Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction Model
To improve the forecasting accuracy of a radiation emulator in a weather prediction model over the Korean peninsula, the learning rate used in neural network training was automatically optimized using the Sherpa. The Sherpa experiment results were compared with two control simulation results using l...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-04-01
|
| Series: | Atmosphere |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4433/13/5/721 |
| _version_ | 1797501647080914944 |
|---|---|
| author | Park Sa Kim Hwan-Jin Song |
| author_facet | Park Sa Kim Hwan-Jin Song |
| author_sort | Park Sa Kim |
| collection | DOAJ |
| description | To improve the forecasting accuracy of a radiation emulator in a weather prediction model over the Korean peninsula, the learning rate used in neural network training was automatically optimized using the Sherpa. The Sherpa experiment results were compared with two control simulation results using learning rates of 0.0001 and 1 for different batch sizes (full to 500). In the offline evaluation, the Sherpa results showed significant improvements in predicting longwave/shortwave heating rates and fluxes compared to the lowest learning rate results, whereas the improvements compared to the highest learning rate were relatively small because the optimized values by the Sherpa were 0.4756–0.6656. The online evaluation results over one month, which were linked with the weather prediction model, demonstrated the usefulness of Sherpa on a universal performance for the radiation emulator. In particular, at the full batch size, Sherpa contributed to reducing the one-week forecast errors for longwave/shortwave fluxes, skin temperature, and precipitation by 39–125%, 137–159%, and 24–26%, respectively, compared with the two control simulations. Considering the widespread use of parallel learning based on full batch, Sherpa can contribute to producing robust results regardless of batch sizes used in neural network training for developing radiation emulators. |
| first_indexed | 2024-03-10T03:21:32Z |
| format | Article |
| id | doaj.art-7ce72538638648f89e93411328b28599 |
| institution | Directory Open Access Journal |
| issn | 2073-4433 |
| language | English |
| last_indexed | 2024-03-10T03:21:32Z |
| publishDate | 2022-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Atmosphere |
| spelling | doaj.art-7ce72538638648f89e93411328b285992023-11-23T10:01:48ZengMDPI AGAtmosphere2073-44332022-04-0113572110.3390/atmos13050721Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction ModelPark Sa Kim0Hwan-Jin Song1National Institute of Meteorological Sciences, Korea Meteorological Administration, Seogwipo-si 63568, Jeju-do, KoreaNational Institute of Meteorological Sciences, Korea Meteorological Administration, Seogwipo-si 63568, Jeju-do, KoreaTo improve the forecasting accuracy of a radiation emulator in a weather prediction model over the Korean peninsula, the learning rate used in neural network training was automatically optimized using the Sherpa. The Sherpa experiment results were compared with two control simulation results using learning rates of 0.0001 and 1 for different batch sizes (full to 500). In the offline evaluation, the Sherpa results showed significant improvements in predicting longwave/shortwave heating rates and fluxes compared to the lowest learning rate results, whereas the improvements compared to the highest learning rate were relatively small because the optimized values by the Sherpa were 0.4756–0.6656. The online evaluation results over one month, which were linked with the weather prediction model, demonstrated the usefulness of Sherpa on a universal performance for the radiation emulator. In particular, at the full batch size, Sherpa contributed to reducing the one-week forecast errors for longwave/shortwave fluxes, skin temperature, and precipitation by 39–125%, 137–159%, and 24–26%, respectively, compared with the two control simulations. Considering the widespread use of parallel learning based on full batch, Sherpa can contribute to producing robust results regardless of batch sizes used in neural network training for developing radiation emulators.https://www.mdpi.com/2073-4433/13/5/721SherpaWRFradiationemulatorneural network |
| spellingShingle | Park Sa Kim Hwan-Jin Song Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction Model Atmosphere Sherpa WRF radiation emulator neural network |
| title | Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction Model |
| title_full | Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction Model |
| title_fullStr | Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction Model |
| title_full_unstemmed | Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction Model |
| title_short | Usefulness of Automatic Hyperparameter Optimization in Developing Radiation Emulator in a Numerical Weather Prediction Model |
| title_sort | usefulness of automatic hyperparameter optimization in developing radiation emulator in a numerical weather prediction model |
| topic | Sherpa WRF radiation emulator neural network |
| url | https://www.mdpi.com/2073-4433/13/5/721 |
| work_keys_str_mv | AT parksakim usefulnessofautomatichyperparameteroptimizationindevelopingradiationemulatorinanumericalweatherpredictionmodel AT hwanjinsong usefulnessofautomatichyperparameteroptimizationindevelopingradiationemulatorinanumericalweatherpredictionmodel |