{"title":"通过在潜空间进行有指导的迭代预测,有效改进关键天气变量预报","authors":"Shuangliang Li, Siwei Li","doi":"arxiv-2407.19187","DOIUrl":null,"url":null,"abstract":"Weather forecasting refers to learning evolutionary patterns of some key\nupper-air and surface variables which is of great significance. Recently, deep\nlearning-based methods have been increasingly applied in the field of weather\nforecasting due to their powerful feature learning capabilities. However,\nprediction methods based on the original space iteration struggle to\neffectively and efficiently utilize large number of weather variables.\nTherefore, we propose an 'encoding-prediction-decoding' prediction network.\nThis network can efficiently benefit to more related input variables with key\nvariables, that is, it can adaptively extract key variable-related\nlow-dimensional latent feature from much more input atmospheric variables for\niterative prediction. And we construct a loss function to guide the iteration\nof latent feature by utilizing multiple atmospheric variables in corresponding\nlead times. The obtained latent features through iterative prediction are then\ndecoded to obtain the predicted values of key variables in multiple lead times.\nIn addition, we improve the HTA algorithm in \\cite{bi2023accurate} by inputting\nmore time steps to enhance the temporal correlation between the prediction\nresults and input variables. Both qualitative and quantitative prediction\nresults on ERA5 dataset validate the superiority of our method over other\nmethods. (The code will be available at https://github.com/rs-lsl/Kvp-lsi)","PeriodicalId":501166,"journal":{"name":"arXiv - PHYS - Atmospheric and Oceanic Physics","volume":"46 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-07-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Efficiently improving key weather variables forecasting by performing the guided iterative prediction in latent space\",\"authors\":\"Shuangliang Li, Siwei Li\",\"doi\":\"arxiv-2407.19187\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Weather forecasting refers to learning evolutionary patterns of some key\\nupper-air and surface variables which is of great significance. Recently, deep\\nlearning-based methods have been increasingly applied in the field of weather\\nforecasting due to their powerful feature learning capabilities. However,\\nprediction methods based on the original space iteration struggle to\\neffectively and efficiently utilize large number of weather variables.\\nTherefore, we propose an 'encoding-prediction-decoding' prediction network.\\nThis network can efficiently benefit to more related input variables with key\\nvariables, that is, it can adaptively extract key variable-related\\nlow-dimensional latent feature from much more input atmospheric variables for\\niterative prediction. And we construct a loss function to guide the iteration\\nof latent feature by utilizing multiple atmospheric variables in corresponding\\nlead times. The obtained latent features through iterative prediction are then\\ndecoded to obtain the predicted values of key variables in multiple lead times.\\nIn addition, we improve the HTA algorithm in \\\\cite{bi2023accurate} by inputting\\nmore time steps to enhance the temporal correlation between the prediction\\nresults and input variables. Both qualitative and quantitative prediction\\nresults on ERA5 dataset validate the superiority of our method over other\\nmethods. (The code will be available at https://github.com/rs-lsl/Kvp-lsi)\",\"PeriodicalId\":501166,\"journal\":{\"name\":\"arXiv - PHYS - Atmospheric and Oceanic Physics\",\"volume\":\"46 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-07-27\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Atmospheric and Oceanic Physics\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2407.19187\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Atmospheric and Oceanic Physics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2407.19187","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Efficiently improving key weather variables forecasting by performing the guided iterative prediction in latent space
Weather forecasting refers to learning evolutionary patterns of some key
upper-air and surface variables which is of great significance. Recently, deep
learning-based methods have been increasingly applied in the field of weather
forecasting due to their powerful feature learning capabilities. However,
prediction methods based on the original space iteration struggle to
effectively and efficiently utilize large number of weather variables.
Therefore, we propose an 'encoding-prediction-decoding' prediction network.
This network can efficiently benefit to more related input variables with key
variables, that is, it can adaptively extract key variable-related
low-dimensional latent feature from much more input atmospheric variables for
iterative prediction. And we construct a loss function to guide the iteration
of latent feature by utilizing multiple atmospheric variables in corresponding
lead times. The obtained latent features through iterative prediction are then
decoded to obtain the predicted values of key variables in multiple lead times.
In addition, we improve the HTA algorithm in \cite{bi2023accurate} by inputting
more time steps to enhance the temporal correlation between the prediction
results and input variables. Both qualitative and quantitative prediction
results on ERA5 dataset validate the superiority of our method over other
methods. (The code will be available at https://github.com/rs-lsl/Kvp-lsi)