Tropical cyclone intensity forecasting using model knowledge guided deep learning model
This paper developed a deep learning (DL) model for forecasting tropical cyclone (TC) intensity in the Northwest Pacific. A dataset containing 20 533 synchronized and collocated samples was assembled, which included ERA5 reanalysis data as well as satellite infrared (IR) imagery, covering the period...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2024-01-01
|
| Series: | Environmental Research Letters |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/1748-9326/ad1bde |
| _version_ | 1797352028817588224 |
|---|---|
| author | Chong Wang Xiaofeng Li Gang Zheng |
| author_facet | Chong Wang Xiaofeng Li Gang Zheng |
| author_sort | Chong Wang |
| collection | DOAJ |
| description | This paper developed a deep learning (DL) model for forecasting tropical cyclone (TC) intensity in the Northwest Pacific. A dataset containing 20 533 synchronized and collocated samples was assembled, which included ERA5 reanalysis data as well as satellite infrared (IR) imagery, covering the period from 1979 to 2021. The u -, v - and w -components of wind, sea surface temperature, IR satellite imagery, and historical TC information were selected as the model inputs. Then, a TC-intensity-forecast-fusion (TCIF-fusion) model was developed, in which two special branches were designed to learn multi-factor information to forecast 24 h TC intensity. Finally, heatmaps capturing the model’s insights are generated and applied to the original input data, creating an enhanced input set that results in more accurate forecasting. Employing this refined input, the heatmaps (model knowledge) were used to guide TCIF-fusion model modeling, and the model-knowledge-guided TCIF-fusion model achieved a 24 h forecast error of 3.56 m s ^−1 for Northwest Pacific TCs spanning 2020–2021. The results show that the performance of our method is significantly better than the official subjective prediction and advanced DL methods in forecasting TC intensity by 4% to 22%. Additionally, compared to operational approaches, model-guided knowledge methods can better forecast the intensity of landfalling TCs. |
| first_indexed | 2024-03-08T13:09:13Z |
| format | Article |
| id | doaj.art-5eac6deda38f4cd1933c7f449cb3bf88 |
| institution | Directory Open Access Journal |
| issn | 1748-9326 |
| language | English |
| last_indexed | 2024-03-08T13:09:13Z |
| publishDate | 2024-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Environmental Research Letters |
| spelling | doaj.art-5eac6deda38f4cd1933c7f449cb3bf882024-01-18T15:03:03ZengIOP PublishingEnvironmental Research Letters1748-93262024-01-0119202400610.1088/1748-9326/ad1bdeTropical cyclone intensity forecasting using model knowledge guided deep learning modelChong Wang0https://orcid.org/0000-0002-8275-8450Xiaofeng Li1https://orcid.org/0000-0001-7038-5119Gang Zheng2Key Laboratory of Ocean Circulation and Waves, Institute of Oceanology , Qingdao, People’s Republic of China; University of the Chinese Academy of Sciences , Beijing, People’s Republic of ChinaKey Laboratory of Ocean Circulation and Waves, Institute of Oceanology , Qingdao, People’s Republic of ChinaState Key Laboratory of Satellite Ocean Environment Dynamics, Second Institute of Oceanography, Ministry of Natural Resources , Hangzhou, People’s Republic of ChinaThis paper developed a deep learning (DL) model for forecasting tropical cyclone (TC) intensity in the Northwest Pacific. A dataset containing 20 533 synchronized and collocated samples was assembled, which included ERA5 reanalysis data as well as satellite infrared (IR) imagery, covering the period from 1979 to 2021. The u -, v - and w -components of wind, sea surface temperature, IR satellite imagery, and historical TC information were selected as the model inputs. Then, a TC-intensity-forecast-fusion (TCIF-fusion) model was developed, in which two special branches were designed to learn multi-factor information to forecast 24 h TC intensity. Finally, heatmaps capturing the model’s insights are generated and applied to the original input data, creating an enhanced input set that results in more accurate forecasting. Employing this refined input, the heatmaps (model knowledge) were used to guide TCIF-fusion model modeling, and the model-knowledge-guided TCIF-fusion model achieved a 24 h forecast error of 3.56 m s ^−1 for Northwest Pacific TCs spanning 2020–2021. The results show that the performance of our method is significantly better than the official subjective prediction and advanced DL methods in forecasting TC intensity by 4% to 22%. Additionally, compared to operational approaches, model-guided knowledge methods can better forecast the intensity of landfalling TCs.https://doi.org/10.1088/1748-9326/ad1bdetropical cyclone intensity forecastdeep learningmodel knowledge |
| spellingShingle | Chong Wang Xiaofeng Li Gang Zheng Tropical cyclone intensity forecasting using model knowledge guided deep learning model Environmental Research Letters tropical cyclone intensity forecast deep learning model knowledge |
| title | Tropical cyclone intensity forecasting using model knowledge guided deep learning model |
| title_full | Tropical cyclone intensity forecasting using model knowledge guided deep learning model |
| title_fullStr | Tropical cyclone intensity forecasting using model knowledge guided deep learning model |
| title_full_unstemmed | Tropical cyclone intensity forecasting using model knowledge guided deep learning model |
| title_short | Tropical cyclone intensity forecasting using model knowledge guided deep learning model |
| title_sort | tropical cyclone intensity forecasting using model knowledge guided deep learning model |
| topic | tropical cyclone intensity forecast deep learning model knowledge |
| url | https://doi.org/10.1088/1748-9326/ad1bde |
| work_keys_str_mv | AT chongwang tropicalcycloneintensityforecastingusingmodelknowledgeguideddeeplearningmodel AT xiaofengli tropicalcycloneintensityforecastingusingmodelknowledgeguideddeeplearningmodel AT gangzheng tropicalcycloneintensityforecastingusingmodelknowledgeguideddeeplearningmodel |