This article utilized hourly temperature, humidity, pressure and wind speed data from passion fruit meteorological observation stations in three southwestern cities of Fujian Province (Longyan, Sanming, Zhangzhou) from 2020 to 2022, as well as national ground conventional meteorological observation stations. BP neural network and stepwise regression method were applied to construct temperature prediction models for the passion fruit planting bases. The results showed that: (1) The simulation effect of the passion fruit station temperature prediction model based on BP neural network (referred to as BP model) was better than that of the model based on stepwise regression method (referred to as regression model). The average absolute error (MSE) of BP model (2.75–3.42 °C) was smaller than that of regression model (3.32–3.94 °C). (2) For the simulation results of daily temperature changes in the passion fruit station, the difference in hourly average temperature between the BP model predictions (regression model predictions) and observed temperatures at passion fruit station was −4.1–4.4 °C (−6.0–10.2 °C). The BP model showed a daily temperature trend that was closer to the measured values; (3) For the simulation results of high and low temperatures in the passion fruit station, the BP neural network model (regression model) showed a prediction error range of −5.6 °C to 5.2 °C compared to observed temperatures, while the stepwise regression model’s error range was −4.1 °C to 8.8 °C. The BP model’s predicted temperature trend was closer to the measured values. (4) Both models have significant shortcomings in the prediction of high-temperature individual cases and hourly averages, with relatively large errors (generally exceeding 3 °C), especially during the period from 10 to 16 o’clock. The future version needs to be optimized.

This study assesses the efficacy of the ZJWARMS model’s AI-based post-processing correction method for temperature and wind speed forecasts in complex terrain. By analyzing 72 h forecasts at four stations with varying elevations (from 273 m to 1327 m) in the Liuchun Lake region during December 2021–December 2022, the study found that AI-based corrections substantially enhanced both forecast accuracy and stability. The results indicate that, after correction, temperature forecast accuracy at all stations exceeded 99%, with the most notable relative gains at higher elevations (up to 48.1%). The mean absolute error (MAE) for temperature declined from 3.08 °C to below 0.8 °C at Octagonal Palace, and from 3.29 °C to below 0.6 °C at Mountaintop. Wind speed forecast accuracy also increased from approximately 60–70% to nearly 100%, with MAE generally constrained to the range of 0.2–0.4 m/s. In terms of extreme error control, the number of samples with temperature errors exceeding ±2 °C was markedly reduced. For instance, at Mountainside, the count dropped from 127 to 0. Extreme wind speed errors were also effectively eliminated. After correction, error distributions became more concentrated, and both temporal stability and spatial consistency showed notable improvement. These gains enhance operational forecasting and risk management in mountainous regions, for example, through threshold-based wind-hazard alerts and support for mountain-road icing, by providing more reliable, high-confidence guidance.

Liuchun Lake area, located in the high-elevation and topographically complex western region of Zhejiang Province, exhibits temperature variability strongly influenced by terrain-induced dynamics and local microclimates. The Zhejiang Operational Consensus Forecasts (ZJOCF) model shows pronounced systematic biases in this area, making it difficult to meet the demand for short-term, fine-scale forecasts in cultural-tourism applications. Using observational data from four stations at different elevations, this study analyzes how ZJOCF temperature forecast errors vary with altitude, develops a station-level machine-learning temperature bias-correction model, and evaluates its performance in terms of accuracy, mean absolute error (MAE), error distribution, and control of extreme errors. Results show that the accuracy of the raw forecasts decreases significantly with increasing elevation, with high-altitude sites exhibiting distinct warm biases and strong fluctuations. After correction, the 72 h forecast accuracy at the four stations increases to 69–71% (up to 40.8% at the mountaintop station), MAE is reduced by more than 60% on average, extreme-error cases decrease by 40–60%, and the error distribution shifts from a scattered multi-peak pattern to a concentrated single-peak structure. These findings demonstrate that station-level machine-learning correction can effectively mitigate structural errors in ZJOCF temperature forecasts over complex terrain, providing a reliable technical pathway for refined meteorological services in mountainous regions.