northwestern China. The results showed that the ARE, nRMSE, and R² values of the alfalfa soil water content, leaf area index, and yield varied between 3.82% and 5.57%, 4.81% and 8.06%, and 0.86 and 0.93, respectively, the accuracy of the calibrated and validated parameters were acceptable, and the model could be applied to this study. The water consumption of alfalfa ranged from 395.6 mm to 421.8 mm during the whole year, and the critical water consumption period was the branching stage and the bud stage. During the branching stage and the bud stage, water consumption was 30–31% and 31–33% of the total water consumption, and the water consumption intensity averaged 2.97–3.04 mm/d and 4.23–4.97 mm/d. The variations of WUE and IWUE were 11.74–14.39 kg·m⁻³ and 7.12–9.31 kg·m⁻³. Irrigation increased the water productivity of rain-fed alfalfa by 49.48–64.70% and increased the yield of alfalfa by 17.87–34.72%. With the highest yield as the goal, the recommended irrigation volumes for normal and dry flow years were 200 mm and 240 mm; with the goal of the highest utilization of groundwater resources, the recommended irrigation volumes for normal and dry flow years were 160 mm and 192 mm. The results of this study are expected to provide scientific and technological support for the rational utilization of groundwater and the scientific improvement of alfalfa yields in the agro-pastoral ecotone of Northwestern China.

Vegetation productivity, as an essential global carbon sink, directly influences the variety and stability of ecosystems. Precise vegetation productivity monitoring and forecasting are crucial for the global carbon cycle. Traditional machine learning algorithms frequently experience overfitting when processing high-dimensional time-series data or substantial numbers of outliers, impeding the accurate prediction of various vegetation metrics. We propose a multimodal regression prediction model utilizing the TCLA framework—comprising the Transient Trigonometric Harris Hawks Optimizer (TTHHO), Convolutional Neural Networks (CNN), Least Squares Support Vector Machine (LSSVM), and Adaptive Bandwidth Kernel Density Estimation (ABKDE)—with the Hetao Irrigation District, a vast irrigation basin in China, serving as the study area. This model employs TTHHO to effectively navigate the search space and adaptively optimize network node positions, integrates CNN-LSSVM for feature extraction and regression analysis, and incorporates ABKDE for probability density function estimation and outlier detection, resulting in accurate interval probability prediction and enhanced model resilience to interference. Experimental data indicate that the TCLA model improves prediction accuracy by 10.57–26.47% compared to conventional models (Long Short-Term Memory (LSTM), Transformer). In the presence of 5–15% outliers, the fusion of multimodal data results in a substantial drop in RMSE (p < 0.05), with a reduction of 45.18–69.66%, yielding values between 0.079 and 0.137, thereby demonstrating the model’s high robustness and resistance to interference in predicting the next three years. This work introduces a scientific approach for precisely forecasting alterations in regional vegetation productivity using the proposed multimodal TCLA model, significantly enhancing global vegetation resource management and ecological conservation techniques.

The shifting irrigation reduction in the Hetao Irrigation District and the inability to effectively discharge salts from the system have led to significant changes in salt migration patterns. Based on the integration of long-term field observations (2017–2023) with soil hydrodynamics and solute transport models, this study explored the impact of irrigation reduction on water and salt migration in a cropland–wasteland–lake system. The results indicated that before and after the reduction in irrigation and decline in groundwater levels, the migration rates of groundwater from croplands to wastelands and from wastelands to lakes remained relatively stable, averaging 78% and 40%. During the crop growth period, after irrigation reduction and groundwater level decline, the volume of groundwater recharging lakes from wastelands decreased by 80–120 mm, causing a water deficit in the lakes of 679–789 mm. After irrigation reduction and groundwater level decline, during the crop growth period, 1402 kg/ha of salt remained in the wasteland groundwater, and 597–861 kg/ha of salt accumulated in the cropland groundwater, exceeding previous levels, leading to salinization in the cropland and wasteland groundwater. This study provides insights relevant to managing groundwater and soil salinity in irrigation areas.

The agro-pastoral ecotone in Yinshanbeilu is the main potato-producing region. In recent years, the shift from rainfed to irrigated agriculture has created challenges in understanding potato water consumption patterns, water use efficiency, and irrigation optimization. This study utilized the DSSAT model to simulate soil moisture, leaf area index, and potato yield based on a two-year in situ observational experiment. The study showed that simulated values of the soil water moisture, leaf area index, and yield, with Absolute Relative Error (ARE) of 4.18–5.27%, Normalized Root Mean Square Error (nRMSE) of 5.64–8.65%, and Coefficient of Determination (R²) values of 0.86–0.921, exhibited acceptable accuracy. Simulated results pointed out that potato water consumption ranged between 375.2 and 414.2 mm, with 50–52% occurring during tuber formation to bulking stages, and the average water consumption intensity was 2.62~2.81 mm/d. Based on DSSAT model simulation, this study found that water use efficiency (WUE) reached 162.17–166.20 kg/(hm²·mm), while irrigation water use efficiency (IWUE) varied between 86.1 and 108.1 kg/(hm²·mm). With the highest yield as the target, the recommended irrigation amounts for potato in normal year and dry year were 180 mm and 240 mm. With the highest utilization rate of groundwater resources as the target, the recommended irrigation amounts in normal year and dry year were 162 mm and 192 mm. These findings offer valuable insights for promoting sustainable groundwater use and enhancing water conservation practices in the Yinshanbeilu agro-pastoral ecotone.