A precipitation forecast method based on transfer learning and Long Short Term Memory
A precipitation forecasting method based on transfer learning and Long Short-Term Memory (LSTM) is proposed to provide an objective reference for intelligent grid heavy precipitation forecasting. Transfer learning is a machine learning method that can transfer knowledge learned from the source domai...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | zho |
Published: |
Editorial Office of Torrential Rain and Disasters
2024-02-01
|
Series: | 暴雨灾害 |
Subjects: | |
Online Access: | http://www.byzh.org.cn/cn/article/doi/10.12406/byzh.2023-118 |
_version_ | 1797272531062751232 |
---|---|
author | Tianwen HUANG Fei JIAO Zhifang WU |
author_facet | Tianwen HUANG Fei JIAO Zhifang WU |
author_sort | Tianwen HUANG |
collection | DOAJ |
description | A precipitation forecasting method based on transfer learning and Long Short-Term Memory (LSTM) is proposed to provide an objective reference for intelligent grid heavy precipitation forecasting. Transfer learning is a machine learning method that can transfer knowledge learned from the source domain to the target domain for application. LSTM is a deep learning model that can handle long-term dependencies in sequence data and can remember long and short periods. In this study, the hourly observation data (rainfall, temperature, air pressure, relative humidity, wind direction, wind speed) from 2009 to 2022 of 6 meteorological observation stations in Zhaoqing City is used. The Gaoyao National Meteorological Observatory is selected as the target domain and the other 5 national meteorological observatories as the source domain, and the transfer learning method is used to transfer the source domain and correct missing values in the target domain. Then the complete training samples are classified in the target domain. Then, the deep learning methods are applied to establish the univariate LSTM daily rainfall prediction models and the multivariate LSTM hourly rainfall prediction models for the target domain, respectively. The daily and hourly rainfall forecast in the target domain for the year 2022 is compared with the actual observations. The results are as follows:
(1) For the daily precipitation forecast of clear rain, the univariate LSTM method from January to February, June, and October to December can achieve an accuracy of over 80%, while for the hourly precipitation forecast of clear rain, the accuracy of the multivariate LSTM precipitation forecast method in March, June, August, and December can be over 80%. (2) The univariate LSTM method can only forecast precipitation with a 24-hour rainfall over 50 mm in June. The multivariate LSTM method can forecast precipitation with a 1-hour rainfall over 20 mm in March, May, and June to August, with the TS score in March and June being higher than 25%. |
first_indexed | 2024-03-07T14:30:41Z |
format | Article |
id | doaj.art-0d71be1f4144488793b9fdf9903fff3c |
institution | Directory Open Access Journal |
issn | 2097-2164 |
language | zho |
last_indexed | 2024-03-07T14:30:41Z |
publishDate | 2024-02-01 |
publisher | Editorial Office of Torrential Rain and Disasters |
record_format | Article |
series | 暴雨灾害 |
spelling | doaj.art-0d71be1f4144488793b9fdf9903fff3c2024-03-06T04:55:30ZzhoEditorial Office of Torrential Rain and Disasters暴雨灾害2097-21642024-02-01431455310.12406/byzh.2023-118byzh-43-1-45A precipitation forecast method based on transfer learning and Long Short Term MemoryTianwen HUANG0Fei JIAO1Zhifang WU2Zhaoqing Meteorological Bureau, Zhaoqing 526040Zhaoqing University, Zhaoqing 526061Guangdong Meteorological Observatory, Guangzhou 510641A precipitation forecasting method based on transfer learning and Long Short-Term Memory (LSTM) is proposed to provide an objective reference for intelligent grid heavy precipitation forecasting. Transfer learning is a machine learning method that can transfer knowledge learned from the source domain to the target domain for application. LSTM is a deep learning model that can handle long-term dependencies in sequence data and can remember long and short periods. In this study, the hourly observation data (rainfall, temperature, air pressure, relative humidity, wind direction, wind speed) from 2009 to 2022 of 6 meteorological observation stations in Zhaoqing City is used. The Gaoyao National Meteorological Observatory is selected as the target domain and the other 5 national meteorological observatories as the source domain, and the transfer learning method is used to transfer the source domain and correct missing values in the target domain. Then the complete training samples are classified in the target domain. Then, the deep learning methods are applied to establish the univariate LSTM daily rainfall prediction models and the multivariate LSTM hourly rainfall prediction models for the target domain, respectively. The daily and hourly rainfall forecast in the target domain for the year 2022 is compared with the actual observations. The results are as follows: (1) For the daily precipitation forecast of clear rain, the univariate LSTM method from January to February, June, and October to December can achieve an accuracy of over 80%, while for the hourly precipitation forecast of clear rain, the accuracy of the multivariate LSTM precipitation forecast method in March, June, August, and December can be over 80%. (2) The univariate LSTM method can only forecast precipitation with a 24-hour rainfall over 50 mm in June. The multivariate LSTM method can forecast precipitation with a 1-hour rainfall over 20 mm in March, May, and June to August, with the TS score in March and June being higher than 25%.http://www.byzh.org.cn/cn/article/doi/10.12406/byzh.2023-118precipitation forecasttime seriestransfer learninglstmdeep learning |
spellingShingle | Tianwen HUANG Fei JIAO Zhifang WU A precipitation forecast method based on transfer learning and Long Short Term Memory 暴雨灾害 precipitation forecast time series transfer learning lstm deep learning |
title | A precipitation forecast method based on transfer learning and Long Short Term Memory |
title_full | A precipitation forecast method based on transfer learning and Long Short Term Memory |
title_fullStr | A precipitation forecast method based on transfer learning and Long Short Term Memory |
title_full_unstemmed | A precipitation forecast method based on transfer learning and Long Short Term Memory |
title_short | A precipitation forecast method based on transfer learning and Long Short Term Memory |
title_sort | precipitation forecast method based on transfer learning and long short term memory |
topic | precipitation forecast time series transfer learning lstm deep learning |
url | http://www.byzh.org.cn/cn/article/doi/10.12406/byzh.2023-118 |
work_keys_str_mv | AT tianwenhuang aprecipitationforecastmethodbasedontransferlearningandlongshorttermmemory AT feijiao aprecipitationforecastmethodbasedontransferlearningandlongshorttermmemory AT zhifangwu aprecipitationforecastmethodbasedontransferlearningandlongshorttermmemory AT tianwenhuang precipitationforecastmethodbasedontransferlearningandlongshorttermmemory AT feijiao precipitationforecastmethodbasedontransferlearningandlongshorttermmemory AT zhifangwu precipitationforecastmethodbasedontransferlearningandlongshorttermmemory |