搜索

x

基于CLDAS和ConvLSTM的海南岛区域网格温度短临预报

Nowcasting of Grid Temperature in Hainan Island Based on CLDAS and ConvLSTM

  • 摘要: 利用中国气象局陆面数据同化系统实时产品数据集(CLDAS),结合ConvLSTM网络提出了新的网格温度短临预报方法,通过实验实现了海南岛区域网格温度的1 h分辨率短临预报.在模型测试当中,12 h连续预报的网格误差超过2 ℃的比率约11.5%,验证了本次研究所提出的模型应用于网格温度短临预报的可能性.

     

    Abstract: In the report, combined with the ConvLSTM network, the China Meteorological Administration's land surface data assimilation system real-time product dataset (CLDAS) was used to propose a new grid temperature short-term forecast method for the Hainan Island area. With the simulation experiment, a one-hour resolution short-term forecast of the grid temperature in Hainan Island was achieved, in which the grid error of the 12-hour continuous forecast exceeds 2 ℃ at a rate of about 11.5%. The results verified the possibility of applying the model proposed in this study to short-term grid temperature prediction.

     

/

返回文章
返回