THE PREDICTION OF METEOROLOGICAL AND PHENOLOGICAL TRENDS AND ROUTE DECISION-MAKING DURING THE QINGMING FESTIVAL BASED ON ARIMA AND RANDOM FOREST REGRESSION
Keywords:
ARIMA model, Random Forest, Multi-objective optimizationAbstract
Addressing the significant challenges in cultural tourism decision-making caused by unpredictable weather and uncertain flowering periods during the Qingming Festival, this study establishes a robust quantitative framework integrating meteorological forecasting, phenological prediction, and route optimization. Specifically, an ARIMA model is constructed to define “continuous drizzle” quantitatively and forecast precipitation probabilities for seven target cities. Concurrently, a Random Forest Regression model is introduced to accurately predict the onset and duration of flowering for cherry blossoms, peonies, and rapeseed flowers. Feature importance analysis reveals temperature as the dominant driver for cherry blossoms (weight 62.3%), with the model achieving a high goodness-of-fit of 0.91. Furthermore, a multi-objective planning model is developed using integer linear programming, generating an optimal closed-loop tour route connecting Wuhan, Xi’an, Luoyang, and Hangzhou. This route effectively balances flower-viewing experiences with rain-avoidance preferences while minimizing travel costs. The research innovates by establishing quantitative meteorological criteria and successfully transforming natural law predictions into personalized tourism decision-making strategies, providing valuable theoretical support and practical guidelines for smart cultural tourism in complex environments.References
[1] Cohen L A, Bradford B, Groves R, et al. Interactions among weather and landscape affect Colorado potato beetle population dynamics. PloS one, 2026, 21(3): e0345180.
[2] Dantas G, Browell J. Seamless Short‐ to Mid‐Term Probabilistic Wind Power Forecasting. Wind Energy, 2025, 29(2): e70079.
[3] Kim J, Hu Y, Khorasani E N, et al. Assessment of deep learning models integrated with weather and environmental variables for wildfire spread prediction and a case study of the 2023 Maui fires. Natural Hazards, 2025, 122(1): 10.
[4] Liu X, Zhu Z, Chen S, et al. Do the S2S Models Have Prediction Skills beyond the Weather Timescale for Winter Snowfall over Eastern China?. Advances in Atmospheric Sciences, 2025, 43(4): 874-888.
[5] Janani K, Alageswaran R, Amirtharajan R. Dynamic SG-SKRDX hybrid framework for precision weather forecasting and crop suitability in the Cauvery Delta. Scientific reports, 2025, 16(1): 2042.
[6] Medina B J, Sengupta A, Steinhoff D, et al. A regional high resolution AI weather model for the prediction of atmospheric rivers and extreme precipitation. npj Climate and Atmospheric Science, 2025, 8(1): 385.
[7] Kasagani R K S V D, Manickam P. Modeling of solar photovoltaic power using a two-stage forecasting system with operation and weather parameters. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 2025, 47(2).
[8] Gerg D I, Tashie M A, Patanaik A, et al. The substantial role of weather data in consumer spending prediction: A robust machine learning assessment. Journal of Retailing and Consumer Services, 2026, 90: 104649.
[9] Amirioun H M, Tabatabaei S S, Asgari A. A learning-based outage prediction method for resilient electricity distribution systems in response to extreme weather events. Computers and Electrical Engineering, 2026, 130: 110881.
[10] Witze A. This AI model ‘studied’ physics — and learnt to forecast extreme weather. Nature, 2025.