Background: Water shortages limit agricultural production in arid and semiarid regions around the world. The accurate estimation of reference evapotranspiration (ET₀) is of the utmost importance for computing crop water requirements, agricultural water management, and irrigation scheduling design. However, due to the combination of insufficient meteorological data and uncertain inputs, the accuracy and stability of the ET₀ prediction model were different to varying degrees. Methods: Six machine learning models were proposed in the current study for daily ET₀ estimation. Information on the weather, such as the maximum and minimum air temperatures, solar radiation, relative humidity, and wind speed, during the period 1960~2019 was obtained from eighteen stations in the northeast of Inner Mongolia, China. Three input combinations were utilized to train and test the proposed models and compared with the corresponding empirical equations, including two temperature-based, three radiation-based, and two humidity-based empirical equations. To evaluate the ET₀ estimation models, two strategies were used: (1) in each weather station, we trained and tested the proposed machine learning model, and then compared it with the empirical equations, and (2) using the K-means algorithm, all weather stations were sorted into three groups based on their average climatic features. Then, each station tested the machine learning model trained using the other stations within the group. Three statistical indicators, namely, determination coefficient (R²), root mean square error (RMSE), and mean absolute error (MAE), were used to evaluate the performance of the models. Results: The results showed the following: (1) The temperature-based temporal convolutional neural network (TCN) model outperformed the empirical equations in the first strategy, as shown by the TCN’s R² values being 0.091, 0.050, and 0.061 higher than those of empirical equations; the RMSE of the TCN being significantly lower than that of empirical equations by 0.224, 0.135, and 0.159 mm/d; and the MAE of the TCN being significantly lower than that of empirical equations by 0.208, 0.151, and 0.097 mm/d, and compared with the temperature-based empirical equations, the TCN model markedly reduced RMSE and MAE while increasing R² in the second strategy. (2) In comparison to the radiation-based empirical equations, all machine learning models reduced RMSE and MAE, while significantly increasing R² in both strategies, particularly the TCN model. (3) In addition, in both strategies, all machine learning models, particularly the TCN model, enhanced R² and reduced RMSE and MAE significantly when compared to humidity-based empirical equations. Conclusions: When the radiation or humidity characteristics were added to the given temperature characteristics, all the proposed machine learning models could estimate ET₀, and its accuracy was higher than the calibrated empirical equations external to the training study area, which makes it possible to develop an ET₀ estimation model for cross-station data with similar meteorological characteristics to obtain a satisfactory ET₀ estimation for the target station.

The sustainable management of Northern China’s vulnerable agro-pastoral ecotone requires a clearer understanding of how tillage systems affect crop productivity through local soil-climate interactions. Therefore, this study was conducted to quantify and compare the long term effects of different tillage practices on soil hydrothermal regimes, resource use efficiency, and maize yield stability in a semi-arid agroecosystem. A long term five-year field experiment with maize was conducted in this ecotone to assess three tillage methods: no tillage (NT), deep ploughing (DP), and conventional rotary tillage (RT). Seasonal monitoring included soil moisture, temperature, bulk density, and straw cover. Analyses focused on soil water use efficiency (WUE), the production efficiency per soil thermal unit (PEsoil), and pathways affecting theoretical calculated yield. Results show that relative to RT and DP, NT consistently elevated soil water content within the 0–30 cm profile during the growing season, with the most marked increases from pre-sowing to the V12 stage. This water-conserving effect was stronger in wet years, highlighting the role of precipitation in NT’s performance. DP also retained more soil water than RT, particularly in deeper layers, though its effect was less pronounced than NT’s. Regarding temperature, NT lowered the daily mean soil temperature and accumulated growing degree days (GDD) in early growth phases, a result of residue cover buffering thermal changes. Despite reduced heat accumulation, NT achieved the greatest efficiencies for both heat and water use (PEsoil and WUE), showing increases of 62.03% and 16.64% over RT, respectively, without yield penalty. Key mechanisms include permanent straw mulch under NT, which curtails evaporation, promotes water infiltration, and stabilizes soil structure, thereby modulating hydrothermal dynamics. Structural equation modeling indicated that soil water content, ear number per hectare, and hundred-kernel weight directly and positively determined final yield. Tillage methods exerted indirect effects on yield by modifying soil physical traits and microclimatic conditions. In this semi-arid setting, both NT and DP outperformed RT in conserving soil water, moderating soil temperature, and boosting resource use efficiency. These practices present viable strategies for strengthening crop resilience and sustaining productivity amid climatic variability.

In arid environments, plant communities often cooperate to fight the harsh conditions. In order to study the interaction between plant populations and explore their survival under drought conditions, an improved Lotka-Volterra model was established by means of mechanism analysis to consider the interaction between different plant species, and the species correlation degree was added. Light, water, air competition among plants and other important parameters affecting plant growth, a relatively complete model of plant drought adaptability was fitted. Then, the fourth order Runge Kutta method was used to solve the model numerically, and the relationship between the amount of various plants in the community with time was obtained, and the relationship between plant populations under drought conditions was explored.