深度学习模型在2021年汛期武汉市雷达回波临近预报中的应用评估

袁凯, 庞晶, 李武阶, 李明

Application evaluation of deep learning models in radar echo nowcasting in Wuhan in flood season of 2021

YUAN Kai, PANG Jing, LI Wujie, LI Ming

表4 2021年汛期4种深度学习模型和光流法对不同强度回波预报的平均得分

Tab.4 The average scores of forecast of radar echo with different intensities by four deep learning models and optical flow method in flood season of 2021

算法 CSI POD FAR 20 dBZ 30 dBZ 40 dBZ 20 dBZ 30 dBZ 40 dBZ 20 dBZ 30 dBZ 40 dBZ 光流法 0.43 0.28 0.08 0.55 0.39 0.14 0.36 0.55 0.72 MIM 0.49 0.28 0.06 0.58 0.34 0.07 0.28 0.31 0.14 CrevNet 0.38 0.24 0.05 0.47 0.33 0.08 0.35 0.52 0.51 PhyDNet 0.42 0.21 0.04 0.49 0.32 0.05 0.25 0.26 0.09 PredRNN++ 0.47 0.25 0.05 0.55 0.30 0.06 0.26 0.27 0.10