基于BP神经网络和FEFLOW模型模拟预测多年冻土活动层温度
郭林茂, 常娟, 徐洪亮, 叶仁政

Simulation and prediction of permafrost active layer temperature based on BP neural network and FEFLOW model: take the Fenghuoshan area on the Tibetan Plateau as an example
Linmao GUO, Juan CHANG, Hongliang XU, Renzheng YE
表2 训练样本数据的统计参数
Table 2 Statistical parameters of air temperature, surface temperature and total solar radiation training sample data
项目最小值(min)最大值(max)均值(mean)标准差(std)离散系数Cv偏态系数Cs
气温/℃-25.3012.70-4.768.25-1.73-0.08
地表温度/℃-20.3020.90-0.739.85-13.49-0.03
总辐射/(MJ·d-1·m-24.8137.6619.116.100.320.59