摘要
选用2021年12月至2022年3月天津市两类新型保暖式钢骨架大棚棚室内、外气象观测数据,应用随机森林法模拟棚室内气温分布,分析新型棚室不同气温年型下的热量资源变化和蔬菜种植的气温适宜度及低温风险状况评估。结果表明:两类保暖式钢骨架大棚棚室内气温的随机森林模型模拟值与实测值的决定系数R^(2)大于0.840,一致性指标D值大于0.980。加温型保暖式钢骨架大棚全年最低气温可维持在10℃以上,变化范围较小,适宜种植半耐寒蔬菜及喜温蔬菜(适宜度分别为0.86~0.97和0.73~0.90,无低温冷害风险)。常规型保暖式钢骨架大棚最低气温可维持在-3℃以上,适宜种植耐寒蔬菜及半耐寒蔬菜(适宜度分别为0.59~0.91和0.43~0.86,低温风险较低)。
Using the indoor and outdoor meteorological observation data from two novel insulated steel frame greenhouses in Tianjin over the period from December of 2021 to March of 2022,we applied the Random Forest(RF)method to simulate the distribution of temperature within the two greenhouses,analyzed the variation in heat resources under different climatic conditions and the temperature suitability of vegetable growing,and assessed the risk of low temperature.The results indicated that the indoor air temperature simulated by the RF method correlates highly with the measured values,with a coefficient of determination exceeding 0.85 and a consistency index D value above 0.988 in both novel insulated steel frame greenhouses.The minimum temperature within the heated insulated steel frame greenhouses is consistently maintained above 10℃,with minimal fluctuations,making it ideal for the cultivation of semi-tolerant and warm vegetables(with a suitability degree ranging from 0.86 to 0.97 and 0.73 to 0.90,respectively,and no risk of low temperature injury).Conventional insulated steel frame greenhouses can maintain a minimum temperature above-3℃,suitable for cold-hardy and half-cold-hardy vegetables(with a suitability degree ranging from 0.59 to 0.91 and 0.43 to 0.86,respectively,and a low risk of low temperature injury).
作者
刘涛
董朝阳
王腾歌
谭方颖
柳芳
麻雪艳
黎贞发
LIU Tao;DONG Chaoyang;WANG Tengge;TAN Fangying;LIU Fang;MA Xueyan;LI Zhenfa(Tianjin Climate Centre,Tianjin 300074,China;Tianjin Agricultural University,Tianjin 300384,China;National Meteorological Centre,Beijing 100081,China)
出处
《气象与环境学报》
2024年第4期131-137,共7页
Journal of Meteorology and Environment
基金
天津市气象局一般项目(202218dgxm01)
天津市蔬菜现代产业技术体系创新团队科研专项(ITTVRS2021017)
国家自然科学基金项目(31901398)共同资助。
关键词
随机森林
神经网络
越冬茬
热量资源
Random Forest(RF)
Neural network
Overwintering crop
Heat resources
