Abstract:

In this study, crab yield and meteorological data from Xinghua were collected for the period 2010−2020, and the corresponding meteorological yield was separated using double exponential smoothing, linear regression, and the Hodrick−Prescott (HP) filter. The key meteorological factors affecting meteorological yield were identified by fitting the relationships between meteorological yield and meteorological factors at three growth stages (the first to third molting stages, the fourth molting stage and the fifth molting stage). Finally, a meteorological yield model based on key meteorological factors was constructed using the multiple linear regression method, and validated with yield data from 2021 to 2023. The results showed that: (1) the HP filter method demonstrated the optimal performance both in extracting the trend yield of Xinghua crab and in stably reflecting its meteorological yield. (2) The key meteorological factors affecting the meteorological yield of Xinghua crab included the number of high−temperature days and precipitation days during the first to third molting stages, the vapor pressure during the fourth molting stage, and the relative humidity, diurnal temperature range and minimum air pressure during the fifth molting stage. (3) The model based on the HP filter method (KMF−H) demonstrated the best simulation performance, with a coefficient of determination (R²) of 0.91, a root mean square error (RMSE) of 50.01kg·ha⁻¹, and the highest validation accuracy (D) of 97.14%. Overall, the model constructed using the HP filter method is suitable for simulating the yield of Xinghua crab.

Key words: Xinghua crab, Meteorological factors, Yield separation, Double exponential smoothing, Linear regression, HP filter