Performance of AquaCrop and DSSAT Models for Greenhouse Grown Tomato Under High-Temperature Conditions

High temperature is a primary meteorological hazard restricting crop yield and quality, particularly in semi-automatic greenhouses. However, the accuracy of crop models under controlled high-temperature conditions in greenhouses has rarely been examined. To address this, our study evaluated the AquaCrop (7.1) and DSSAT (4.8.2) models for greenhouse-grown tomato under high-temperature conditions. Field experiments were conducted over two consecutive years (2024 and 2025) with two temperature treatments (TH: daily maximum temperature > 35 °C; TD: daily maximum temperature < 35 °C) and two irrigation levels (WH: 100%Epan; WD: 60%Epan, where Epan is cumulative pan evaporation). Key indicators were continuously measured, including canopy cover, soil water content, above-ground biomass, yield, and water consumption. Additionally, the XGBoost algorithm was introduced to estimate reference crop evapotranspiration across different treatment combinations, facilitating a multi-scenario analysis. The results indicated that under both TD and TH conditions, the AquaCrop model accurately simulated canopy cover (RMSE ≤ 6.84%, NRMSE ≤ 12.27%, EF ≥ 0.92), yield (PE ≤ 12.20%), above-ground biomass (PE ≤ 2.34%), and water consumption (PE ≤ 10.33%) across different irrigation levels. However, its performance in simulating soil water content declined markedly under the TH treatment, with EF becoming negative in some cases (−1.66 ≤ EF ≤ 0.43), indicating that the model was largely ineffective for SWC under these conditions. The DSSAT model exhibited lower simulation accuracy under the TH treatment compared to the TD treatment, and its overall performance was slightly inferior to that of the AquaCrop model (RMSE ≤ 15.79 mm, NRMSE ≤ 14.54%, −1.68 ≤ EF ≤ 0.46). Scenario analysis of 20 temperature and irrigation combinations, integrating XGBoost-derived ET0 into the AquaCrop model, revealed that maximizing greenhouse tomato yield requires a combination strategy of 1.1Epan and 34 °C, whereas optimizing water use efficiency is best achieved under 0.9Epan and 32 °C. These findings provide a theoretical basis and technical support for the application of crop models and the development of environmental regulation strategies in greenhouse cultivation under high-temperature conditions.