A Framework for Four-Dimensional Variational Data Assimilation Based on Machine Learning
The initial field has a crucial influence on numerical weather prediction (NWP). Data assimilation (DA) is a reliable method to obtain the initial field of the forecast model. At the same time, data are the carriers of information. Observational data are a concrete representation of information. DA...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/24/2/264 |
| _version_ | 1797480542942265344 |
|---|---|
| author | Renze Dong Hongze Leng Juan Zhao Junqiang Song Shutian Liang |
| author_facet | Renze Dong Hongze Leng Juan Zhao Junqiang Song Shutian Liang |
| author_sort | Renze Dong |
| collection | DOAJ |
| description | The initial field has a crucial influence on numerical weather prediction (NWP). Data assimilation (DA) is a reliable method to obtain the initial field of the forecast model. At the same time, data are the carriers of information. Observational data are a concrete representation of information. DA is also the process of sorting observation data, during which entropy gradually decreases. Four-dimensional variational assimilation (4D-Var) is the most popular approach. However, due to the complexity of the physical model, the tangent linear and adjoint models, and other processes, the realization of a 4D-Var system is complicated, and the computational efficiency is expensive. Machine learning (ML) is a method of gaining simulation results by training a large amount of data. It achieves remarkable success in various applications, and operational NWP and DA are no exception. In this work, we synthesize insights and techniques from previous studies to design a pure data-driven 4D-Var implementation framework named ML-4DVAR based on the bilinear neural network (BNN). The framework replaces the traditional physical model with the BNN model for prediction. Moreover, it directly makes use of the ML model obtained from the simulation data to implement the primary process of 4D-Var, including the realization of the short-term forecast process and the tangent linear and adjoint models. We test a strong-constraint 4D-Var system with the Lorenz-96 model, and we compared the traditional 4D-Var system with ML-4DVAR. The experimental results demonstrate that the ML-4DVAR framework can achieve better assimilation results and significantly improve computational efficiency. |
| first_indexed | 2024-03-09T22:01:33Z |
| format | Article |
| id | doaj.art-6775497302474dfebb720b018ac0c668 |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-09T22:01:33Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-6775497302474dfebb720b018ac0c6682023-11-23T19:48:36ZengMDPI AGEntropy1099-43002022-02-0124226410.3390/e24020264A Framework for Four-Dimensional Variational Data Assimilation Based on Machine LearningRenze Dong0Hongze Leng1Juan Zhao2Junqiang Song3Shutian Liang4College of Meteorology and Oceanography, National University of Defense Technology, Changsha 410000, ChinaCollege of Meteorology and Oceanography, National University of Defense Technology, Changsha 410000, ChinaCollege of Meteorology and Oceanography, National University of Defense Technology, Changsha 410000, ChinaCollege of Meteorology and Oceanography, National University of Defense Technology, Changsha 410000, ChinaSchool of Resources and Environmental Engineering, Hefei University of Technology, Hefei 230000, ChinaThe initial field has a crucial influence on numerical weather prediction (NWP). Data assimilation (DA) is a reliable method to obtain the initial field of the forecast model. At the same time, data are the carriers of information. Observational data are a concrete representation of information. DA is also the process of sorting observation data, during which entropy gradually decreases. Four-dimensional variational assimilation (4D-Var) is the most popular approach. However, due to the complexity of the physical model, the tangent linear and adjoint models, and other processes, the realization of a 4D-Var system is complicated, and the computational efficiency is expensive. Machine learning (ML) is a method of gaining simulation results by training a large amount of data. It achieves remarkable success in various applications, and operational NWP and DA are no exception. In this work, we synthesize insights and techniques from previous studies to design a pure data-driven 4D-Var implementation framework named ML-4DVAR based on the bilinear neural network (BNN). The framework replaces the traditional physical model with the BNN model for prediction. Moreover, it directly makes use of the ML model obtained from the simulation data to implement the primary process of 4D-Var, including the realization of the short-term forecast process and the tangent linear and adjoint models. We test a strong-constraint 4D-Var system with the Lorenz-96 model, and we compared the traditional 4D-Var system with ML-4DVAR. The experimental results demonstrate that the ML-4DVAR framework can achieve better assimilation results and significantly improve computational efficiency.https://www.mdpi.com/1099-4300/24/2/264numerical weather predictionfour-dimensional variational assimilationmachine learningtangent linear and adjoint models |
| spellingShingle | Renze Dong Hongze Leng Juan Zhao Junqiang Song Shutian Liang A Framework for Four-Dimensional Variational Data Assimilation Based on Machine Learning Entropy numerical weather prediction four-dimensional variational assimilation machine learning tangent linear and adjoint models |
| title | A Framework for Four-Dimensional Variational Data Assimilation Based on Machine Learning |
| title_full | A Framework for Four-Dimensional Variational Data Assimilation Based on Machine Learning |
| title_fullStr | A Framework for Four-Dimensional Variational Data Assimilation Based on Machine Learning |
| title_full_unstemmed | A Framework for Four-Dimensional Variational Data Assimilation Based on Machine Learning |
| title_short | A Framework for Four-Dimensional Variational Data Assimilation Based on Machine Learning |
| title_sort | framework for four dimensional variational data assimilation based on machine learning |
| topic | numerical weather prediction four-dimensional variational assimilation machine learning tangent linear and adjoint models |
| url | https://www.mdpi.com/1099-4300/24/2/264 |
| work_keys_str_mv | AT renzedong aframeworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT hongzeleng aframeworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT juanzhao aframeworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT junqiangsong aframeworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT shutianliang aframeworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT renzedong frameworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT hongzeleng frameworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT juanzhao frameworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT junqiangsong frameworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning AT shutianliang frameworkforfourdimensionalvariationaldataassimilationbasedonmachinelearning |