Using Deep Learning to Improve Short-term Climate Prediction of Summer Precipitation in Southwestern China 

Expand
  • Data Science Research CenterFaculty of ScienceKunming University of Science and TechnologyKunming 650500YunnanChina

Online published: 2025-04-29

Abstract

In recent yearsSouthwestern Chinaincluding YunnanGuizhouSichuanand Chongqinghas been frequently hit by flood disasters caused by climate changeresulting in severe casualties and enormous property losses. The occurrence of these disasters is closely related to abnormal precipitation. Although traditional statistical methods and atmospheric models have achieved certain effectiveness in precipitation forecastingeffective approaches for dealing with the complex spatiotemporal characteristics of precipitation data are still lacking. With the development of machine learning technologythe convolutional long short-term memory networkCon‐ vLSTM),which integrates convolutional neural networks CNNand long short-term memory networks LSTM),has shown outstanding performance in addressing spatiotemporal sequence problemsparticularly in the field of precipitation forecasting. In order to more accurately predict the summer precipitation in the south‐ western region of China for the next yearshort-term climate prediction of precipitation),this study constructed a dataset by integrating global sea surface temperature and precipitation data in Southwestern China. The ConvL‐ STM was used for training and named SST-ConvLSTM. This model not only captures the spatiotemporal characteristics in real precipitation data but also learns some information from global sea surface temperature datathereby enhancing the accuracy of short-term climate prediction of precipitation. The results show that compared to ConvLSTM that does not consider sea surface temperature and a traditional atmospheric modelSST-ConvL‐ STM model has significant advantages in short-term climate prediction of summer precipitation in Southwestern China.1Numericallythe predictions of the SST-ConvLSTM model are closest to the real precipitation datawith similar trend changes. In contrastboth ConvLSTM and the traditional atmospheric model show certain de‐ viations in their predictions.2Spatiallythe SST-ConvLSTM model also performs well. Its predictions are consistent with the spatial distribution of real precipitation data and accurately reflect the spatial distribution of precipitation.3In model evaluationthree evaluation metrics were used to assess the performance of the SSTConvLSTM model. The results show that the SST-ConvLSTM model performs well in all evaluation metrics and achieves the best scores. These findings provide important references and insights for future research on precipitation prediction in Southwestern China.

Cite this article

ZHANG Haoyuan, QIAO Panjie, LIU Wenqi, ZHANG Yongwen . Using Deep Learning to Improve Short-term Climate Prediction of Summer Precipitation in Southwestern China [J]. Plateau Meteorology, 0 : 1 . DOI: 10.7522/j.issn.1000-0534.2024

Outlines

/